People v. Peterson

Full title: THE PEOPLE, Plaintiff and Respondent, v. SCOTT LEE PETERSON, Defendant and…

Court: SUPREME COURT OF CALIFORNIA

Date published: Aug 24, 2020

10 Cal.5th 409 (Cal. 2020)

268 Cal. Rptr. 3d 56

472 P.3d 382

Opinion

S132449

08-24-2020

The PEOPLE, Plaintiff and Respondent, v. Scott Lee PETERSON, Defendant and Appellant.

Cliff Gardner, under appointment by the Supreme Court; Catherine White and Lazuli Whitt for Defendant and Appellant. Kamala D. Harris and Xavier Becerra, Attorneys General, Gerald A. Engler and Ronald S. Matthias, Assistant Attorneys General, Glenn R. Pruden and Donna M. Provenzano, Deputy Attorneys General, for Plaintiff and Respondent.

Opinion of the Court by Kruger, J.

Cliff Gardner, under appointment by the Supreme Court; Catherine White and Lazuli Whitt for Defendant and Appellant.

Kamala D. Harris and Xavier Becerra, Attorneys General, Gerald A. Engler and Ronald S. Matthias, Assistant Attorneys General, Glenn R. Pruden and Donna M. Provenzano, Deputy Attorneys General, for Plaintiff and Respondent.

Opinion of the Court by Kruger, J. A jury convicted defendant Scott Lee Peterson of one count of first degree murder for killing his wife, Laci Peterson, and one count of second degree murder for killing their unborn son. It found true the special circumstance that Peterson had committed multiple murders. At the penalty phase, the jury returned a verdict of death. This appeal is automatic. ( Pen. Code, § 1239, subd. (b).)

Peterson contends his trial was flawed for multiple reasons, beginning with the unusual amount of pretrial publicity that surrounded the case. We reject Peterson's claim that he received an unfair trial as to guilt and thus affirm his convictions for murder. But before the trial began, the trial court made a series of clear and significant errors in jury selection that, under long-standing United States Supreme Court precedent, undermined Peterson's right to an impartial jury at the penalty phase. While a court may dismiss a prospective juror as unqualified to sit on a capital case if the juror's views on capital punishment would substantially impair his or her ability to follow the law, a juror may not be dismissed merely because he or she has expressed opposition to the death penalty as a general matter. (See Witherspoon v. Illinois (1968) 391 U.S. 510, 88 S.Ct. 1770, 20 L.Ed.2d 776 ; Wainwright v. Witt (1985) 469 U.S. 412, 105 S.Ct. 844, 83 L.Ed.2d 841.) Here, the trial court erroneously dismissed many prospective jurors because of written questionnaire responses expressing opposition to the death penalty, even though the jurors gave no indication that their views would prevent them from following the law — and, indeed, specifically attested in their questionnaire responses that they would have no such difficulty. Under United States Supreme Court precedent, these errors require us to reverse the death sentence in this case. ( Gray v. Mississippi (1987) 481 U.S. 648, 107 S.Ct. 2045, 95 L.Ed.2d 622 ; see People v. Riccardi (2012) 54 Cal.4th 758, 778, 144 Cal.Rptr.3d 84, 281 P.3d 1.) On remand, the People may retry the penalty phase if they so choose.

I. FACTUAL AND PROCEDURAL BACKGROUND

A. Guilt Phase Trial

1. Prosecution Evidence

Peterson and Laci Rocha met in San Luis Obispo, where Laci was attending college and Peterson was working in a restaurant. They married in 1997. They opened and ran a restaurant together in San Luis Obispo. In 2000, they moved to Modesto and bought a house. Laci took a job as a substitute teacher, while Peterson ran a start-up fertilizer company named TradeCorp U.S.A. out of a leased warehouse. Some years after the two married, Laci became pregnant; the baby — whom the couple had named Conner — was due in February 2003.¹

1 For clarity, we generally will refer to Laci Peterson (neé Rocha) and Conner by their first names. We will also sometimes refer to members of Laci's immediate family — her mother, Sharon Rocha; her sister, Amy Rocha; and her brother, Brent Rocha — by their first names. No disrespect is intended to any of these individuals.

On December 23, 2002, Laci went grocery shopping around midday. She also had a prenatal medical checkup. In the late afternoon, both Laci and Peterson went to a salon where Laci's sister, Amy Rocha, worked. Amy mentioned that she had ordered a gift basket for a family member that needed to be picked up the next day by 3:00 p.m. Peterson volunteered to get it for her, as he was going to be golfing nearby. Peterson also invited Amy to dinner, but she declined because she had prior plans. That night, Laci and her mother, Sharon, spoke on the phone and confirmed that Laci and Peterson would join Sharon and Sharon's longtime partner, Ron Grantski, for dinner the following night, Christmas Eve.

At 10:18 the following morning, a neighbor, Karen Servas, saw the Petersons' dog, McKenzie, wandering unaccompanied on the street, wearing his leash. Peterson's truck was gone; Laci's car was still in their driveway. There were no signs of activity at the house, so Servas put McKenzie in the Petersons' backyard and closed the gate. That afternoon, Grantski tried to reach Laci, without success. Around 3:45 p.m., Amy received a call that her gift basket had not been picked up. She was unable to reach Peterson. Neighbors reported Peterson's truck still absent at 4:05 p.m., but back by 5:30 p.m.

At around 5:15 p.m., Peterson called Sharon and asked if Laci was there. He described Laci as "missing." Sharon suggested he check with friends and neighbors. Peterson called Sharon back shortly afterwards and reported the people he had spoken to had not seen Laci either. Sharon told Grantski to call the police. Officers soon met Peterson, Sharon, and Grantski at a nearby park. Neighbors and other relatives gathered at the park as well. Grantski spoke with Peterson and asked if he had gone golfing that day. Peterson said he had changed his mind and gone fishing instead. Told what time Peterson had gone, Grantski suggested it was an unusually late time to be fishing. Peterson walked off without responding. Peterson told a cousin of Sharon's and two neighbors that he had been golfing all day. He volunteered to Sandy Rickard, a friend of Sharon's, that he would not be surprised if the police found blood in his truck because he cut his hands all the time.

Police inspected the Peterson home. There were no signs of forced entry, nothing appeared missing, and Laci's purse was still there. Peterson told officers he and Laci had watched television that morning, and Laci had planned to walk the dog and go grocery shopping. Peterson decided to go fishing in the San Francisco Bay. He went to his company warehouse where he stored a boat, drove to the Berkeley Marina, fished for two hours, and quit because the day was cold and rainy. He tried calling Laci on the home phone and her cell phone but did not reach her. Peterson got home around 4:30 p.m. He washed his clothes, ate some pizza, and then called Sharon to track down Laci.

Officer Matthew Spurlock asked what time Peterson was fishing. He also asked what Peterson was fishing for and what lure he used. According to Spurlock and Officer Derrick Letsinger, Peterson gave slow and initially noncommittal answers. He "really didn't give a responsive time" and, when asked what he was fishing for, paused, gave a blank look, and "mumbled some stuff" without really answering. Peterson likewise responded with a blank look when asked about his lure, but after some delay came up with a size and color description.

Detective Allen Brocchini was called to the Peterson home. He found wet towels on top of the washing machine. Peterson explained that he had taken them out so that he could wash the clothes he had worn that day. Inside the washing machine were Peterson's jeans, shirt, and green pullover jacket. In the bedroom, officers observed a laundry hamper nearly full of clothes. With consent, Detective Brocchini examined Peterson's truck and saw large patio umbrellas and a tarp in the truck bed. Inside the truck cab, he found a fishing rod and a bag containing a package of unused fishing lures and a receipt indicating the items had all been purchased on December 20. Peterson handed him a Berkeley Marina parking receipt that indicated Peterson had entered at 12:54 p.m. On the backseat was a camouflage jacket Peterson said he had worn fishing that day. Brocchini and Peterson then went to Peterson's warehouse. There, Brocchini observed what he described as a "homemade anchor" made of concrete in Peterson's boat, but no long rope to attach it to the boat.

Peterson agreed to a further interview at the Modesto police station. Peterson repeated that Laci had planned to walk the dog and go grocery shopping. For his part, Peterson decided to go fishing because it was too cold to golf. He went to his warehouse, then to the Berkeley Marina around 1:00 p.m., and fished for 90 minutes near an area that was later identified as Brooks Island.² Peterson did not pack a lunch or stop to eat on the way to or from the marina. On the way back, Peterson called Laci on their home phone and left two messages on her cell phone.³ He dropped off his boat at the warehouse and went home. Peterson told officers that there were no problems in his marriage.

2 Peterson said he left the house with no jacket on, put on a green pullover jacket, and then put the camouflage jacket over that when it started raining. The camouflage jacket, when Detective Brocchini saw it in Peterson's truck a few hours later, was dry.

3 When messages on Laci's cell phone were played, only one voice message from Peterson was found.

Peterson had a followup interview with Detective Craig Grogan and an investigator from the state's Department of Justice on Christmas Day, December 25. Peterson explained that he had never fished on the San Francisco Bay before but wanted to test out his boat. He troll fished⁴ for an hour on the way out to Brooks Island from the marina dock. Peterson suggested Laci might have been robbed of her jewelry by a transient and then kidnapped. He denied being involved in an affair with anyone. Later that day, Peterson called Detective Brocchini to check on the investigation. He asked if the police would be using cadaver dogs⁵ to search for Laci. Brocchini explained that they would not, because no one assumed Laci was dead.

4 Troll fishing involves dragging a baited line through the water.

5 Cadaver dogs are trained to scent and alert to decomposing human remains.

In the days after Christmas, the Modesto Police Department executed search warrants on the Peterson home and Peterson's warehouse. Police confirmed that there had been no forced entry at the house. None of Laci's jewelry was missing, other than one pair of diamond earrings. Traces of Peterson's blood were found on a comforter in the master bedroom. In sheds in the backyard, police found the cover to Peterson's boat, smelling heavily of gasoline,⁶ as well as a blue tarp. The boat cover had chunks of concrete in it. In Peterson's truck, police found additional spots of Peterson's blood. Peterson explained that he had cut his hand on the truck door. Police found small chunks of cement and a claw hammer with cement powder on it in the truck's bed.

6 At trial, evidence was introduced that gasoline makes it extremely difficult for trailing dogs to identify a human scent.

At the warehouse, the police inspected the boat and found a pair of pliers under the middle seat. The pliers had hair clamped in their teeth. Subsequent mitochondrial DNA testing of a hair fragment determined that the hair matched a reference sample from Sharon, which meant that its donor had the same maternal lineage as Sharon. The hair did not match Peterson's.

During the search of the Peterson home, articles that Laci would have touched were collected to give to trailing dogs to enable them to search for Laci's scent. These included a slipper and a pair of sunglasses. On December 28, four days after Laci disappeared, Trimble, a trailing dog, was presented Laci's sunglasses at the Berkeley Marina. Trimble alerted to Laci's scent along a path that led out onto a dock and ended at the water.

On December 30, a woman named Amber Frey contacted the police after a friend advised her that Peterson, who she thought was unmarried with no children, and with whom she had been having a relationship since November, was connected to the disappearance of his pregnant wife. Frey and Peterson had had their first date on November 20 and had immediately become sexually intimate. Their relationship had progressed to the point where Peterson had stayed over at Frey's home, picked up Frey's young daughter from daycare, gone to various parties with Frey, alone and with her daughter, picked out a Christmas tree with Frey, and discussed their views on having children. Peterson initially told Frey he had never been married and had no children, but on December 6 a friend of Frey's discovered otherwise and gave him an ultimatum to tell Frey by December 9 or else she would. On December 9, Peterson explained to Frey that he had in fact been married, but had "lost" his wife, and the upcoming holidays would be his first without her. On December 15, Peterson told Frey he would be in Europe on business through the rest of the month and much of January. On December 23, after Frey asked where she should send him things while he was away, Peterson rented a private mailbox to which Frey could send letters. He called Frey that day, claiming to be in Maine duck hunting with his father, and again on Christmas Day, supposedly still from Maine.

After meeting with police, Frey agreed to cooperate and tape future calls from Peterson. On New Year's Eve, Peterson called Frey from a vigil for Laci, claiming to be in Paris watching fireworks over the Eiffel Tower. He called Frey again on New Year's Day and in the days after, maintaining the fiction that he was in Europe. On January 3, 2003, when police confronted Peterson with a picture of himself and Frey, Peterson denied that it was him in the photo and that he was having an affair.

On January 6, at the instigation of police, Frey dropped hints that a friend had learned the truth and would tell her in a matter of hours.⁷ In response, Peterson finally admitted to Frey that he was married to a woman named Laci and had been in Modesto the entire time. The next day, when Frey asked if Peterson had told Laci about her, Peterson said he had and that Laci was "fine" with his having an affair. Later in the month, once news media had made the affair public, Peterson, in an interview aired nationwide, repeated that Laci was fine with his having an affair and said he had disclosed the affair to the police immediately. On February 19, at the direction of police investigators, Frey told Peterson they should stop talking.

7 Laci's disappearance swiftly became the subject of widespread media attention. To maintain the pretense that she did not know the truth about Peterson yet, Frey denied watching the news.

In January, after obtaining a warrant, police placed a surveillance camera outside the Peterson home and GPS tracking devices on Peterson's vehicles, including a series of cars and trucks Peterson rented for a few days at a time. Surveillance data from these devices and visual surveillance by the police showed Peterson driving the approximately 90 miles from his home to the Berkeley Marina at least five times in January, each time using a different vehicle. On January 5, he drove there in a gray Subaru, spent five or ten minutes, and left. On January 6, he returned to the marina in a red Honda and again spent only a few minutes. On January 9, Peterson drove there in a white pickup truck. On January 11, after determining that their cover had been blown, the Modesto Police Department shut down surveillance at the Peterson home. Nonetheless, from tracking data supplied by the automobiles' manufacturers, police were able to determine that Peterson returned to the marina on January 26 in Laci's Land Rover and on January 27 in a rented Dodge Dakota.

During the same period, Peterson began to make various changes to his work and living situations. On January 13, Peterson gave 30 days' notice that he was terminating his warehouse lease, which was not up until October. That same month, he started discussions to sell the Peterson home. On January 29, Peterson sold Laci's car, trading it in for a Dodge Dakota pickup truck. On January 30, he stopped home mail delivery and directed that all mail be delivered to the post office box he had set up on December 23. The nursery for Conner was converted into storage space. On February 18, satellite television service to the Peterson home was canceled; the satellite company's records indicated the customer had explained he was moving overseas.

A $500,000 reward was posted by a private foundation for information leading to Laci's return. For months, no useful leads turned up. Even when potentially promising sightings were reported, Peterson appeared to show little interest. For example, the prosecution presented evidence collected from an authorized wiretap of Peterson's phone that showed he took days to follow up with police about a possible sighting in Washington, though he told others — including his mother — that he had followed up with police immediately. Peterson similarly told a business associate he was waiting near the airport in case he needed to fly up to Washington, though at the time, Peterson was not near any airport.

In mid-April, a significant storm hit the San Francisco Bay Area. On April 13, after the storm had passed, a couple walking their dog came upon Conner's badly decomposed body, apparently washed ashore along with other storm debris. The location was just over a mile from the southern tip of Brooks Island. The next morning, Laci's body was discovered on the shoreline at Point Isabel, south of Conner's body and again just over a mile from Brooks Island. Laci's body had barnacles and duct tape on it. From residual clumps of fabric, it was possible to determine that she had been wearing light-colored capris. The clothing was consistent with the recollection of Amy, who testified that Laci was wearing cream-colored pants when she last saw her sister on December 23. It was, however, inconsistent with the recollection of Peterson, who told police that Laci was wearing black pants when he last saw her on December 24. Days later, DNA testing confirmed the identities of the two bodies.

Dr. Brian Peterson (no relation to the Petersons) performed autopsies on both bodies. Laci's body had several parts missing, including her head, forearms, and one lower leg. Changes to the tissue suggested her body had been in a marine environment. Tidal action and marine animals could explain the missing body parts. Laci's uterus was still enlarged, her birth canal was closed, and there was no evidence of a Caesarian-section birth, which indicated she had died while still pregnant. Dr. Allison Galloway, a forensic anthropologist given the remains to analyze, testified that Laci had been in water for three to six months. Given the condition of the body, it was not possible to determine a cause of death. Conner's body was intact. There was tape on his neck but no associated injuries, which led Dr. Peterson to conclude the tape was just debris that had become attached to Conner after his death. There was no clothing on the body. Conner still had part of his umbilical cord and meconium in his intestines, which indicated he had died before birth. Based on his size and the softness of his tissue, Dr. Peterson opined that Conner must have remained protected inside Laci's uterus for some time after death; had he spent a significant period of time exposed in the water, he would have been eaten by marine animals.

As mentioned, Laci had had a prenatal checkup on December 23. Based on ultrasounds, Conner was at 32 to 33 weeks of gestation. Post-mortem measurements of his bone growth allowed Dr. Greggory DeVore to estimate Conner's date of death as falling between December 21 and December 24, with an average of December 23. Both Dr. Esther Towder, Laci's gynecologist who conducted the December 23 checkup, and Dr. Peterson testified that based on his age and health, Conner would have survived had he been born that day.

Dr. Ralph Cheng, a hydrologist with the United States Geological Survey, was contacted by the Modesto Police Department in February, while Laci was still missing, and again in May, after she and Conner had been found. The first time, he was asked to assume that Laci's body had been dumped with weights into the San Francisco Bay and, based on that assumption, to estimate where the body might be found. The second time, after the bodies had been found, Dr. Cheng was asked to estimate where they might have originated. He was able to estimate a location for Conner near the southern tip of Brooks Island, but no likely location for Laci. Divers searching the bay at Dr. Cheng's target location were unable to find any relevant evidence.

On April 12, the day before Conner's body was found, Peterson bought a car using his mother's name, Jacqueline, as his own, providing a fake driver's license number, and paying $3,600 in cash. He had grown a goatee and mustache and appeared to have dyed his hair. On April 15, when Sharon called him about the discovery of the (as-yet unidentified) bodies of Conner and Laci, Peterson did not return her call. Believing Peterson might flee, police arrested him on April 18. When arrested, Peterson had nearly $15,000 in cash, foreign currency, two drivers' licenses (his own and his brother's), a family member's credit card, camping gear, considerable extra clothing, and multiple cell phones.

The prosecution introduced evidence concerning the Petersons' finances. The Petersons' expenses were high in relation to their current income. TradeCorp U.S.A. had never been profitable, posting operating losses of $40,000 and $200,000 in consecutive years; the company was not meeting sales goals, and it owed its parent company $190,000. Peterson had signed multiple credit card applications in the company's name containing misrepresentations as to the company's income.

In fall 2002, Laci inherited jewelry and, at Peterson's request, had some of the items appraised. They were valued at more than $100,000. Computers seized from the Peterson home and the warehouse showed e-mails sent from an account bearing the username "slpete1" discussing the sale of jewelry, and eBay records likewise showed Peterson had posted jewelry items for sale. Laci also stood to inherit one-third of the proceeds from the sale of her grandfather's house, an interest estimated to be worth around $140,000. Laci's interest would terminate on her death, with no right of survivorship to Peterson, but it was unclear whether Peterson was aware of the limitation; Brent, the cotrustee of the grandparents' estate, had not told Peterson about the provision.

The prosecution also submitted additional background concerning Peterson's fishing. Computers seized from the Peterson home and the warehouse showed that someone had conducted searches of classified advertisements for boats on December 7, the day after Peterson learned he would no longer be able to conceal his marriage from Frey. That same day, Peterson called Bruce Peterson (no relation) about a boat for sale. Peterson inspected the boat the next day and bought it on December 9, without the anchors that came with the boat. Peterson never registered the boat, nor did he ever mention the purchase to his father; to Grantski, an avid fisherman who had invited Peterson to fish several times; to other members of the Rocha family; or to his friend Gregory Reed, with whom he frequently discussed fishing. Review of the seized computers' browser histories also showed someone conducting searches on December 8 for boat ramps on the Pacific Ocean, then examining nautical charts, currents, and maps for the Berkeley Marina and San Francisco Bay, including the area around Brooks Island. There were also visits to fishing-related websites.

December 24, the day Peterson said he was fishing, was gray, damp, and cold with a bit of wind. Few people were at the Berkeley Marina. When questioned by police, Peterson would not say what he was hoping to catch, but the fishing searches performed from his computer earlier in the month had included searches relating to sturgeon and striped bass. Angelo Cuanang, a published author on fishing in the San Francisco Bay who was accepted by the court as an expert fisherman, testified that Brooks Island was the wrong place to seek sturgeon, which congregated in a different part of the bay that time of year. Sturgeon also preferred live bait to lures, and Peterson's rod was too weak to catch them. Anchoring was essential to reel in sturgeon; the homemade cement anchor in his boat would have been inadequate. Finally, it was illegal to troll for sturgeon, as Peterson claimed to have done. Peterson's lures and the time of year he was fishing were also wrong for catching striped bass. The prosecution's theory was as follows: Peterson killed Laci sometime on the night of December 23 or morning of December 24. On the morning of the 24th, Peterson let their dog McKenzie out with his leash on to make it appear something had happened while Laci was walking him. He wrapped Laci's body in a tarp in the bed of his truck, covered her with the patio umbrellas, drove to the warehouse, and then moved her body into his boat.⁸ He drove to the Berkeley Marina, motored out to an area near Brooks Island, and slipped her body, attached to homemade concrete weights like the homemade anchor Peterson had made, into the bay.⁹ Peterson then returned to Modesto, dropped off the boat at the warehouse, put the boat cover out back under a leaky gas blower so that any scent would be obscured, washed his clothes, and proceeded with the ruse that Laci was missing, hoping her body would never be discovered.

8 The prosecution introduced photographs of a district attorney's office employee, at approximately the same stage of pregnancy and weight as Laci at her disappearance, fitting into the bottom of Peterson's boat.

9 Through an engineer for the company that manufactured the boat, the prosecution introduced stability tests the boat model underwent to obtain certification before it was sold.

2. Defense Evidence

The defense argued the police had not diligently pursued whether a person or persons other than Peterson were more likely responsible for Laci's disappearance and murder. The defense presented evidence that a burglary had occurred on the Petersons' street the week of her disappearance and argued that the police failed adequately to follow up on whether that burglary had any connection to Laci's disappearance. It also presented evidence that a stranger had gone to several houses on December 23 asking for money and, one neighbor thought, casing houses for burglaries, and so might have had something to do with her disappearance. Testimony was presented that the same neighbor, walking with a police officer on Christmas Day to look for the stranger, had seen a pair of sandals lying in the road 150 feet from the Petersons' home; the neighbor wondered at the time if they might have any connection to Laci's disappearance, but the officer just left them there. To support the possibility of a third party's involvement, the defense challenged the prosecution's theory that Conner died December 23 or 24, presenting its own expert who testified based on ultrasounds and other evidence that Conner lived until after Christmas. The defense also sought to challenge other aspects of the prosecution's case. To rebut the dog-trailing evidence, the defense called Ronald Seitz, a second dog handler who also had his dog try to find Laci's scent at the Berkeley Marina on December 28. The dog, T.J., was given Laci's slipper as a scent object, but discovered no scent trail. To rebut the inference that Peterson had a financial incentive to kill Laci, the defense presented a financial expert who testified that TradeCorp U.S.A. and the Petersons were both reasonably financially healthy. To portray the prosecution's theory as physically impossible, the defense also sought to introduce video of a demonstration with a weighted 150-pound dummy in a boat on the bay in which a defense firm employee, trying to dump the dummy out, sank the boat. As will be discussed below, the trial court excluded the video.

The defense offered explanations for the circumstances of Peterson's behavior in April. His use of his mother's name to purchase a car was at her suggestion, to avoid having it impounded. He had large amounts of cash because she gave it to him to reimburse him for money erroneously withdrawn from his bank account rather than hers. Finally, he had his brother's driver's license because the club where he was going to golf that day gave discounts for local residents such as his brother.

3. Guilt Phase Verdict

The jury found Peterson guilty of murder in the first degree for killing Laci and murder in the second degree for killing Conner. (See Pen. Code, §§ 187, 189.) It found true the sole charged special circumstance, for multiple murder. (See id. , § 190.2, subd. (a)(3).)

B. Penalty Phase Trial

1. Prosecution Evidence

Peterson had no criminal record nor any history of violent acts. At the penalty phase, the prosecution relied exclusively on the circumstances of the crime and victim impact evidence. Four members of Laci's immediate family — her mother, Sharon; her stepfather, Ron Grantski; her brother, Brent; and her sister, Amy — testified. They described who Laci was as a person, shared photographs, memories, and vignettes from her life, and conveyed the grief and loss they each felt after the deaths of Laci and her unborn child.

2. Defense Evidence

Through friends, family, neighbors, teachers, coworkers, employers, and other witnesses, the defense offered evidence that Peterson had been a kind and positive member of the community. Peterson grew up in a loving family, displayed a patient and gentle disposition, and was a solid student. As part of his high school community service requirement, Peterson worked at a home for the elderly and tutored homeless children. He started his own business and worked a variety of other jobs while in college.

According to the defense, Peterson was always calm with Laci. Indeed, witnesses testified Peterson was calm at all times — at work, on the golf course, and in his dealings with all those around him.

Friends and family testified to the impact the trial had had on Peterson's relatives and indicated they believed, if sentenced to life in prison, Peterson could make a positive impact on the lives of others.

In closing argument, defense counsel described Peterson's life as one worth saving and argued that lingering doubt about Peterson's guilt should also weigh in favor of a life verdict.

3. Penalty Phase Verdict and Sentence

The jury returned a death verdict. The court denied a motion for new trial, denied the automatic motion for modification of the verdict, and imposed a sentence of death.

II. DISCUSSION

A. Excusal of Prospective Jurors for Cause Based on Questionnaire Answers Reflecting Opposition to the Death Penalty

Peterson claims errors occurred during every phase of his trial. We begin with his challenge to the manner in which the jury was chosen. During jury selection, multiple prospective jurors were excused based solely on written questionnaire responses indicating they were personally opposed to the death penalty. Peterson contends that, absent any indication these jurors would be unable to faithfully and impartially apply the law, it was error to remove them from the juror pool. On this initial point, Peterson is correct. Long-standing United States Supreme Court precedent makes clear that prospective jurors may not be disqualified from service in a capital case solely because of their general objections to the death penalty. (See Witherspoon v. Illinois, supra , 391 U.S. at pp. 518–523, 88 S.Ct. 1770 ( Witherspoon ); Wainwright v. Witt, supra , 469 U.S. at p. 424, 105 S.Ct. 844 ( Witt ).) That is just what happened here. And as this court has repeatedly explained, under high court precedent, even one such error requires "automatic reversal of any ensuing death penalty judgment." ( People v. Heard (2003) 31 Cal.4th 946, 966–967, 4 Cal.Rptr.3d 131, 75 P.3d 53 ; accord, e.g., People v. Armstrong (2019) 6 Cal.5th 735, 764, 243 Cal.Rptr.3d 105, 433 P.3d 987 ; People v. Woodruff (2018) 5 Cal.5th 697, 745, 235 Cal.Rptr.3d 513, 421 P.3d 588 ; People v. Riccardi, supra , 54 Cal.4th at p. 783, 144 Cal.Rptr.3d 84, 281 P.3d 1.) Here, there was not just one error; there were many. We are therefore required to reverse the death judgment. Contrary to Peterson's argument, however, we are not also required to reverse the judgment of guilt.

Jury selection in a capital case typically begins with prospective jurors filling out written questionnaires. These questionnaires allow the court and counsel to explore potential jurors' views and past experiences that might affect how they evaluate the evidence to be presented. They also address views a juror might have concerning the death penalty, in the event a defendant is found guilty and findings are made that would render him or her eligible to be punished by death. The questionnaire in this case, based on proposals from the parties and as approved by the court, contained roughly 120 questions on a range of topics, including 13 directed to potential jurors' views on the death penalty. To supplement the picture painted by answers to written questionnaires, the court and counsel may ask prospective jurors individual questions orally to clarify the nature of any views and further evaluate their ability to serve, a process known as voir dire. (See People v. Armstrong, supra , 6 Cal.5th at p. 749, 243 Cal.Rptr.3d 105, 433 P.3d 987 ; Hovey v. Superior Court (1980) 28 Cal.3d 1, 80–81, 168 Cal.Rptr. 128, 616 P.2d 1301.)

Prospective jurors may be excused from jury service through one of two primary mechanisms. First, the court may excuse jurors for cause based on a determination that bias or another substantial impairment disqualifies them from service. In a capital case, prospective jurors are subject to excusal if they would be unable or unwilling to impose the death penalty (or, conversely, if they would be unable or unwilling to vote against death) should the defendant be found guilty. (See, e.g., Ross v. Oklahoma (1988) 487 U.S. 81, 83–86, 108 S.Ct. 2273, 101 L.Ed.2d 80 ; Lockett v. Ohio (1978) 438 U.S. 586, 595–596, 98 S.Ct. 2954, 57 L.Ed.2d 973.) Second, prospective jurors may be excused by the parties by means of peremptory challenge. Each party is given a number of peremptory strikes set by law and allowed to cull from the remaining pool, up to the limit of their strikes, additional potential jurors they believe would be less favorably disposed to their side and to the verdict they seek. The result of this process is a final jury of 12, plus alternates to guard against the need to excuse one or more jurors during trial itself.

We are concerned here only with the first mechanism, excusal for cause. The question is whether the process the trial court employed to remove jurors complied with the standards the United States Supreme Court has established for the disqualification of jurors for bias in a capital case. More than half a century ago, the United States Supreme Court held in Witherspoon, supra , 391 U.S. at page 522, 88 S.Ct. 1770, that "a sentence of death cannot be carried out if the jury that imposed or recommended it was chosen by excluding veniremen for cause simply because they voiced general objections to the death penalty or expressed conscientious or religious scruples against its infliction." The court explained the reason for this rule: "A man who opposes the death penalty, no less than one who favors it, can make the discretionary judgment entrusted to him by the State and can thus obey the oath he takes as a juror. But a jury from which all such men have been excluded cannot perform the task demanded of it," namely, to "express the conscience of the community on the ultimate question of life or death." ( Id. at p. 519, 88 S.Ct. 1770.) Put differently, "a criminal defendant has the right to an impartial jury drawn from a venire that has not been tilted in favor of capital punishment by selective prosecutorial challenges for cause" based solely on general opposition to the death penalty. ( Uttecht v. Brown (2007) 551 U.S. 1, 9, 127 S.Ct. 2218, 167 L.Ed.2d 1014.)

The law also recognizes, however, that states must have a way to ensure capital cases are tried before juries "able to apply capital punishment within the framework state law prescribes." ( Uttecht v. Brown, supra , 551 U.S. at p. 9, 127 S.Ct. 2218.) In Witt, supra , 469 U.S. 412, 105 S.Ct. 844, the court held that trial courts may excuse a prospective juror for cause based on the juror's views of capital punishment if those views "would ‘prevent or substantially impair the performance of his duties as a juror in accordance with his instructions and his oath.’ " ( Id. at p. 424, 105 S.Ct. 844, quoting Adams v. Texas (1980) 448 U.S. 38, 45, 100 S.Ct. 2521, 65 L.Ed.2d 581 ; accord, People v. Jones (2017) 3 Cal.5th 583, 614, 220 Cal.Rptr.3d 618, 398 P.3d 529.) But to protect the right to trial by impartial jury, a trial court may not remove jurors for cause based on views that do not substantially impair their ability to serve. "[I]f prospective jurors are barred from jury service because of their views about capital punishment on ‘any broader basis’ than inability to follow the law or abide by their oaths, the death sentence cannot be carried out." ( Adams , at p. 48, 100 S.Ct. 2521 ; accord, Uttecht , at p. 9, 127 S.Ct. 2218 ; Jones , at p. 614, 220 Cal.Rptr.3d 618, 398 P.3d 529.)

Taken together, Witherspoon and Witt make clear that prospective jurors may not be disqualified from service simply because they object to the death penalty as a general matter. "[N]ot all who oppose the death penalty are subject to removal for cause in capital cases; those who firmly believe that the death penalty is unjust may nevertheless serve as jurors in capital cases so long as they state clearly that they are willing to temporarily set aside their own beliefs in deference to the rule of law." ( Lockhart v. McCree (1986) 476 U.S. 162, 176, 106 S.Ct. 1758, 90 L.Ed.2d 137.) Nor may a juror be disqualified from service because he or she might "impose a higher threshold before concluding that the death penalty is appropriate." ( People v. Stewart (2004) 33 Cal.4th 425, 447, 15 Cal.Rptr.3d 656, 93 P.3d 271.) "The critical issue is whether a life-leaning prospective juror — that is, one generally (but not invariably) favoring life in prison instead of the death penalty as an appropriate punishment — can set aside his or her personal views about capital punishment and follow the law as the trial judge instructs." ( People v. Thompson (2016) 1 Cal.5th 1043, 1065, 210 Cal.Rptr.3d 667, 384 P.3d 693.) If a juror can obey those instructions and determine whether death is appropriate based on a sincere consideration of aggravating and mitigating circumstances, the juror may not be excused for cause. ( People v. Armstrong, supra , 6 Cal.5th at p. 750, 243 Cal.Rptr.3d 105, 433 P.3d 987 ; Stewart , at p. 447, 15 Cal.Rptr.3d 656, 93 P.3d 271 ; People v. Lewis (2001) 25 Cal.4th 610, 633, 106 Cal.Rptr.2d 629, 22 P.3d 392.)

Here, Peterson directs our attention to a group of prospective jurors who were excused based solely on their questionnaire answers, without further questioning by the court or counsel. When dismissal is based solely on the written record, we independently review whether that record establishes that a juror was categorically unable to serve. ( People v. Woodruff, supra , 5 Cal.5th at p. 743, 235 Cal.Rptr.3d 513, 421 P.3d 588 ; People v. Riccardi, supra , 54 Cal.4th at p. 779, 144 Cal.Rptr.3d 84, 281 P.3d 1.) " ‘[W]hen an excusal was based on questionnaire responses alone, the excusal may be upheld if those answers, "taken together," clearly demonstrate the juror's unwillingness or inability, because of attitudes about the death penalty, to perform his or her duties in a capital trial.’ " ( Riccardi , at p. 779, 144 Cal.Rptr.3d 84, 281 P.3d 1.)

The record reveals that many jurors were summarily excused based on their responses to a single question, No. 109: "How would you rate your attitude towards the death penalty?" That question gave prospective jurors six possible answers — Strongly Oppose, Oppose, Weakly Oppose, Weakly Support, Support, and Strongly Support — and spaces to mark which one of these most closely reflected their general attitude. Although the prospective jurors' answers to this question could well have prompted further inquiry during voir dire, these answers alone offered little insight into the controlling issue for purposes of their qualification to serve as jurors — whether they, whatever their general views on the death penalty might be, could accept and follow the court's instructions and be able to choose either life or death based on a sincere consideration of any aggravating or mitigating circumstances. ( People v. Armstrong, supra , 6 Cal.5th at p. 750, 243 Cal.Rptr.3d 105, 433 P.3d 987.) On that issue, a second question, No. 115, was more illuminating. It asked: "Do you have any moral, religious, or philosophical opposition to the death penalty so strong that you would be unable to impose the death penalty regardless of the facts?"¹⁰

10 A related question, No. 116, screened for jurors on the other end of the spectrum. It asked: "Do you have any moral, religious, or philosophical beliefs in favor of the death penalty so strong that you would be unable to impose life without possibility of parole regardless of the facts?" (Italics added.)

Peterson asks us to focus in particular on 13 prospective jurors who were excused without further questioning after expressing some degree of opposition to the death penalty in response to question No. 109, even though these same jurors also answered "no" to question No. 115 — meaning that, so far as their questionnaires revealed, none held views so strong that they would be unable to vote for death if the circumstances warranted. The trial court dismissed these 13 prospective jurors as not qualified to serve without, insofar as the record reveals, ever probing their views on the ultimate issue, whether they were substantially impaired.

Take, for example, Prospective Juror No. 4841. Asked in her questionnaire her feelings toward the death penalty, the juror wrote: "Have no feeling." Two questions later, asked to rate her attitude toward the death penalty and given the aforementioned range of options from "Strongly Oppose" to "Strongly Support," she checked "Strongly Oppose." But in response to the additional question, "Do you have any moral, religious, or philosophical opposition to the death penalty so strong that you would be unable to impose the death penalty regardless of the facts ?" (italics added), the prospective juror checked "No." Indeed, the juror implicitly confirmed that in some instances she could impose the death penalty; asked whether it would "be difficult for you to vote for the death penalty if the crime was the guilty party's first offense?," she checked "No."

Discussing the juror and whether counsel would stipulate to her excusal, the court said: "Reading these [questionnaire answers], I am of the opinion this juror wouldn't qualify. [¶] Look at 4841. Go to that one. 19 — page 19, answer 109. Strongly opposes the death penalty. I could dispose of this juror in a couple of questions. So if you don't want to stipulate, fine. But if they oppose the death penalty, they are not qualified under Wainwright [v. ] Witt . " That afternoon, the trial court read off the juror's answer to the attitude question, No. 109, again noting that she had checked, "Strongly Opposed." On that basis, he concluded "this juror ... would not qualify" and excused her without any questioning.

In response to these same questions, Prospective Juror No. 6960 said of the death penalty, "I wish it was not a thing needed" and checked that she opposed (not that she strongly opposed) the death penalty. But she too indicated on the very next page that her opposition to the death penalty was not such as would render her "unable to impose the death penalty regardless of the facts." Based on these answers, the court described the juror as "one who is opposed to the death penalty without qualification" and concluded it would "excuse 6960 for cause, because that juror is opposed to the death penalty, without reservation."

Prospective Juror No. 16727 indicated he was strongly opposed to the death penalty based on his "spirituality," but his opposition was not such that he could never impose it no matter the facts. His questionnaire indicated that, if the crime was a defendant's first offense, whether it would be difficult to vote for death would "[d]epend[ ] on the [e]vidence" — implying the juror could, in some instances, vote for death even for a first-time offender. The trial court, noting that the juror wrote he was "against the death penalty" and checked that he "strongly oppose[d]" it, said, "I don't think this person would qualify because of his answers, and he's opposed to the death penalty. So I would be inclined to excuse him. Over the objection of [defense counsel]." After defense counsel confirmed his "vehement objection," the court excused the juror without questioning him: "[J]uror number 16727 is excused because he is opposed to the death penalty."

Again, these were not isolated occurrences. The trial court excused more than a dozen prospective jurors based solely on their written opposition to the death penalty (question No. 109), without also considering their answers to question No. 115, which reflected the jurors' ability to impose the death penalty in some circumstances. Perhaps further questioning might have established that one or more of these jurors in truth could not conscientiously consider death as an option. But no such questioning occurred here, because the trial court rejected counsel's requests to question these prospective jurors and declined to ask any questions itself. As a result, these jurors were excused for cause based on a written expression of opposition to the death penalty, without more. We do not suggest that a trial court errs any time it exercises its discretion to limit counsel's opportunity to question prospective jurors directly. In a case with a venire of this size — nearly 1,500 prospective jurors — tight controls on voir dire were necessary and inevitable. But a court must still ensure that, by whatever means, sufficient inquiry is made so only those properly excusable under the governing standards are dismissed for cause: " ‘Before granting a challenge for cause, the "court must have sufficient information regarding the prospective juror's state of mind to permit a reliable determination as to whether the juror's views would ‘ "prevent or substantially impair" ’ " performance as a capital juror. [Citation.] Trial courts must therefore make "a conscientious attempt to determine a prospective juror's views regarding capital punishment to ensure that any juror excused from jury service meets the constitutional standard." ’ " ( People v. Buenrostro (2018) 6 Cal.5th 367, 412, 240 Cal.Rptr.3d 704, 430 P.3d 1179.) Sufficient inquiries were not made in this case.

The People concede Peterson's claims are all preserved because at the time of trial no objection was required to preserve claims of Witt / Witherspoon error. (See People v. Jones (2013) 57 Cal.4th 899, 914–915, 161 Cal.Rptr.3d 295, 306 P.3d 1136 [describing evolution of the forfeiture rule for such error].) In any event, the record shows defense counsel consistently resisted these dismissals, arguing on numerous occasions that just because a juror indicated opposition to the death penalty, that did not mean he or she could not vote for death in appropriate circumstances. These objections gained no traction with the trial court, and ultimately the defense had no choice but to accede to the for-cause standard the court had adopted: "Obviously[,] I believe that opposition to the death penalty should not be a for-cause challenge. The Court has ruled on it. I'm not going to continue to raise it each time. Although I want the record to reflect that I am submitting [subject to a standing objection] on the Court's previous rulings."

Crucially, "it is the adversary seeking exclusion who must demonstrate, through questioning, that the potential juror lacks impartiality." ( Witt, supra , 469 U.S. at p. 423, 105 S.Ct. 844 ; accord, People v. Stewart, supra , 33 Cal.4th at p. 445, 15 Cal.Rptr.3d 656, 93 P.3d 271.) It is thus incumbent on the party seeking excusal, or the court, to ask questions sufficient to differentiate between mere opposition and an actual inability to impose the death penalty. The exclusion of prospective jurors as impaired, in the absence of a record demonstrating they were impaired, is a violation of the Sixth Amendment right to an impartial jury. (See People v. Woodruff, supra , 5 Cal.5th at pp. 744–745, 235 Cal.Rptr.3d 513, 421 P.3d 588 [error not to ask clarifying questions to determine whether written indication of opposition to the death penalty would impair juror's ability to serve]; People v. Riccardi, supra , 54 Cal.4th at p. 782, 144 Cal.Rptr.3d 84, 281 P.3d 1 [same]; Stewart , at pp. 449–452, 15 Cal.Rptr.3d 656, 93 P.3d 271 [same].)

Even though neither the court nor the prosecution questioned these 13 excused jurors, the People argue that all 13 dismissals are supported by substantial evidence in the record. Given the absence of questioning, the People rely solely on the jurors' questionnaire responses. In addition to the opposition to the death penalty reflected in the jurors' answers to question No. 109, the People point to other responses to questions about the jurors' death penalty attitudes — for example, that in many cases their views had not changed in the last 10 years, or that some jurors anticipated it would be difficult for them to choose death for a first-time offender. There are two difficulties with the People's response. The first is that it misstates the standard of review; as discussed, when dismissal for cause based on an inability to impose the death penalty rests solely on written questionnaire answers, we review the record de novo, rather than under the deferential substantial evidence standard. ( People v. Woodruff, supra , 5 Cal.5th at p. 743, 235 Cal.Rptr.3d 513, 421 P.3d 588.) The second, more fundamental difficulty is that nothing in the prospective jurors' questionnaire responses establishes that the jurors would have been unable to follow the law and impose the death penalty if circumstances warranted, which is the only relevant question under high court precedent. There is no dispute that each of the 13 prospective jurors opposed the death penalty, sometimes strongly so. But the issue is whether that opposition would have disabled them from following the court's instructions and ever imposing the death penalty. On that point, in response to question No. 115, all 13 prospective jurors expressly indicated their death penalty views were not so strong that they would be unable to impose the death penalty. That answer required further inquiry before the court could conclude these jurors were impaired. (See People v. Leon (2015) 61 Cal.4th 569, 592–593, 189 Cal.Rptr.3d 703, 352 P.3d 289.)

In sum: The law is clear that a capital jury may include those who, as an abstract matter, oppose — or even strongly oppose — the death penalty, though a prosecutor might seek to limit the number of such jurors. It may include those who favor — or even strongly favor — the death penalty, though defense counsel might seek to limit their numbers. Eligibility for service does not depend on a juror's abstract views of capital punishment. It depends, instead, on the prospective juror's willingness and ability to follow a court's instructions and conscientiously consider both penalties in light of the evidence presented by each side. This is the meaning of the guarantee of an impartial jury, drawn from the community at large, for the trial of a defendant facing the death penalty.

Under that standard, the questionnaire answers submitted by these prospective jurors did not establish they were unfit to serve. Voir dire might have painted a different picture, with the court and counsel through oral questions exploring whether each individual juror had the necessary ability and willingness to consider both life and death as options. But for these 13 jurors, there was no such questioning. Thus, we know only that in the abstract they opposed the death penalty. The record made in the trial court does not offer a basis sufficient to uphold excusal of these jurors for cause under the clear standards laid out by the United States Supreme Court.

The People contend these errors in jury selection should be treated as harmless. But precedent requires otherwise. Witherspoon itself held "that a sentence of death cannot be carried out" if it has been imposed by a jury chosen after disqualifying prospective jurors for having "voiced general objections to the death penalty." ( Witherspoon, supra , 391 U.S. at p. 522, 88 S.Ct. 1770.) The United States Supreme Court has since instructed that even a single erroneous exclusion of a prospective juror based solely on his or her general views about the death penalty requires that any death sentence thereafter imposed be set aside. ( Gray v. Mississippi, supra , 481 U.S. at p. 665, 107 S.Ct. 2045 (plur. opn. of Blackmun, J.) ["The nature of the jury selection process defies any attempt to establish that an erroneous Witherspoon - Witt exclusion of a juror is harmless"]; see id. at p. 672, 107 S.Ct. 2045 (conc. opn. of Powell, J.); see People v. Riccardi, supra , 54 Cal.4th at p. 778, 144 Cal.Rptr.3d 84, 281 P.3d 1 [ Witherspoon - Witt error "requires automatic reversal of defendant's sentence of death under existing United States Supreme Court precedent"].) Here, the trial court excluded no fewer than 13 jurors based on their general views about the death penalty, even though all 13 attested that their views would not prevent them from following the law. The death sentence must be reversed, and the People given another opportunity to seek that penalty before a properly selected jury if they so choose. In their brief, the People asked us to reconsider our cases, including principally People v. Riccardi, supra , 54 Cal.4th 758, 144 Cal.Rptr.3d 84, 281 P.3d 1, that hold the erroneous exclusion of a juror based on death penalty views can never be harmless. But at oral argument, the People acknowledged that Gray v. Mississippi, supra , 481 U.S. 648, 107 S.Ct. 2045 is controlling and that error of this sort requires automatic reversal of the penalty judgment. (See People v. Armstrong, supra , 6 Cal.5th at p. 764, 243 Cal.Rptr.3d 105, 433 P.3d 987.) Riccardi is but one in a long line of cases, dating back to Witherspoon itself, that have held that reversal is required.

In applying these rules here, we break no new ground; the governing law was as clear at the time of trial as it is today. Indeed, in a case decided several months before jury selection began in this case, we pointedly reminded trial judges about the critical importance of carefully adhering to the well-settled standards and procedures for death-qualifying a jury: "In view of the extremely serious consequence — an automatic reversal of any ensuing death penalty judgment — that results from a trial court's error in improperly excluding a prospective juror for cause during the death-qualification stage of jury selection, we expect a trial court to make a special effort to be apprised of and to follow the well-established principles and protocols pertaining to the death qualification of a capital jury. As the present case demonstrates, an inadequate or incomplete examination of potential jurors can have disastrous consequences as to the validity of a judgment. The error that occurred in this case — introducing a fatal flaw that tainted the outcome of the penalty phase even before the jury was sworn — underscores the need for trial courts to proceed with special care and clarity in conducting voir dire in death penalty trials. The circumstance that the error in this case was committed by a trial judge with substantial experience in criminal law renders the voir dire examination at issue all the more inexplicable and disappointing." ( People v. Heard, supra , 31 Cal.4th at pp. 966–967, 4 Cal.Rptr.3d 131, 75 P.3d 53.) Those remarks apply with full force to this case. Because the trial court failed to develop a record sufficient to support excusal of jurors for cause, and because the prosecution did not speak up as these errors were occurring to ensure an adequate record, the penalty phase in this case was over before it ever began. It is in no one's interest for a capital case to begin with the certainty that any ensuing death verdict will have to be reversed and the entire penalty case retried. (See People v. Heard, supra , 31 Cal.4th at p. 968, 4 Cal.Rptr.3d 131, 75 P.3d 53.) We emphasize what we said in Heard , and have said many times before and since: Before a prospective capital juror is dismissed on the basis of death penalty views, it is imperative that the party seeking dismissal and the trial court ensure a record adequately establishing those views is developed. (See, e.g., People v. Buenrostro, supra , 6 Cal.5th at p. 412, 240 Cal.Rptr.3d 704, 430 P.3d 1179 ; People v. Covarrubias (2016) 1 Cal.5th 838, 863–866, 207 Cal.Rptr.3d 228, 378 P.3d 615 ; People v. Zaragoza (2016) 1 Cal.5th 21, 37–41, 204 Cal.Rptr.3d 131, 374 P.3d 344 ; People v. Leon, supra , 61 Cal.4th at p. 592, 189 Cal.Rptr.3d 703, 352 P.3d 289.) Jurors may not be excused merely for opposition to the death penalty, but only for views rendering them unable to fairly consider imposing that penalty in accordance with their oath. The record must make manifest that inability — not with " ‘unmistakable clarity’ " ( Witt, supra , 469 U.S. at p. 424, 105 S.Ct. 844 ), to be sure, but with sufficient clarity that a reviewing court can identify a basis for the trial court's conclusion that the juror actually lacked the requisite ability. In the absence of such a record, we have no choice but to reverse the death sentence. ( Buenrostro , at pp. 415–418, 240 Cal.Rptr.3d 704, 430 P.3d 1179.)

Peterson, however, asks us to go one step further. He argues that the errors in jury selection affected all parts of his trial, not just the penalty phase, and rendered the results of the jury's guilt phase deliberations unreliable as well. He accordingly asks that we set aside not only his sentence but the murder convictions that preceded it. Both the United States Supreme Court and this court have previously declined to take this additional step, and Peterson offers no persuasive ground for doing so here.

The United States Supreme Court in Witherspoon rejected the argument that errors in death qualifying a jury necessarily undermine its guilt phase verdict in addition to its penalty judgment. Invoking surveys and academic studies, Witherspoon had argued that "the kind of juror who would be unperturbed by the prospect of sending a man to his death ... is the kind of juror who would too readily ignore the presumption of the defendant's innocence, accept the prosecution's version of the facts, and return a verdict of guilt," and thus his jury was biased as to guilt too. ( Witherspoon, supra , 391 U.S. at pp. 516–517, 88 S.Ct. 1770 ; see id. at p. 517, fns. 10, 11, 88 S.Ct. 1770.) The high court was unpersuaded. While the court considered it "self-evident" that errors in death qualification would undermine the jury's impartiality "in its role as arbiter of the punishment to be imposed" ( id. at p. 518, 88 S.Ct. 1770 ), Witherspoon's studies failed to show that the same was true of the jury in its different capacity as finder of fact ( id. at pp. 517–518, 88 S.Ct. 1770 ; see id. at p. 521, fn. 20, 88 S.Ct. 1770 [determination whether to impose the death penalty "is different in kind from a finding that the defendant committed a specified criminal offense"] ). The court explained: "We simply cannot conclude, either on the basis of the record now before us or as a matter of judicial notice, that the exclusion of jurors opposed to capital punishment results in an unrepresentative jury on the issue of guilt or substantially increases the risk of conviction." ( Id. at pp. 517–518, 88 S.Ct. 1770.) Despite a record in which "the prosecution eliminated nearly half the venire of prospective jurors by challenging ... any venireman who expressed qualms about capital punishment" ( id. at p. 513, 88 S.Ct. 1770 ), the high court refused to disturb the guilty verdict ( id. at pp. 518, 522–523, fn. 21, 88 S.Ct. 1770 ). Our own cases have followed the same path, reversing the death judgment in cases of jury-selection error but declining to disturb the guilty verdict. (See, e.g., People v. Armstrong, supra , 6 Cal.5th at p. 764 ; People v. Stewart, supra , 33 Cal.4th at p. 455, 15 Cal.Rptr.3d 656, 93 P.3d 271.)

Peterson candidly acknowledges this precedent, but asks us to distinguish it on the ground that Witherspoon and subsequent cases considered only whether the Sixth and Fourteenth Amendments to the United States Constitution require a guilty verdict be set aside; Peterson argues the Eighth Amendment compels a different result. But Peterson's Eighth Amendment claim "appears to be merely a restatement of [the] Sixth Amendment claim[ ]." ( People v. Johnson (1992) 3 Cal.4th 1183, 1213, 14 Cal.Rptr.2d 702, 842 P.2d 1.) Though couched in terms of Eighth Amendment reliability, as opposed to Sixth Amendment representativeness, it depends on the identical premise — that the exclusion of jurors generally opposed to the death penalty tilts the venire to an unacceptable degree in favor of the prosecution on questions of guilt as well as sentence. On the record before it, the court in Witherspoon was unable to say that even systematic exclusion of those opposed to the death penalty "substantially increase[d] the risk of conviction." ( Witherspoon, supra , 391 U.S. at p. 518, 88 S.Ct. 1770.) The 1984, 1986, and 1994 studies Peterson cites do not lead to a different conclusion. He relies on research analyzing how the constitutionally accepted process of "death qualification" — i.e., the dismissal of potential jurors who would either always or never vote for death — alters the willingness to convict. These studies do not establish that excluding one — or even 13 — prospective jurors, from a pool of nearly 1,500, "substantially increase[s] the risk of error in the factfinding process." ( Beck v. Alabama (1980) 447 U.S. 625, 632, 100 S.Ct. 2382, 65 L.Ed.2d 392.) Nor does Peterson point to any evidence in the record to support this claim.

Peterson argues that even if the errors in jury selection do not require reversal of his convictions, there were several other errors that do command that result. We therefore turn to whether any of these asserted errors in the guilt phase trial requires that Peterson's murder convictions be set aside. B. Denial of Motion to Change Venue

Peterson was arrested and charged in Stanislaus County, where he and Laci had lived. He moved for a change of venue, arguing that a fair and impartial trial could not be had in Stanislaus because of the extensive publicity the case had received. The Stanislaus County Superior Court granted the motion and, after considering the alternatives recommended by the Administrative Office of the Courts,¹¹ including Alameda, Santa Clara, and Orange Counties, selected San Mateo County as the best venue. It found that local media coverage and the degree of community involvement in the case would preclude a fair trial in Stanislaus, but any county of sufficient size outside the Central Valley could provide a fair trial. San Mateo was selected based on its facilities and its relative proximity compared to Southern California, which would minimize travel for the many Modesto-area witnesses.

11 The Administrative Office of the Courts was the name of a body serving the Judicial Council of California. In 2014, the name was retired to better reflect that the office was, and is, a subpart of the Judicial Council.

Jury selection began in March 2004. In May 2004, Peterson filed a second change of venue motion, this time seeking transfer of the case to Los Angeles County. He argued that examination of questionnaire answers from prospective jurors showed, once again, that extensive pretrial publicity was affecting the juror pool. The trial court considered the parties' papers and arguments and made a detailed oral ruling denying a further venue change. Peterson contends the denial of his second motion violated his federal constitutional right to trial by an impartial jury ( U.S. Const., 6th & 14 Amends.) and similar state guarantees.

In a series of cases in the 1960s, the United States Supreme Court recognized that media publicity about a criminal trial could in some circumstances deprive the defendant of the right to trial by an impartial jury. ( Sheppard v. Maxwell (1966) 384 U.S. 333, 86 S.Ct. 1507, 16 L.Ed.2d 600 ; Estes v. Texas (1965) 381 U.S. 532, 85 S.Ct. 1628, 14 L.Ed.2d 543 ; Rideau v. Louisiana (1963) 373 U.S. 723, 83 S.Ct. 1417, 10 L.Ed.2d 663 ; Irvin v. Dowd (1961) 366 U.S. 717, 81 S.Ct. 1639, 6 L.Ed.2d 751 ; see generally Skilling v. United States (2010) 561 U.S. 358, 378–381, 130 S.Ct. 2896, 177 L.Ed.2d 619.) In the wake of these decisions, to ensure "the requirement basic to our jurisprudence that every person accused of crime is entitled to a trial by a fair and impartial jury" ( Maine v. Superior Court (1968) 68 Cal.2d 375, 384, 66 Cal.Rptr. 724, 438 P.2d 372 ), this court adopted a new standard for pretrial change of venue motions: Such a motion should " ‘be granted whenever it is determined that because of the dissemination of potentially prejudicial material, there is a reasonable likelihood that in the absence of such relief, a fair trial cannot be had’ " ( id. at p. 383, 66 Cal.Rptr. 724, 438 P.2d 372 ; see Pen. Code, § 1033, added by Stats. 1971, ch. 1476, § 3, p. 2915 [codifying the Maine standard] ).

Over time, we have elaborated on the prophylactic Maine standard and identified a series of considerations courts must weigh to ensure the constitutional right to a fair trial is preserved. "The factors to be considered are the nature and gravity of the offense, the nature and extent of the news coverage, the size of the community, the status of the defendant in the community, and the popularity and prominence of the victim." ( People v. Harris (1981) 28 Cal.3d 935, 948, 171 Cal.Rptr. 679, 623 P.2d 240 ; accord, e.g., People v. Rices (2017) 4 Cal.5th 49, 72, 226 Cal.Rptr.3d 118, 406 P.3d 788.) "On appeal, the defense bears the burden of showing both error and prejudice. It must establish a reasonable likelihood both that a fair trial could not be had at the time of the motion, and that the defendant did not actually receive a fair trial." ( People v. Smith (2015) 61 Cal.4th 18, 39, 186 Cal.Rptr.3d 550, 347 P.3d 530 ; see People v. McCurdy (2014) 59 Cal.4th 1063, 1075, 176 Cal.Rptr.3d 103, 331 P.3d 265.) Alternatively, in rare and "exceptional cases," a defendant may show circumstances so " ‘extraordinary’ " that a court may assume no fair trial could be had. ( People v. Prince (2007) 40 Cal.4th 1179, 1216, 57 Cal.Rptr.3d 543, 156 P.3d 1015.) We accept when supported by substantial evidence the facts a trial court finds in connection with a motion to change venue. ( Smith , at p. 39, 186 Cal.Rptr.3d 550, 347 P.3d 530 ; McCurdy , at p. 1075, 176 Cal.Rptr.3d 103, 331 P.3d 265.) As dictated by Sheppard v. Maxwell, supra , 384 U.S. at page 362, 86 S.Ct. 1507 and Maine v. Superior Court, supra , 68 Cal.2d at page 382, 66 Cal.Rptr. 724, 438 P.2d 372, however, we will independently review all the circumstances to determine whether there was a reasonable likelihood of an unfair trial. ( Smith , at p. 39, 186 Cal.Rptr.3d 550, 347 P.3d 530 ; McCurdy , at p. 1075, 176 Cal.Rptr.3d 103, 331 P.3d 265.)

Conceding that the other factors were "largely neutral," Peterson rests his argument that denial of a second change in venue was error on a single consideration, the nature and extent of pretrial publicity. This case was the subject of massive, worldwide media attention. Peterson asserts, and the People do not dispute, "that the combination of print media, radio coverage, television and cable and internet coverage made this perhaps the most widely covered trial in American history." The trial judge remarked, "The only place you could send this case probably where they wouldn't [have] hear[d] about it — I'm not so sure about that — would be send it to Mars, you know. That's the only place where you could try this case where nobody would know anything about it. It's been all over the world." We also take as a given that much of this publicity portrayed Peterson in a negative light. We nonetheless find no error in the trial court's denial of Peterson's second change of venue motion.

Preliminarily, we reject Peterson's argument that when unfavorable publicity reaches the saturation level of this case, any denial of a change of venue motion is presumptively prejudicial and a defendant need not show an impartial jury could not be seated. To the contrary, " ‘it is well-settled that pretrial publicity itself — "even pervasive, adverse publicity — does not inevitably lead to an unfair trial." ’ " ( People v. Prince, supra , 40 Cal.4th at p. 1216, 57 Cal.Rptr.3d 543, 156 P.3d 1015 ; see Murphy v. Florida (1975) 421 U.S. 794, 799, 95 S.Ct. 2031, 44 L.Ed.2d 589 ; People v. Harris, supra , 28 Cal.3d at pp. 949–950, 171 Cal.Rptr. 679, 623 P.2d 240.) Rather, a "presumption of prejudice" from excessive publicity "attends only the extreme case" ( Skilling v. United States, supra , 561 U.S. at p. 381, 130 S.Ct. 2896 ), one in which the " ‘trial atmosphere [is] utterly corrupted by press coverage’ " ( id. at p. 380, 130 S.Ct. 2896 ). Peterson has not shown that the atmosphere outside the courtroom reached inside the courtroom and categorically precluded a fair trial.

In similar cases, courts have held that a change of venue motion may be denied if the change would be futile: "Where pretrial publicity has been geographically widespread and pervasive ..., a court may deny change of venue on the sensible ground that it would do no good." ( People v. Venegas (1994) 25 Cal.App.4th 1731, 1738, 31 Cal.Rptr.2d 114.) For cases of a certain profile, it would be "speculation to suppose the results of jury selection would have been significantly different in any county. The media report local trials of notorious crimes in all counties. People read newspapers and watch television" — and, we may now add, use the Internet — "in all counties." ( People v. Cooper (1991) 53 Cal.3d 771, 807, 281 Cal.Rptr. 90, 809 P.2d 865.) Discussing the trial of members of the Charles Manson family, the Court of Appeal wrote: "It is patently clear that the crimes charged, as well as the identity and the involvement of appellants, permeated every corner of this state with varying degrees of intensity. The ubiquity of media coverage made any such differential one of insignificant degree. A change of venue offered no solution to the publicity problem. Even if venue had been changed, nothing could have prevented the public media from swinging its attention to that place. The magnetic pull of such notorious cases is compelling." ( People v. Manson (1976) 61 Cal.App.3d 102, 176–177, 132 Cal.Rptr. 265 ; see People v. Davis (2009) 46 Cal.4th 539, 578–579, 94 Cal.Rptr.3d 322, 208 P.3d 78.)

The same is true here. Precisely because this case was the subject of such widespread media attention, it is unclear what purpose a second change of venue would have served. The publicity the Peterson trial generated, like the trials of O.J. Simpson, the Manson family, and any number of other so-called trials of the century before them, was intrinsic to the case , not the place. This was even more true than in earlier times because, as Peterson rightly notes, his trial followed the explosion of cable television and the Internet as sources of information, facilitating nationwide coverage of the case. For these reasons, as the trial court aptly observed, "[i]t is speculation to suppose [the] results of jury selection would be any different anywhere else." Media attention followed this case from Stanislaus County to San Mateo County. Given the level of public interest, this case would have attracted attention in any venue in which it was held. There is no rational reason to think coverage would have been any less in Los Angeles County — one of the media capitals of the world — if Peterson's motion had been granted.

Against this conclusion, Peterson relies on survey statistics measuring the levels in Los Angeles County of awareness and prejudgment of the case, but those statistics do not support his argument. First, the surveys did not include San Mateo, so no direct comparisons can be drawn between Los Angeles and San Mateo. Second, the two other Bay Area counties surveyed, Alameda and Santa Clara, generally showed no statistically significant differences from Los Angeles, although all three counties were statistically significantly different from Stanislaus, where the case was originally pending. This led Peterson's own expert to opine, in support of moving the case from Stanislaus, that for a fair trial the "best three counties that are [statistically] fairly close are Los Angeles, Alameda, and Santa Clara County." Third, the statistics derive from a survey taken a full year before the second change of venue motion and do not show how prejudgment levels would have changed if the case were transferred to Los Angeles, and with it the national media spotlight. (See People v. Davis, supra , 46 Cal.4th at p. 575, 94 Cal.Rptr.3d 322, 208 P.3d 78 ["Because it is impossible to control heightened media attention in any new venue, it also is virtually impossible to prevent the knowledge and prejudgment rates for potential jurors living in a new venue from increasing after the change of venue has occurred"].)

Instead, in a high-profile case such as this one, provided a sufficiently large pool is available — and we agree with the trial court that San Mateo's 700,000-plus residents provided such a pool — the better answer is not to change venue yet again but to rigorously vet potential jurors to screen out those tainted and irrevocably biased by pretrial publicity, to find 12, plus alternates, who can decide only on the evidence admitted at trial. Almost inevitably even those qualified for potential service by a court may have had some prior exposure to the case, but "[p]rominence does not necessarily produce prejudice, and juror impartiality , we have reiterated, does not require ignorance. " ( Skilling v. United States, supra , 561 U.S. at p. 381, 130 S.Ct. 2896.) What matters is whether each prospective juror can " ‘lay aside his [or her] impression or opinion and render a verdict based on the evidence presented in court.’ " ( People v. Harris, supra , 28 Cal.3d at p. 950, 171 Cal.Rptr. 679, 623 P.2d 240, quoting Irvin v. Dowd, supra , 366 U.S. at p. 723, 81 S.Ct. 1639.)

The trial court here did just this. Indeed, defense counsel himself acknowledged, "I think the court has exercised Herculean efforts in trying to get a fair panel here." Nearly 1,500 prospective jurors were scrutinized. (See People v. Famalaro (2011) 52 Cal.4th 1, 30, 127 Cal.Rptr.3d 40, 253 P.3d 1185 ["The huge number of prospective jurors initially summoned (1,200) ensured that an ample number of unbiased prospective jurors remained after the biased ones had been excused"].) The court observed that "[t]he 66 people we've qualified so far have given us [the] assurance" that they could set aside any prior impressions or opinions and decide the case only on the evidence, "and the court is satisfied that that is, in fact, the case." Ultimately, all 18 jurors and alternates chosen to serve had indicated in their juror questionnaires that they had not formed any opinion as to Peterson's guilt or innocence because they did not have enough information to decide. The actual seated jurors thus had no preliminary opinions on guilt — never mind preset views they would be unable to set aside.

Peterson argues that the seated jurors' declarations of impartiality should be rejected based on his survey data drawn from the venire as a whole. We agree that "the juror's assurances that he [or she] is equal to this task cannot be dispositive of the accused's rights." ( Murphy v. Florida, supra , 421 U.S. at p. 800, 95 S.Ct. 2031.) But Peterson offers no sound basis to believe the jurors' assurances in this case were insincere. Almost every prospective juror had been exposed to publicity about the case, but this is unsurprising for a case Peterson describes as the most publicized in American history, and would have been unlikely to change much in any other county. According to Peterson's review of juror questionnaires filled out before the second motion, 43 percent of prospective jurors (426 of 998) had formed a preliminary opinion that Peterson was guilty, and 19 percent (190 of 998), roughly one in five, would not be able to set aside that view. But this still means that, by the time jury selection was complete, the parties would have had well in excess of 1,000 avowedly impartial jurors to choose from. "When, as here, there is a ‘large, diverse pool of potential jurors, the suggestion that 12 impartial individuals could not be empanelled is hard to sustain.’ " ( People v. Famalaro, supra , 52 Cal.4th at p. 23, 127 Cal.Rptr.3d 40, 253 P.3d 1185, quoting Skilling v. United States, supra , 561 U.S. at p. 382, 130 S.Ct. 2896.) Our confidence in the fairness of the actual jury rests as well on our independent review of their voir dire, during which the parties and court carefully vetted them to ferret out bias. (See Famalaro , at p. 31, 127 Cal.Rptr.3d 40, 253 P.3d 1185.) Peterson has not demonstrated the trial court could not, and did not, find 12 impartial individuals. His claim that he was denied his right to trial before a fair jury is without merit.¹² We acknowledge Peterson's concern was not just that the court would be unable to find 12 unbiased jurors. Rather, it was that they would not stay unbiased; according to defense counsel, "the problem is that you take that panel that may be fair here, and you stick them back in the community and you do that on a daily basis over five months, and you're not going to be able to get over the kind of community passion and what I consider to be a polluted atmosphere." This challenge, however, is one that would have had to be addressed in any community to which the case was relocated, and one the trial court took appropriate efforts to address. A second change of venue was not the answer, and the trial court did not err in saying so.

12 Irvin v. Dowd, supra , 366 U.S. 717, 81 S.Ct. 1639, People v. Tidwell (1970) 3 Cal.3d 62, 89 Cal.Rptr. 44, 473 P.2d 748, and People v. Williams (1989) 48 Cal.3d 1112, 259 Cal.Rptr. 473, 774 P.2d 146, upon which Peterson relies to argue it was error to deny a change of venue, are not comparable. In Irvin , in marked contrast to this case, eight of the 12 actual jurors thought the defendant guilty before trial even began. (Irvin , at p. 727, 81 S.Ct. 1639.) And in both Tidwell and Williams , the small size of the venues (Lassen and Placer Counties, respectively) and the comparative prominence of the victims and defendants within those communities — factors notably absent here — weighed in favor of concluding that a change of venue would have made a difference. (Tidwell , at pp. 72–73, 89 Cal.Rptr. 44, 473 P.2d 748 ; Williams , at pp. 1126–1129, 259 Cal.Rptr. 473, 774 P.2d 146.)

C. Admission of Dog Scent Trailing Evidence

Before trial, the court held an Evidence Code section 402 hearing to determine whether to admit the prosecution's evidence of dogs trailing Laci's scent from her home in Modesto, in and around Peterson's warehouse in Modesto, on highways leaving Modesto, and at the Berkeley Marina. After extensive testimony from and cross-examination of various dog handlers concerning their training, as well as their dogs' training and past performance, the court excluded all dog scent trailing evidence as lacking in sufficient foundation and corroboration, except for evidence of scent trailing at the Berkeley Marina. The court explained that all the dog-trailing evidence in and around Modesto lacked sufficient independent corroboration that Laci had ever been where dogs purported to trail her scent. The Berkeley Marina evidence differed because there was corroborating evidence that Laci had been present in the area — namely, that Laci's remains washed ashore on the edge of the San Francisco Bay, very near the marina.

Consistent with the trial court's ruling, dog handler Eloise Anderson, a member of the Contra Costa County Sheriff's Department search and rescue team, testified at trial about scent trailing she conducted at the Berkeley Marina with her dog Trimble. Anderson reported to the marina with Trimble on December 28, 2002, four days after Laci's disappearance. Anderson was provided a glasses case containing a pair of Laci's sunglasses and asked to work with her dog to determine whether any trail of Laci's scent could be detected. The marina harbor had two separate access points where someone might enter, so Anderson had Trimble check each entry point. Wearing rubber gloves to conceal her own scent, Anderson opened the glasses case to expose the sunglasses inside, presented the case with the sunglasses to Trimble, and gave a trailing command. In the first location, Trimble responded with a "no scent trail" signal. In the second location, Trimble "lined out," pulling her harness line taut, with level head, and taking Anderson from an area near the parking lot down one of the marina piers to a pylon on the pier where a boat could have been tied, then giving Anderson an "end of trail" signal. Peterson argues admission of this evidence of dog trailing at the Berkeley Marina was error because the trial court failed to require a hearing under People v. Kelly (1976) 17 Cal.3d 24, 130 Cal.Rptr. 144, 549 P.2d 1240 ( Kelly ), and the prosecution failed to establish the legally required foundation. We disagree.

When admission of expert testimony relating the use of novel scientific methods or techniques is at issue, the proponent of the evidence must demonstrate the technique's reliability through testimony from an expert qualified to opine on the subject. The technique's reliability, in turn, depends on a showing that it has achieved general acceptance among practitioners in the relevant field. Finally, the proponent of the evidence must show any procedures necessary to ensure the technique's validity were properly followed in the given case. ( Kelly, supra , 17 Cal.3d at p. 30, 130 Cal.Rptr. 144, 549 P.2d 1240.) The purpose of these threshold requirements — commonly referred to as the Kelly test — is to protect against the risk of credulous juries attributing to evidence cloaked in scientific terminology an aura of infallibility. ( Id. at pp. 31–32, 130 Cal.Rptr. 144, 549 P.2d 1240.) Not every subject of expert testimony needs to satisfy the Kelly test. Courts determining whether Kelly applies must consider, first, whether the technique at issue is novel, because Kelly " ‘only applies to that limited class of expert testimony which is based, in whole or part, on a technique, process, or theory which is new to science and, even more so, the law.’ " ( People v. Jackson (2016) 1 Cal.5th 269, 316, 205 Cal.Rptr.3d 386, 376 P.3d 528 ( Jackson ).) Second, courts should consider whether the technique is one whose reliability would be difficult for laypersons to evaluate. A " Kelly hearing may be warranted when ‘the unproven technique or procedure appears in both name and description to provide some definitive truth which the expert need only accurately recognize and relay to the jury.’ " ( Jackson , at p. 316, 205 Cal.Rptr.3d 386, 376 P.3d 528.) Conversely, no Kelly hearing is needed when "[j]urors are capable of understanding and evaluating" the reliability of expert testimony based in whole or in part on the novel technique. ( Jackson , at p. 317, 205 Cal.Rptr.3d 386, 376 P.3d 528.)

Several decades ago, in People v. Craig (1978) 86 Cal.App.3d 905, 150 Cal.Rptr. 676 ( Craig ), addressing an issue then of first impression in California, the Court of Appeal concluded no Kelly hearing was needed before introducing dog-trailing evidence. The court explained that dog trailing does not involve standardized techniques and inanimate, fungible instruments whose accepted use in the scientific community may be established, but individual dogs, whose "ability and reliability [should] be shown on a case-by-case basis." ( Craig , at p. 915, 150 Cal.Rptr. 676.) As such, Kelly does not apply; the evidence is admissible if proper foundation is laid concerning the present ability of a particular well-trained dog to trail a human. This "is a fact which, like other facts, may be proven by expert testimony." ( Ibid. ) The Court of Appeal in People v. Malgren (1983) 139 Cal.App.3d 234, 188 Cal.Rptr. 569 ( Malgren ) elaborated on the foundation necessary to introduce dog-trailing evidence. Drawing from the analyses of sister state courts that had wrestled with the same problem, the court identified five points a proponent must establish: "(1) the dog's handler was qualified by training and experience to use the dog; (2) the dog was adequately trained in tracking humans; (3) the dog has been found to be reliable in tracking humans; (4) the dog was placed on the track where circumstances indicated the guilty party to have been; and (5) the trail had not become stale or contaminated." ( Id. at p. 238, 188 Cal.Rptr. 569, disapproved in part by Jackson, supra , 1 Cal.5th at p. 325, 205 Cal.Rptr.3d 386, 376 P.3d 528 [last factor duplicative; see discussion, post ].)

When this court first addressed these issues in Jackson, supra , 1 Cal.5th 269, 205 Cal.Rptr.3d 386, 376 P.3d 528, we endorsed the general approach of Craig and Malgren. We agreed that unlike the sorts of scientific evidence a juror might uncritically accept, for which a threshold Kelly hearing should be held, "[s]cent trailing evidence is not so foreign to everyday experience that it would be unusually difficult for jurors to evaluate. Jurors are capable of understanding and evaluating testimony about a particular dog's sensory perceptions, its training, its reliability, the experience and technique of its handler, and its performance in scent trailing" in a given case. ( Jackson , at p. 317, 205 Cal.Rptr.3d 386, 376 P.3d 528.) We thus approved the general admissibility of "[e]vidence grounded in the ability of particular dogs to perform scent trailing on command ... so long as a proper foundation is laid." ( Id. at p. 320, 205 Cal.Rptr.3d 386, 376 P.3d 528.) In discussing the necessary foundation, we adopted the Malgren factors with one modification. A proponent must establish as background qualifications the adequacy of the handler's and dog's training and supply evidence of the dog's reliability in trailing humans. ( Jackson, supra , 1 Cal.5th at pp. 321–322, 205 Cal.Rptr.3d 386, 376 P.3d 528.) The party must also show the adequacy of the manner in which the dog was given a scent to trail, whether (as in Craig and Malgren ) by being allowed to sniff the beginning of a known trail or (as in Jackson ) by being "presented with a scent article" and then asked to smell for a corresponding trail of the same scent. ( Jackson , at p. 322, 205 Cal.Rptr.3d 386, 376 P.3d 528.) We also approved the corroboration requirement established in Craig, Malgren , and People v. Gonzales (1990) 218 Cal.App.3d 403, 409, 267 Cal.Rptr. 138 ( Gonzales ) — the need for some independent evidence tending to confirm that a person found at the end of the trail the dog followed was indeed the person who left the scent trail and supplied the initial scent. ( Jackson , at p. 321, 205 Cal.Rptr.3d 386, 376 P.3d 528.) We concluded the remaining Malgren condition for admissibility — evidence the source of the initial scent had not become stale or contaminated — was essentially duplicative of the other elements and thus need not be shown independently. "If a well-qualified handler trains a dog who has reliably trailed human scent and is well trained in ignoring or forgetting past smells and in indicating negative trails, then the dog will not trail if the scent on the scent item is stale or nonexistent, or if there is no trail that matches the scent on the scent item." ( Jackson , at p. 325, 205 Cal.Rptr.3d 386, 376 P.3d 528.) Thus, evidence of the remaining Malgren requirements ordinarily will suffice to support admission of dog-trailing evidence. (See People v. Westerfield (2019) 6 Cal.5th 632, 706, 243 Cal.Rptr.3d 18, 433 P.3d 914.)

As in Jackson , we conclude the trial court did not err in declining to subject the dog-trailing evidence to the threshold Kelly test. The nature of the dog-trailing technique at issue here is not meaningfully different from the technique at issue in Jackson , which we concluded was not subject to Kelly. In Jackson , on one occasion, a trained dog was given a gauze pad infused with scent from a fresh shoe print left outside a victim's house and then taken to a lobby through which a suspect had passed. The dog was able to trail from the lobby to an interview room where the suspect was sitting, whereupon the dog alerted to the suspect. ( Jackson, supra , 1 Cal.5th at pp. 308–309, 205 Cal.Rptr.3d 386, 376 P.3d 528.) On a second occasion, a dog was given a gauze pad infused with scent from an envelope left on a different victim's bed and believed to have been handled by the perpetrator. The dog was again asked to seek out and follow any matching trail, and trailed from a point outside the police station to a locker room inside the station, where the suspect was sitting, and again alerted to the suspect. ( Id. at p. 309, 205 Cal.Rptr.3d 386, 376 P.3d 528.) The essential task the dog was asked to perform was to "follow any human scent that she could pick up from the envelope." ( Id. at p. 320, 205 Cal.Rptr.3d 386, 376 P.3d 528.) Here, Trimble's task was not fundamentally different in nature: She was presented with Laci's sunglasses and then directed to smell for trails of the same human scent, if any, at the Berkeley Marina. This was not a novel technique; indeed, Anderson testified that teaching a dog to scent off an object and then seek a corresponding trail is a routine part of training dogs to trail humans. Nor was it a technique whose fallibility would have been opaque to laypersons. Under Jackson , no Kelly hearing was necessary before the evidence was admitted, provided that the requisite foundational requirements were satisfied.

Turning to that necessary foundation, Peterson argues we should supplement the requirements set out in Jackson with additional requirements derived from the Court of Appeal's decisions in People v. Willis (2004) 115 Cal.App.4th 379, 9 Cal.Rptr.3d 235 ( Willis ) and People v. Mitchell (2003) 110 Cal.App.4th 772, 790–794, 2 Cal.Rptr.3d 49 ( Mitchell ). We declined to impose these requirements in Jackson , and there is no reason for a different result here. ( Jackson, supra , 1 Cal.5th at pp. 319–320, 205 Cal.Rptr.3d 386, 376 P.3d 528.)

In Willis , the dog was not asked to smell for a scent trail, but instead was exposed to a scent and then "watched to see if the dog ‘show[ed] interest’ in various locales frequented by the defendant." ( Willis, supra , 115 Cal.App.4th at p. 386, 9 Cal.Rptr.3d 235.) This sort of scent identification, the court held, should be admissible only upon foundation concerning such matters as "how long scent remains on an object or at a location" and "whether every person has a scent that is so unique that it provides an accurate basis for scent identification." ( Ibid. , citing Mitchell, supra , 110 Cal.App.4th at pp. 791–792, 2 Cal.Rptr.3d 49.) In Mitchell , on which Willis relied, a dog had been given pads with scent from murder shell casings and the victim's shirt and a lineup of pads with scents from various people, including the defendants. The dog alerted to a match with pads containing one defendant's scent, but not the other's. ( Mitchell , at pp. 780–781, 2 Cal.Rptr.3d 49.) The Court of Appeal distinguished between scent trailing, which was established as admissible without any Kelly hearing, and a scent lineup of the sort performed in Mitchell. ( Mitchell , at p. 790, 2 Cal.Rptr.3d 49.) Such a lineup, the court reasoned, did not have the kind of centuries-long lineage that scent trailing does, and thus ought to be supported by additional foundation establishing the uniqueness of human scents and their persistence and rate of degradation. ( Id. at pp. 793–794, 2 Cal.Rptr.3d 49.)

Here, the dog-trailing evidence admitted at trial did not resemble either the open-ended identification at issue in Willis or the scent lineup at issue in Mitchell. Trimble was not asked to match a scent to a general location the target may have frequented at unspecified times in the past, nor was she asked to distinguish among multiple people or objects on the basis of their scents. Rather, Trimble was asked to seek out and follow a trail, if any could be found, based on a given scent. Jackson rejected the need for additional foundation tied to scent identification before introducing evidence of a dog's performance of a very similar task. ( Jackson, supra , 1 Cal.5th at pp. 319–320, 205 Cal.Rptr.3d 386, 376 P.3d 528.) No more foundation than that required by Malgren and Jackson for analogous tasks was necessary here.

Again, under the Malgren test, as modified in Jackson , the prosecution was required to show that Anderson was sufficiently trained, that Trimble was sufficiently trained and reliable in tracking humans, that Trimble was properly given an initial scent to trail, and that some evidence tended to corroborate Trimble's trailing. We review the trial court's determination that adequate foundation was laid for abuse of discretion. ( Jackson, supra , 1 Cal.5th at p. 321, 205 Cal.Rptr.3d 386, 376 P.3d 528.) If substantial evidence supports each foundational element, the decision to admit dog-trailing evidence will be upheld. ( Ibid. ; Craig, supra , 86 Cal.App.3d at p. 917, 150 Cal.Rptr. 676.)

Anderson testified about her experience over more than two decades training dogs. She described her dog Trimble's certificates and training, which included certification in trailing from the California Rescue Dog Association and repeated seminars and training exercises over a period of years leading up to her use in trailing Laci's scent in December 2002. Trimble's training included out-of-state sessions to give her experience trailing under varied environmental factors, terrain, and weather conditions. Anderson described many of Trimble's successful human scent trailings, including numerous instances of trailing noncontact trails¹³ and successfully following trails four days old or even older.¹⁴ Anderson opined that Trimble was reliable in trailing humans. Based on the foregoing evidence, which was substantial, the trial court made specific findings that Anderson and Trimble were adequately trained and Trimble was reliable at tracking humans.

13 A noncontact trail is one left by a person not in contact with the ground, as on a bicycle or in a vehicle.

14 For example, in a November 1999 exercise, Trimble successfully trailed a person who had left a trail largely on asphalt, which retains scent more poorly than vegetation. The trail was five days old and there had been intervening rainstorms. In a December 2001 exercise, Trimble successfully followed a four-day-old noncontact trail left by a bicyclist riding the 30 miles from Walnut Creek to Dublin. Given the length of the trail, it was not physically possible for Trimble to cover the entire distance, so at various points she was driven from one section of the trail to a later section, but she was able to relocate and follow the bicyclist's scent each time.

The fourth Malgren element, the adequacy of the fashion in which the dog was given a scent to trail, was established with testimony that Trimble was exposed to a scent object of Laci's, her sunglasses. (See Jackson, supra , 1 Cal.5th at p. 322, 205 Cal.Rptr.3d 386, 376 P.3d 528.) This case is thus unlike Willis , where the dog was given an initial scent from a matchbook but the prosecution failed to present any "proof that [the target] ever touched the matchbook from which a scent was collected." ( Willis, supra , 115 Cal.App.4th at p. 386, 9 Cal.Rptr.3d 235.) Finally, the court found corroboration of Trimble's trailing — reason to think her identification of Laci's scent along a path down to the water was accurate — from evidence that Laci's body would, some months later, wash ashore not far from the marina. That evidence, and its tendency to give credence to Trimble's trailing, are beyond reasonable dispute.¹⁵ Substantial evidence supports the court's determination that the Berkeley Marina dog-trailing evidence was admissible.

15 In his reply brief, Peterson challenges the trial court's ruling on the ground that the prosecution presented no other evidence proving Laci was at the Berkeley Marina. This is true, but it was not an unreasonable inference based on the evidence that was presented. Laci disappeared in Modesto, California, a 90-mile drive inland from Berkeley; her remains washed ashore in the San Francisco Bay, just two to two and one-half miles from the marina; and the marina was one of the closest access points to the bay from Modesto. These facts offer some reason to believe Laci, or her body, was taken to the San Francisco Bay by way of the marina. This is to say nothing of the other evidence that Peterson was responsible for Laci's disappearance and, by his own admission, was at the Berkeley Marina on December 24. The trial court did not err in holding that the corroboration element had been satisfied.

Peterson does not contest the sufficiency of this testimony to establish that Anderson and Trimble had sufficient training in trailing scents and that Trimble had shown herself generally capable of identifying and following human scents. He argues, however, that the prosecution failed to establish specific essential qualifications for both Anderson and Trimble: that Anderson had specialized training to avoid certain handler behaviors, and that Trimble had demonstrated success in particular conditions mirroring those at the Berkeley Marina.

Regarding Anderson's training, Peterson argues that Anderson never testified her training included specific instruction on how to avoid cueing dogs to go in a desired, predetermined direction when trailing. Neither we nor the Courts of Appeal have ever held such specific testimony a mandatory prerequisite. (Cf. Florida v. Harris (2013) 568 U.S. 237, 244, 133 S.Ct. 1050, 185 L.Ed.2d 61 [in the context of warrantless searches challenged under the 4th Amend., the reliability of a drug-sniffing dog should be evaluated under the totality of the circumstances, rather than according to a "strict evidentiary checklist"].)¹⁶ Nor did Peterson argue to the trial court that this more specific foundation should have been supplied, at a time when the prosecution would have had the opportunity to rectify any perceived omission. It is therefore too late to raise the issue now. (See Florida v. Harris , at pp. 248–249, 133 S.Ct. 1050 [failure to voice in the trial court doubts about the sufficiency of dog's training bars raising those concerns for the first time on appeal].)

16 Peterson argues that other jurisdictions require testimony concerning cue avoidance, but review of the published case law he relies on and subsequent decisions belies the assertion. Although some of the cases he cites have referred to such testimony, or the absence thereof, in passing, none holds that such testimony is necessary before dog-trailing evidence may be admitted. (See U.S. v. Trayer (D.C. Cir. 1990) 898 F.2d 805, 809 ; U.S. v. One Million, Thirty-Two Thousand, Nine Hundred Eighty Dollars in U.S. Currency ($1,032,980.00) (N.D.Ohio 2012) 855 F.Supp.2d 678, 699 ; Harris v. State (Fla. 2011) 71 So.3d 756, 768–769, revd. sub nom. Florida v. Harris, supra , 568 U.S. 237, 133 S.Ct. 1050 ; State v. Helzer (2011) 350 Or. 153, 158–159, ; State v. England (Tenn. 2000) 19 S.W.3d 762, 768–769.)

In any event, the record does demonstrate that Anderson was aware of the dangers of trainer cueing and sought to avoid them. She testified that she did everything possible to remain neutral and trained Trimble in specific ways to ignore her as much as possible, and that their joint training included successfully working numerous blind trails where neither knew in advance where the subject had gone. During cross-examination of prosecution witnesses, in the defense case, and in closing argument, Peterson was free to address whether Trimble's performance at the marina was tainted by trainer cueing. (See Florida v. Harris, supra , 568 U.S. at pp. 246–247, 133 S.Ct. 1050 [proof of certification or successful completion of a training program supports presumption of dog's reliability, but subject to cross-examination and introduction of conflicting evidence by the defendant]; Jackson, supra , 1 Cal.5th at p. 297, 205 Cal.Rptr.3d 386, 376 P.3d 528 [defendant introduced expert testimony that particular dog-trailing evidence was unreliable because the procedure used permitted inadvertent cueing].) He chose not to. Of course, had Peterson done so, the prosecution could have argued that if Trimble was merely responding to her handler's subconscious desire to find a path, she would have done so in the first place they checked, rather than giving her handler a "no trail" signal. Ultimately, whether cueing might have affected Trimble's performance is a matter the jury could consider in deciding what weight to give it, but the absence of testimony about specific training designed to avoid cueing did not render the evidence categorically inadmissible. Peterson next argues that the prosecution was required to present foundation that Trimble was reliable in the specific conditions under which she performed scent trailing at the marina. We have never required, and see no basis for demanding, that the proponent of dog-trailing evidence demonstrate past performance in conditions that are identical to the case at hand. All the training and experience a dog receives has relevance in assessing the dog's reliability in trailing humans. ( Jackson, supra ,, 205 Cal.Rptr.3d 386, 376 P.3d 528 [trial court finding of reliability supported by generalized training]; see Florida v. Harris, supra , 568 U.S. at pp. 244–247, 133 S.Ct. 1050 [totality of the circumstances relevant to assessing drug-sniffing dog's reliability].) As with the cueing issue just addressed, a defendant is free to introduce evidence, cross-examine witnesses, and present argument that aspects of the trailing environment made the detection of a scent trail more fraught and subject to error. But the failure of a particular dog to have trailed in circumstances precisely mirroring those in a given case does not preclude the jury from considering trailing evidence.

Nor, in any event, do the specific features of the marina trailing undermine the soundness of the trial court's determination that the prosecution's foundation was adequate. Peterson emphasizes that the trailing involved a "marine" environment. Trimble was not asked to trail at sea; she was asked to trail in a field, parking lot, and dock that were adjacent to a body of water. Peterson offered no evidence that trailing on land in an area adjacent to a body of water poses significant difficulties.¹⁷ Peterson also stresses the noncontact nature of the trail, but Anderson testified at length about Trimble's many successful trailing exercises involving noncontact trails.

17 Indeed, Christopher Boyer, the head of the Contra Costa County Sheriff's Department's search and rescue team, testified that humidity enhances the ability of scent to adhere to surfaces and may make it easier for dogs to trail.

Lastly, under the prosecution's theory, Laci's body was covered by a tarp when she was in the marina, and so Peterson highlights the fact that Trimble's one undisputed failure in a training exercise had involved an enclosed target. But the prosecution presented evidence that on a different occasion, Trimble had previously successfully trailed an enclosed target, following a trail generated by a subject in the closed trunk of a car.¹⁸ Of course, Peterson was also entitled to try to undercut Anderson's testimony by pointing to Trimble's track record in trailing enclosed targets, and he did just that: He cross-examined Anderson about Trimble's failure, as well as the more general difficulties with following a trail left by someone in a closed vehicle. The jury was then specifically instructed that in weighing testimony concerning Trimble's performance, it could consider "any other factor that could affect the accuracy of the dog tracking evidence." The particular circumstances involved in the trailing exercise here may have made Trimble's task more or less difficult, but they did not alter the fundamental nature of the task — which, again, is the type of task dogs have performed for centuries. Those circumstances were thus relevant to the weight the jury might accord this evidence, but not its admissibility.

18 A third test Peterson cites, from October 2002, was the subject of vigorous debate as to whether Trimble succeeded and whether the test was even an enclosed-target test.

Finally, Peterson argues that the dog-trailing evidence should have been excluded because Anderson failed to perform a "missing member" test on Laci's sunglasses. Anderson explained that when a scenting object has two "equally intense" scents on it, a handler may present the object and then have the dog smell the target she does not want the dog to trail. Trailing dogs are then taught to seek out a trail that matches the remaining scent — the person that is missing, i.e., the "missing member." Peterson suggests it is possible that he may have touched Laci's sunglasses at some point in the past. He argues Anderson therefore should have performed a missing member check to eliminate the possibility that Trimble was following Peterson's own scent from when he was, concededly, at the Berkeley Marina a few days earlier.

There are two difficulties with this argument. First, Anderson testified, and Peterson does not dispute, that trailing dogs are trained to follow the freshest, most recent scent on a particular object. There is no evidence in the record that Peterson handled Laci's sunglasses so recently that his scent would be equally fresh and risk confusing Trimble. Indeed, there is nothing in the record to show Peterson handled the sunglasses at all. The most Peterson can point to is hearsay, inadmissible for its truth, that Peterson had handled Laci's purse at some unspecified time in the past. Second, and in any event, conducting a missing member test is not a foundational requirement for admission of dog-trailing evidence. Peterson does not dispute that Laci handled her own sunglasses, and that Trimble was then given the sunglasses as a scent object before attempting to find a trail. Given this, the prosecution adequately established foundation for the fourth Malgren factor — "that the dog was presented with a scent article that the jury could infer was handled by" the target. ( Jackson, supra , 1 Cal.5th at p. 322, 205 Cal.Rptr.3d 386, 376 P.3d 528.) Whether Peterson also handled the sunglasses so recently Trimble became confused was a matter he could, and did, argue to the jury. (See id. at pp. 325–326, 205 Cal.Rptr.3d 386, 376 P.3d 528 [even when foundation established, defendant free to cross-examine witnesses and present rebuttal evidence].) But Peterson fails to establish that the evidence rested on an inadequate foundation to support its admission.

For the same reasons, Peterson fails to establish that admission of the evidence violated his federal constitutional right to a reliable guilt determination in a capital case. ( U.S. Const., 8th & 14th Amends.) The dog-trailing evidence was not, as defendant claims, inherently unreliable; it rested on a solid foundation and could fairly be considered by the jury alongside whatever arguments against its significance and accuracy Peterson chose to muster.

While we see no error in the admission of Trimble's scent trailing, we also see no reasonable possibility that the jury would have returned a different verdict had the scent trailing evidence been excluded. Peterson contends that the evidence was crucial because the prosecution relied heavily on it. But the prosecutor devoted only a few sentences to the subject in the course of closing and rebuttal arguments that stretched for 150 pages of transcript. If credited by the jury, the trailing evidence would have shown Laci was at the Berkeley Marina in late December 2002. The jury, however, already knew Laci's body had been deposited in the San Francisco Bay months before she washed ashore in April 2003, and there were a limited number of access points, among which the Berkeley Marina was closest to where she was found. The trailing evidence did not add much beyond what could already be inferred from other evidence. It also did not rule out the defense's theory "that somebody [else] abducted her" and then disposed of her in the bay. And the trailing testimony aside, there was considerable other circumstantial evidence incriminating Peterson, from the simple fact that Laci's and Conner's bodies washed ashore 90 miles from their home but within sight of where Peterson admitted he went fishing the day they disappeared; to the research Peterson did on bay currents in the weeks preceding her disappearance and the fishing boat he bought but mentioned to no one; to Peterson's inability to explain what he was fishing for in the middle of the day; to his repeated subsequent surreptitious trips to the marina in the weeks after her disappearance; to the many steps he took in the weeks after she went missing — selling her car, exploring sale of the house, turning the nursery into a storage room — that indicated he already knew Laci and Conner were never coming back. Even under the most stringent harmlessness standard, for federal constitutional error ( Chapman v. California (1967) 386 U.S. 18, 24, 87 S.Ct. 824, 17 L.Ed.2d 705 ), it is clear beyond a reasonable doubt exclusion of the limited dog-trailing evidence admitted by the trial court would have had no impact on the jury's determination that Peterson was guilty. D. Instructions on Dog Scent Trailing Evidence

1. CALJIC No. 2.16 Does Not Provide an Alternate Theory of First Degree Murder

The jury was instructed with a modified version of CALJIC No. 2.16, which provided guidance concerning how to evaluate the dog-trailing evidence that had been introduced. Peterson argues that the instruction permitted the jury to find him guilty of first degree murder, without proof of malice, based solely on the dog-trailing evidence once its accuracy was corroborated. This is incorrect.

Because the dog-trailing evidence involved following the scent of a missing victim, not a suspect, the court slightly modified CALJIC No. 2.16. The jury was instructed: "Evidence of dog tracking of the victim has been received for your consideration. This evidence is not, by itself, sufficient to permit an inference that the defendant is guilty of the crime of murder. Before guilt may be inferred, there must be other evidence that supports the accuracy of the dog tracking evidence. The evidence can be direct or circumstantial, and must support the accuracy of the dog tracking evidence. [¶] In determining the weight to give to dog tracking evidence, you should consider: [¶] One, whether or not the handler was qualified by training and experience to use the dog; [¶] Two, whether or not the dog was adequately trained in tracking humans; [¶] Three, whether or not the dog has been found reliable in tracking humans; [¶] Four, whether the dog was placed on the track where circumstances have shown the victim to have been; [¶] Five, whether or not the trail has become stale or contaminated by the environment, weather, or any other factor; [¶] And, six, any other factor that could affect the accuracy of the dog tracking evidence."

Peterson focuses specifically on one sentence in the instruction: "Before guilt may be inferred, there must be other evidence that supports the accuracy of the dog tracking evidence." As we have previously noted (see Jackson, supra , 1 Cal.5th at p. 336, 205 Cal.Rptr.3d 386, 376 P.3d 528 ), this portion of the instruction restates the law as set forth in a 1990 Court of Appeal decision, Gonzales, supra , 218 Cal.App.3d 403, 267 Cal.Rptr. 138. Like other California courts that had previously addressed the issue (see Malgren, supra , 139 Cal.App.3d at pp. 241–242, 188 Cal.Rptr. 569 ; Craig, supra , 86 Cal.App.3d at p. 918, 150 Cal.Rptr. 676 ), Gonzales held that dog-trailing evidence requires corroboration, and it explained why. Corroboration is necessary not because a dog might be untrustworthy — the reason corroboration is required in the case of accomplice testimony — but because it might be inaccurate. ( Gonzales , at pp. 410–412, 267 Cal.Rptr. 138.) Without the ability to question a dog, we cannot be as sure that on a given occasion the dog correctly identified a starting scent and thereafter correctly located and followed a trail of that scent to a given destination. ( Id. at pp. 412–413, 267 Cal.Rptr. 138.) Uncertainty as to the underlying accuracy of the trailing undermines the reliability of any inferences one might otherwise draw from the trailing evidence. (See id. at p. 412, 267 Cal.Rptr. 138.) The corroboration requirement mitigates that uncertainty, ensuring there are "other circumstances supporting the accuracy of the inferences drawn from the dog-tracking evidence." ( Id. at pp. 413–414, 267 Cal.Rptr. 138 ; see id. at p. 414, 267 Cal.Rptr. 138 ["corroborating evidence ... allows assurance that the inferences we draw from any of the various pieces of circumstantial evidence, including the dog-tracking evidence, are correct"].)

As relevant here, however, Gonzales made clear that corroboration does not automatically permit a jury to leap from acceptance of the dog's accuracy to the ultimate conclusion of guilt. The inferences each bit of circumstantial evidence may support are but an intermediate step in the process. (See Gonzales, supra , 218 Cal.App.3d at p. 414, 267 Cal.Rptr. 138 ["As with any circumstantial reasoning process, the ultimate conclusion is predicated upon many inferences that are drawn" from various pieces of circumstantial evidence, including dog trailing evidence].) Without corroboration, the accuracy of dog trail evidence is too uncertain to support any inferences; with corroboration, that evidence may support one or more inferences, which in turn may be relied upon by a jury to reason its way to a final conclusion concerning guilt.

Although Peterson contends otherwise, CALJIC No. 2.16 does correctly capture these principles. (See Jackson, supra , 1 Cal.5th at p. 336, 205 Cal.Rptr.3d 386, 376 P.3d 528.) According to the portion in dispute here, dog trailing "evidence is not, by itself, sufficient to permit an inference that the defendant is guilty of the crime of murder. Before guilt may be inferred, there must be other evidence that supports the accuracy of the dog tracking evidence. The evidence can be direct or circumstantial, and must support the accuracy of the dog tracking evidence." In other words, independent corroborating evidence of the dog's accuracy on the occasion in question is required foundation before any inference pointing toward guilt may be drawn. The instruction is not reasonably read, as Peterson suggests, to equate an inference of guilt based on a single piece of evidence with the ultimate conclusion that the defendant is guilty beyond a reasonable doubt. (See Gonzales, supra , 218 Cal.App.3d at p. 414, 267 Cal.Rptr. 138.)

We see no reasonable likelihood the jury was confused on this point. Here, the trial court concluded the Berkeley Marina dog-trailing evidence could go to the jury because there was corroboration that Trimble had accurately extracted Laci's scent from her sunglasses and then trailed that scent to the end of a pier in the marina — specifically, the fact Laci's body was later found on the shores of the San Francisco Bay, at a point a relatively short distance from the marina. If the jury credited Trimble as accurate, it could then draw inferences pointing toward Peterson's guilt — such as that, within days of the December 28, 2002, trailing, Laci (whether then alive or dead) was at a pier in the marina where Peterson admitted he had been. But as the remainder of the jury instructions made clear, the ultimate conclusion that Peterson had committed premeditated murder would require far more. Moments after instructing the jury with CALJIC No. 2.16, the court gave the jury complete, detailed instructions reminding it that a guilty verdict required every element of a charged crime be proved beyond a reasonable doubt, and spelling out each element that must be shown to prove first degree murder. Reasonably understood, CALJIC No. 2.16 did not supplant these instructions or suggest the jury could bypass considering whether each and every element had been proven beyond a reasonable doubt based solely on the dog-trailing evidence.

2. CALJIC No. 2.16 Did Not Violate Any Right to Balanced Instructions

Peterson argues that the court's dog-trailing evidence instruction was flawed in a second way: It referred to the inculpatory value of dog-trailing evidence without also alluding to its potential exculpatory value. In support of the argument, he emphasizes that the prosecution's evidence was not the only dog-trailing evidence introduced at trial. Peterson presented evidence that on December 28, 2002, the same day Trimble trailed Laci's scent at the Berkeley Marina, a different dog, T.J., was unable to trail Laci's scent at the marina. Peterson contends the trial court's dog-trailing instruction was unfairly one-sided in that it told the jury about the circumstances under which the dog-trailing evidence could be used to convict, without also mentioning that the evidence could be used to support acquittal. We are unpersuaded.

Peterson's argument rests on an analogy to Cool v. United States (1972) 409 U.S. 100, 93 S.Ct. 354, 34 L.Ed.2d 335 ( Cool ). There, an alleged accomplice of the defendant gave exculpatory testimony. ( Id. at p. 103, fn. 4, 93 S.Ct. 354.) Despite the exculpatory character of the evidence, the court instructed the jury that it could rely on the evidence to convict, never mentioning that the jury could also rely on the evidence to acquit: " ‘I further instruct you that testimony of an accomplice may alone and uncorroborated support your verdict of guilty of the charges in the Indictment if believed by you to prove beyond a reasonable doubt the essential elements of the charges in the Indictment against the defendants.’ " ( Ibid. ) The Supreme Court concluded this instruction required reversal: "[E]ven if it is assumed that [the alleged accomplice's] testimony was to some extent inculpatory, the instruction was still fundamentally unfair in that it told the jury that it could convict solely on the basis of accomplice testimony without telling it that it could acquit on this basis." ( Ibid. ) Peterson contends that here, as in Cool , due process required an instruction that explicitly informed the jury it could acquit on the basis of dog-trailing evidence. The analogy to Cool fails for several reasons. First, as just discussed, the instruction did not tell the jury it could convict based on the dog-trailing evidence alone. Rather, it told the jury that such evidence, if (and only if) corroborated, could be used to support an inference of guilt. An inference is not the same as a conclusion that each element has been shown beyond a reasonable doubt (see ante , 268 Cal.Rptr.3d at pp. 98–100, 472 P.3d at pp. 417–419), and so the instruction here did not replicate the defect in Cool : It did not put a thumb on the scale of the jury's deliberations by informing them they could return a guilty verdict based entirely on one piece of the prosecution's evidence. It instead placed limits on the circumstances in which the jury could consider the prosecution's dog-trailing evidence, along with other evidence, as supporting a conclusion of guilt.

Second, it would have made no sense to impose similar limits on the defense's dog-trailing evidence; the dog the defense pointed to did not find any trail, so there was no trailing evidence that would require corroboration. And, of course, no limiting instruction was necessary to allow the defense to use its dog-trailing evidence for the purpose it desired: that is, as reason for the jury not to give the prosecution's dog-trailing evidence much weight. The defense could, and did, argue that T.J.'s failure to find a trail on December 28 pointed to Trimble having been mistaken.

Finally, and in any event, the dog-trailing evidence here was of substantially less moment in the context of the overall case than the accomplice testimony in Cool. There, the defendant "relied primarily on the testimony of" the alleged accomplice. ( Cool, supra , 409 U.S. at p. 101, 93 S.Ct. 354.) The "Government's position clearly depended upon its ability to discredit [the alleged accomplice], since his testimony was completely exculpatory." ( Ibid. ) In such a context, an instruction that told the jury that that testimony could be used by itself to convict, but not to acquit, rose to the level of a due process violation and warranted reversal of the entire conviction. Here, in contrast, the lone dog-trailing witness called by the defense testified for a few transcript pages, during the course of a guilt phase trial that lasted more than five months. The failure of the modified version of CALJIC No. 2.16 to tell the jury explicitly that it could acquit based on that scant absence-of-trailing evidence did not give rise to any comparable due process violation.¹⁹ E. Admission of Expert Testimony Concerning the Trajectory of Conner's Body in the San Francisco Bay

19 Peterson makes one other argument — that the dog-trailing evidence immediately followed a court instruction on motive that allowed the jury to consider the presence of a motive as weighing in favor of guilt and the absence of a motive as weighing in favor of acquittal, and a reasonable juror would have inferred from this sequence that by negative inference it was not permitted to consider dog-trailing evidence as supporting acquittal. The motive instruction actually came after the dog-trailing evidence instruction, and, in any event, we do not read the dog-trailing instruction as reasonably susceptible to the understanding that jurors could not consider such evidence as supporting acquittal.

Dr. Cheng, a senior research hydrologist for the United States Geological Survey, was called as an expert witness by the prosecution. Dr. Cheng testified to his training, publications, and experience analyzing fluid dynamics, with a special focus on the San Francisco Bay. The trial court accepted him as an expert hydrologist qualified to testify about the movement of water in the bay and related topics. Dr. Cheng then described two analyses he did for the Modesto Police Department. In February 2003, while Laci was still missing, Dr. Cheng analyzed where she might most likely be found based on an assumption given him by the police that Laci's body had been placed in the bay in a particular area. In May 2003, after Laci's and Conner's bodies were found, Dr. Cheng performed a second analysis to supply his best estimate concerning where they might have originated, and thus where divers should search for missing body parts and any weights or other evidence. Dr. Cheng offered his opinion that a particular region near Brooks Island represented the most likely starting point for Conner, but was unable to estimate any most likely origin for Laci.

There is no dispute that Dr. Cheng's training and experience qualified him as an expert in the area of fluid dynamics and, specifically, the flow of the waters of the San Francisco Bay. Defense counsel conceded Dr. Cheng was qualified to give expert testimony on hydrology. But during the Evidence Code section 402 hearing that preceded Dr. Cheng's testimony, Peterson asked the court to require that foundation for Dr. Cheng's testimony be laid under Kelly, supra , 17 Cal.3d 24, 130 Cal.Rptr. 144, 549 P.2d 1240. The court denied the request. Peterson now contends Dr. Cheng's opinions as to the movement of the victims' bodies in the bay were inadmissible. We review the decision to admit the expert testimony for abuse of discretion. ( People v. Banks (2014) 59 Cal.4th 1113, 1190, 176 Cal.Rptr.3d 185, 331 P.3d 1206.) We conclude the court did not err.

As discussed earlier, Kelly imposes certain preconditions on the admission of evidence derived from a novel scientific technique or procedure. The additional scrutiny "is justified because ‘[l]ay jurors tend to give considerable weight to "scientific" evidence when presented by "experts" with impressive credentials. We have acknowledged the existence of a "... misleading aura of certainty which often envelops a new scientific process, obscuring its currently experimental nature." ’ " ( People v. Jones, supra , 57 Cal.4th at p. 952, 161 Cal.Rptr.3d 295, 306 P.3d 1136, quoting Kelly, supra , 17 Cal.3d at pp. 31–32, 130 Cal.Rptr. 144, 549 P.2d 1240.)

But in most cases no similar caution is required before a jury considers expert opinion testimony. Unlike results "produced by a machine," to which jurors may "ascribe an inordinately high degree of certainty," jurors presented with the personal opinion of a witness, even an expert witness, "may temper their acceptance of his [or her] testimony with a healthy skepticism born of their knowledge that all human beings are fallible." ( People v. McDonald (1984) 37 Cal.3d 351, 372, 208 Cal.Rptr. 236, 690 P.2d 709 ; accord, People v. Jones, supra , 57 Cal.4th at p. 953, 161 Cal.Rptr.3d 295, 306 P.3d 1136.) For this reason, " ‘[a]bsent some special feature which effectively blindsides the jury, expert opinion testimony is not subject to Kelly [ ].’ " ( People v. Eubanks (2011) 53 Cal.4th 110, 140, 134 Cal.Rptr.3d 795, 266 P.3d 301, quoting People v. Stoll (1989) 49 Cal.3d 1136, 1157, 265 Cal.Rptr. 111, 783 P.2d 698.) Of course, some expert testimony may be "based, in whole or part, on a technique, process, or theory which is new to science and, even more so, the law" ( Stoll , at p. 1156, 265 Cal.Rptr. 111, 783 P.2d 698 ); where the novel technique "appears in both name and description to provide some definitive truth which the expert need only accurately recognize and relay to the jury," additional scrutiny under Kelly is warranted. ( Ibid. ; see People v. Cowan (2010) 50 Cal.4th 401, 470, 113 Cal.Rptr.3d 850, 236 P.3d 1074.) But this case does not fit that description.

Dr. Cheng began with an overview for the jury of how various forces, including tides, currents, and wind, interact and affect the waters of the bay. He then explained how, given the time and location where Conner's body was found, he worked backward to estimate where Conner was most likely to have started. Tidal currents in the area were weak and would likely have canceled out, so Dr. Cheng treated wind-driven drift as the principal force that would have moved Conner's body to shore. He accepted as a starting assumption the hypothesis that Laci and Conner had been weighted down and then broken free and treated Conner as a floating body thereafter. Performing calculations using an equation drawn from the United States Army Corps of Engineers Coastal Engineering Handbook to translate measured wind speeds into a corresponding rate of water movement, Dr. Cheng worked backward to estimate the most probable path Conner's body would have followed.

Dr. Cheng's modeling involved no novel technique. The study of tides and currents and their effect on the motion of bodies in water is hardly new. Nor is the scientific understanding of how wind affects the movement of water — and thus bodies in water — of recent origin. Indeed, as Dr. Cheng explained to the jury, much of his modeling involved applying established, published equations to the known conditions in the hours and days before Conner's and Laci's bodies were found. The application of settled principles to estimate the motions of bodies in water did not require a Kelly hearing.

While the absence of a novel technique alone disposes of Peterson's Kelly argument (see People v. Stoll, supra , 49 Cal.3d at p. 1156, 265 Cal.Rptr. 111, 783 P.2d 698 ), we also note that, as described by Dr. Cheng, the technique was not one that carried a "misleading aura of scientific infallibility" ( id. at p. 1157, 265 Cal.Rptr. 111, 783 P.2d 698 ) — the primary danger Kelly is designed to guard against. Dr. Cheng apprised the jury of the ways in which the model's accuracy was dependent on a host of initial assumptions, many of which carried with them a considerable degree of uncertainty. For example, it was unknown when Conner's body had actually washed ashore, which would have altered best estimates of his starting point, and no starting point could be estimated at all for Laci, who was found a day later at a point nearly one mile away from Conner. Moreover, because divers found no evidence to confirm Conner or Laci had ever been at the starting location Dr. Cheng suggested, the jury had ample basis to take his opinions with more than a dash of salt.²⁰ Because Dr. Cheng's testimony did not describe to the jury a mechanistic process that pointed unerringly to firm conclusions, it presented no risk of blindsiding the jury and precluding it from critically evaluating his ultimate opinion. (See People v. Garlinger (2016) 247 Cal.App.4th 1185, 1196–1197, 203 Cal.Rptr.3d 171 [no risk jury would accord aura of infallibility to cell phone networks expert who "did not purport to be able to determine the precise location" of a cell phone].)

20 Indeed, recognizing that Dr. Cheng's testimony might help the defense as much as it hurt, defense counsel said before Dr. Cheng testified, "[I]n some ways, I want it to come in because I believe his ultimate conclusion is that he can't say anything about Laci."

Peterson relies heavily on People v. Leahy (1994) 8 Cal.4th 587, 34 Cal.Rptr.2d 663, 882 P.2d 321, but Leahy bears no resemblance to the case here. There, police officers testified to the results of a " ‘horizontal gaze nystagmus’ " test purported to reliably indicate intoxication. ( Id. at p. 605, 34 Cal.Rptr.2d 663, 882 P.2d 321.) We explained that the test was a " ‘new’ " technique for purposes of Kelly. ( Leahy , at p. 606, 34 Cal.Rptr.2d 663, 882 P.2d 321.) Further, with its " ‘ "pretentiously scientific name," ’ " the test was one that might "appear[ ] to provide to the jury a ‘definitive truth.’ " ( Ibid. ) As such, the test involved the precise risk underlying the exception to the rule that expert testimony need not satisfy Kelly — it was a novel technique that "appears in both name and description to provide some definitive truth which the expert need only accurately recognize and relay to the jury." ( People v. Stoll, supra , 49 Cal.3d at p. 1156, 265 Cal.Rptr. 111, 783 P.2d 698.) The same is not true of the methodology underlying Dr. Cheng's testimony.

Peterson also criticizes the extent of Dr. Cheng's expertise as it relates to the motion of objects in fluids, as opposed to the motion of fluids themselves. The criticism is not well taken. For one, Dr. Cheng had a Ph.D. in aeronautical engineering and had studied the movement of drifters, devices used to measure currents at the surface and at depth. For another, such criticisms go to the weight a jury might accord Dr. Cheng's testimony, but not its admissibility or whether Kelly foundation was required. (See People v. Eubanks, supra , 53 Cal.4th at p. 143, 134 Cal.Rptr.3d 795, 266 P.3d 301.) Peterson did, in fact, raise his concern about Dr. Cheng's background during cross-examination, and the jury could consider that concern. In sum, Dr. Cheng, an expert hydrologist, could offer his opinion as to the most probable movement of Conner's body in the San Francisco Bay without the court first conducting a Kelly hearing. The court did not abuse its discretion in admitting his testimony.

F. Issues Concerning the Stability of Peterson's Boat

1. Exclusion of Videotaped Defense Demonstration

During the defense's presentation of evidence, the court held an Evidence Code section 402 hearing to evaluate the admissibility of a videotaped experiment the defense had conducted from a boat on the San Francisco Bay. In the experiment, an employee of defense counsel's law firm tried to push a 150- to 155-pound dummy out of the boat. The boat partially sank before the employee abandoned ship.

The prosecution submitted a litany of objections. From the videotape, it was not possible to determine the kind of boat used because identifying features had been covered up; the boat's seats were mounted on wood, thus raising the boat's center of gravity; the boat had a different motor than Peterson's boat; the person performing the experiment was wearing a weight belt that impeded his movements; nothing established the prevailing tides, currents, and waves, or how closely they approximated the conditions on December 24, 2002; nothing established the location of the experiment; the boat's gas tank and batteries were located in different positions than on Peterson's boat; the boat had plywood decking added, which could affect its stability; the dummy got wet, which would have added weight; the boat was already starting to take on water before the employee even tried to dump the dummy; and the employee performed the experiment while standing on the boat's gunwales, which suggested he was "intentionally trying to sink the boat." After reviewing the tape, the court ruled it inadmissible under Evidence Code section 352. Peterson contends this was error.

" ‘Under Evidence Code section 352, the trial court has wide discretion to admit or reject experimental evidence. We reverse decisions to admit or exclude such evidence only when the trial court has clearly abused its discretion.’ " ( People v. Jones (2011) 51 Cal.4th 346, 375–376, 121 Cal.Rptr.3d 1, 247 P.3d 82.) Before experimental evidence may be admitted, the proponent must establish that the experiment is relevant, was " ‘ "conducted under substantially similar conditions as those of the actual occurrence," ’ " and will not mislead or confuse the jury or take undue time. ( Jackson, supra , 1 Cal.5th at p. 342, 205 Cal.Rptr.3d 386, 376 P.3d 528.)

The prosecution raised a host of ways in which the defense did not carry its burden of establishing that its demonstration's conditions sufficiently resembled conditions on December 24, 2002. The court could and did rely on these points, including Peterson's failure to establish the similarity of the boat used, the weather, and the location. From our own review of the proffered video, we cannot say the court abused its broad discretion to determine the video's admissibility. (See People v. Jones, supra , 51 Cal.4th at pp. 375–376, 121 Cal.Rptr.3d 1, 247 P.3d 82.) The video shows the waters of the bay were extremely choppy during the experiment. Even before the defense employee tried to hoist the dummy, the boat was taking on water because the employee was stepping on the boat's gunwale, on the same side as an elevated seat (Peterson's boat had no similar elevated seat) and the side toward which the boat's roughly 75-pound motor had been angled. This concentration of weight allowed waves to break over the boat's gunwales. However stable the boat might have been if attempts were made to counterbalance, no such attempts were made in the defense's demonstration. The boat also had plywood decking added, which would have not only raised the center of gravity, but also concealed whatever might have lain underneath.

Peterson emphasized to the trial court that it had already admitted evidence of a prosecution experiment in which a district attorney's office employee, also late in her third trimester of pregnancy and weighing the same as Laci, lay down in Peterson's boat to show how someone could fit without being noticeable. Peterson strenuously argued that, by parity of reasoning, his experiment should be admissible too. He makes the same point on appeal. But the two situations are not symmetric, and that asymmetry explains why a court could exercise its discretion to admit one experiment but not the other. As discussed, Peterson laid no foundation establishing the extent to which the conditions during his videotaped experiment mirrored those that would have obtained during the alleged disposal of the body. In contrast, the prosecution's photographs showed precisely the feasibility of what they posited had occurred — that the body of a pregnant, 153-pound woman could be curled to fit in the bottom of Peterson's actual boat.

The two experiments are asymmetric in a second way as well. A person trying to dispose of a body by dumping it out of a boat into the bay would have had every incentive to counterbalance against the body's weight in order to avoid capsizing the boat. The defense firm's employee, in contrast, had an incentive to undermine the experiment's results, and indeed the prosecution argued that the video showed he was trying to do just that. To be sure, the prosecution's employee had an incentive to make herself fit, but that incentive does nothing to detract from the persuasive force of the prosecution's experiment as evidence of the possible. No matter how much the pregnant employee might have wanted to make the demonstration work, she could not have unless it was, in fact, possible for her to fit her body in that space. As the court noted, "All she had to do was lay there. She didn't have to demonstrate throwing something — throwing something in the water." Consequently, while the defense demonstration was not irrelevant, it was within the court's discretion to conclude the defense had not laid sufficient foundation to establish the demonstration — conducted in a different boat, under different conditions, by a defense employee — bore a sufficiently close resemblance to how Peterson allegedly disposed of Laci's body, so as to avoid misleading the jury. That the trial court admitted photographs of the prosecution's demonstration does not alter this conclusion. The refusal to admit the video was likewise not erroneous under federal law. Peterson argues that the exclusion of the evidence violated his rights under the Fifth and Sixth Amendments to a fair trial and to submit evidence relevant to the question of guilt. (See Pennsylvania v. Ritchie (1987) 480 U.S. 39, 56, 107 S.Ct. 989, 94 L.Ed.2d 40 ; Chambers v. Mississippi (1973) 410 U.S. 284, 294, 93 S.Ct. 1038, 35 L.Ed.2d 297.) Here, however, the court reasonably found the evidence to be unreliable and misleading. The Constitution contains no right to admit such evidence.

2. Conditions on Reprise of Defense Experiment

In the course of ruling the defense's boat experiment inadmissible, the court made clear it was not categorically barring any such experiment, only the experiment as originally conducted and submitted. Counsel complained that the ruling gave the defense no way to oppose the prosecution's case. While cautioning that it was "not here to give you advice," the court offered two ways in which the defense might choose to modify the experiment to increase its odds of admissibility: "Number one, you take out the original boat instead of this boat. [Number two, y]ou have someone that doesn't work for you conduct the experiment, you know. That would be two things that are out of the way." The court also suggested the defense try to establish what the conditions were like on the day Peterson was alleged to have dumped Laci's body in the bay.

Shortly after denying admission of the defense's boat experiment, the court voluntarily revisited the issue and again made clear the defense could redo the experiment in a way that might make it admissible, if it so chose. The court offered the defense access to Peterson's boat to redo the experiment, adding "I think you should have representatives of the [P]eople there to observe what happens." Later, the court reiterated, "I'm willing to let you have the boat. I'm willing to let you put the 150-pound person in it. I'm willing to let you have somebody, but I want the prosecution to be present and watch the way this thing went down. What's wrong with that?" The court also said it would be helpful if the reenactment included footage showing where Brooks Island was, relative to the demonstration, and was conducted in the general area the prosecution's expert, Dr. Cheng, had suggested was most likely. If the defense wanted to redo the experiment with some adjustments, the court offered to revisit its ruling. Ultimately, Peterson elected not to redo the experiment.

Peterson now contends that by offering to revisit its ruling, but conditioning access to the boat on the prosecution being able to observe any reenactment, the court committed presumptively prejudicial error compelling automatic reversal. Peterson has not preserved any claim of error.

During the Evidence Code section 402 hearing on admissibility of the defense's video and its aftermath, Peterson did not object to the trial court's suggested condition, even as the court invited objection by asking, "What's wrong with that?" Nor did Peterson move for the court to grant him access to the boat and to lift any condition that a member of the prosecution view experiments with that boat, or in any other way object to the court's conditional offer of access to Peterson's boat. Instead, counsel thanked the court for its openness to revisiting its decision and said he would check the weather for suitable conditions and confer with his client.

At the close of the Evidence Code section 402 hearing, then, here is where matters stood. The defense had presented a video of a demonstration on the bay, and the court had — permissibly, as we have just discussed — exercised its discretion to exclude that video. The court had also, sua sponte, made clear that its ruling did not foreclose a future attempt to videotape a further demonstration and resubmit a video for consideration. And finally, the court had suggested different ways the defense could modify the original experiment, if it so chose, including (1) using a different test subject, unaffiliated with the defense, to perform the experiment or (2) using the Peterson boat, with a prosecution observer. Peterson did not object to the latter condition; he instead let the matter drop.

The People argue that because Peterson elected not to conduct an experiment with the prosecution present, he has forfeited any claim on appeal. This goes too far. To preserve a claim that the court has unconstitutionally placed a condition on the defendant's introduction of evidence, a defendant need not always comply with the condition. (See, e.g., United States v. Nobles (1975) 422 U.S. 225, 95 S.Ct. 2160, 45 L.Ed.2d 141 [addressing the merits after the defendant declined to comply with an allegedly unconstitutional condition and elected not to have a witness testify].) But if the defendant objects to the condition, he does need to make his objection known (see, e.g., People v. Varghese (2008) 162 Cal.App.4th 1084, 1090–1091, 76 Cal.Rptr.3d 449 ), unless doing so would be futile ( People v. Perez (2020) 9 Cal.5th 1, 7–8, 259 Cal.Rptr.3d 195, 459 P.3d 1 ). Here, nothing in the record suggests it would have been futile to raise the issue with the trial court. It appears the trial court was merely throwing out ideas as to how the defense could redo the experiment in a way that would address the deficiencies of its first effort, not giving its final word on the subject.

In this and other respects, this case differs from Prince v. Superior Court (1992) 8 Cal.App.4th 1176, 1179–1181, 10 Cal.Rptr.2d 855, which held that a trial court could not condition the ability of the defense to test a critical DNA sample on disclosure of any results to the People. Here, we do not know whether the trial court actually and finally conditioned the defense team's ability to test the stability of the Peterson boat on the presence of a prosecution observer, because Peterson never raised any concerns about the trial court's offer to revisit its exclusion ruling under those circumstances. Had the court said nothing at all after excluding the original, flawed demonstration, Peterson would have no argument. That the court volunteered one way in which Peterson could rectify the problems with the original demonstration did not inject constitutional error into the proceedings, let alone presumptively prejudicial error compelling automatic reversal.

3. Prosecutorial Misconduct in Closing Argument

During closing argument, the prosecution addressed the stability of Peterson's boat. According to the prosecutor, although the defense had insinuated a boat like his was "ready to tip over at the drop of a hat" and would have capsized if Peterson had tried to push Laci's body overboard, "there's no evidence [the boat] would have done that." He recounted the testimony of prosecution witnesses indicating the boat was stable enough to pull heavy fish on or push heavy weights off and concluded, "There's no evidence to contradict that whatsoever." Peterson argues these comments were reversible misconduct because they improperly took advantage of the exclusion of the defense's evidence about the boat's instability. We disagree; this was not misconduct.

"Prosecutorial misconduct requires reversal when it ‘so infect[s] a trial with unfairness [as to] create a denial of due process. [Citations.] Conduct by a prosecutor that does not reach that level nevertheless constitutes misconduct under state law, but only if it involves the use of deceptive or reprehensible methods to persuade the court or jury.’ " ( People v. Armstrong, supra , 6 Cal.5th at p. 795, 243 Cal.Rptr.3d 105, 433 P.3d 987.)

Preliminarily, Peterson has forfeited his claim that the prosecutor engaged in prejudicial misconduct. To preserve a claim for review, a defendant must object and ask that the jury be admonished concerning the misconduct. ( People v. Daveggio and Michaud (2018) 4 Cal.5th 790, 853, 231 Cal.Rptr.3d 646, 415 P.3d 717 ; People v. Bryant, Smith and Wheeler (2014) 60 Cal.4th 335, 426–427, 178 Cal.Rptr.3d 185, 334 P.3d 573 ; People v. Watkins (2012) 55 Cal.4th 999, 1031, 150 Cal.Rptr.3d 299, 290 P.3d 364.) Peterson concedes he did not object, but contends objection was excused because prosecutors are held in such high regard that no admonishment from a court could cure any harm. If Peterson were correct, then no criminal defendant would ever need object to perceived prosecutorial misconduct. The case law is, of course, to the contrary. (See People v. Panah (2005) 35 Cal.4th 395, 462, 25 Cal.Rptr.3d 672, 107 P.3d 790 ["A defendant claiming that one of these exceptions [to the objection requirement] applies must find support for his or her claim in the record. [Citation.] The ritual incantation that an exception applies is not enough."].) Alternatively, Peterson argues that counsel was ineffective for failing to object. But counsel's performance was not deficient; any such objection would have been meritless and properly overruled.

As the record stood at the close of the guilt phase, there was, in fact, no evidence affirmatively supporting the argument that Peterson's boat was too unstable for Peterson to have thrown Laci off it. Peterson cross-examined the prosecution's witnesses in an attempt to cast doubt as to whether their testimony adequately established the boat's stability. But Peterson introduced no evidence of instability to contradict that testimony. The prosecution's observations about this omission were thus fair comment on the state of the evidence.

Peterson does not contend otherwise, but instead urges that the only reason the record contains no such evidence is that the prosecution moved, successfully, to exclude the aforementioned video of the defense's demonstration in which a surrogate boat capsized. In Peterson's view, it was misconduct for the prosecution to obtain the exclusion of evidence and then comment on the resulting evidentiary vacuum.

That is not the law, as we explained in People v. Lawley (2002) 27 Cal.4th 102, 115 Cal.Rptr.2d 614, 38 P.3d 461 ( Lawley ). In Lawley , the defendant sought to introduce evidence of third party culpability, but the court excluded it at the prosecution's urging — a ruling we upheld as correct. ( Id. at pp. 151–155, 115 Cal.Rptr.2d 614, 38 P.3d 461.) The prosecutor then argued in closing that no one but the defendant had a motive to kill the victim. The defendant urged this as misconduct, but we explained that, in light of the court's correct evidentiary rulings, the prosecution's argument was fair comment on the record as it stood. ( Id. at p. 156, 115 Cal.Rptr.2d 614, 38 P.3d 461.) Lawley makes clear that it is not misconduct for the prosecutor to argue in closing that there was no evidence supporting a particular proposition after the trial court has properly excluded evidence the defense had sought to introduce on that point.

Peterson relies on other cases, but they do not establish that the prosecution is barred from ever remarking on an evidentiary gap after successfully moving to exclude evidence that would have filled that gap. The only two California cases he cites, People v. Daggett (1990) 225 Cal.App.3d 751, 275 Cal.Rptr. 287 and People v. Varona (1983) 143 Cal.App.3d 566, 192 Cal.Rptr. 44, were likewise invoked by the defendant in Lawley ; we distinguished them as "inapposite" because "each involved erroneous evidentiary rulings on which the prosecutor improperly capitalized during his closing argument." ( Lawley, supra , 27 Cal.4th at p. 156, 115 Cal.Rptr.2d 614, 38 P.3d 461, italics added.) Here, by contrast, there was no error in the exclusion of the evidence.²¹

21 Our recent decision in People v. Armstrong, supra , 6 Cal.5th 735, 243 Cal.Rptr.3d 105, 433 P.3d 987, is similarly distinguishable. There, we found misconduct after a prosecutor persuaded the trial court to exclude defense evidence of what the victim said before she was attacked — evidence that should have been admitted — and then attributed to the victim a different statement nowhere supported in the record. (Id. at pp. 785–787, 796–797, 243 Cal.Rptr.3d 105, 433 P.3d 987.) The prosecutor here, in contrast, commented only on the admissible evidence and did so accurately.

Peterson's cases from other jurisdictions are likewise distinguishable because each involved erroneously excluded evidence or a prosecutor deceiving the jury into making inferences the prosecutor knew to be untrue. (Cf. Paxton v. Ward (10th Cir. 1999) 199 F.3d 1197 [prosecution misled the jury as to the reason charges against the defendant for a prior shooting had been dropped]; U.S. v. Ebens (6th Cir. 1986) 800 F.2d 1422 [reversal required where the prosecution suggested to the jury inferences it knew to be untrue]; U.S. v. Toney (6th Cir. 1979) 599 F.2d 787 [erroneous exclusion of evidence corroborating the defendant's story prejudicial because the prosecution stressed the absence of corroborating evidence]; State v. Bass (1996) 121 N.C.App. 306 [prosecution argued inference to the jury that it knew to be untrue].) These cases involved " ‘deceptive or reprehensible methods to persuade the court or jury.’ " ( People v. Armstrong, supra , 6 Cal.5th at p. 795, 243 Cal.Rptr.3d 105, 433 P.3d 987.) Here, instead, the prosecutor commented fairly on the record. There was no misconduct.

4. Juror Examination of the Physical Evidence During Deliberations

Peterson's boat was admitted into evidence. During trial, the jury was given the opportunity to inspect it. During deliberations, the jurors asked to look at the boat again. The court agreed and had them do so with the prosecution and defense present. Some jurors asked to sit in the boat. The court acquiesced, reasoning that they should be permitted to test for themselves the evidence the prosecution had submitted concerning how a person (or body) could fit in the bottom of the boat. While in the boat, at least two jurors stood and, by shifting their weight back and forth, rocked the boat. The court cautioned the jurors that they should keep in mind the boat was secured on a trailer, not in the water. Peterson argued that the jurors' attempts to rock the boat constituted an impermissible experiment and sought an opportunity to reopen the evidence and submit his excluded boat demonstration or, in the alternative, a mistrial. The court denied both motions.

Peterson argues the denial of a mistrial based on juror experimentation violated both state and federal law. The denial of a motion for a mistrial is generally reviewed for abuse of discretion. ( People v. Bell (2019) 7 Cal.5th 70, 121, 246 Cal.Rptr.3d 527, 439 P.3d 1102.) Where, however, the motion rests on allegations of juror misconduct and the facts underlying those allegations are essentially undisputed, we review de novo whether misconduct occurred. ( People v. Collins (2010) 49 Cal.4th 175, 242, 110 Cal.Rptr.3d 384, 232 P.3d 32 ( Collins ).) There was no misconduct, and thus no error in denying the motion for a mistrial.

The framework for analyzing whether jury experimentation is permissible or misconduct was established more than a century ago in Higgins v. L.A. Gas & Electric Co. (1911) 159 Cal. 651, 656–657, 115 P. 313 ( Higgins ): "It is a fundamental rule that all evidence shall be taken in open court and that each party to a controversy shall have knowledge of, and thus be enabled to meet and answer, any evidence brought against him. It is this fundamental rule which is to govern the use of such exhibits by the jury. They may use the exhibit according to its nature to aid them in weighing the evidence which has been given and in reaching a conclusion upon a controverted matter. They may carry out experiments within the lines of offered evidence, but if their experiments shall invade new fields and they shall be influenced in their verdict by discoveries from such experiments which will not fall fairly within the scope and purview of the evidence, then, manifestly, the jury has been itself taking evidence without the knowledge of either party, evidence which it is not possible for the party injured to meet, answer, or explain."

We reviewed these principles and endorsed them anew in Collins, supra , 49 Cal.4th at pages 243 to 249, 110 Cal.Rptr.3d 384, 232 P.3d 32. After discussing Higgins , the cases it relied on, and the many cases that had followed it, we elaborated on the guiding principles: "Not every jury experiment constitutes misconduct. Improper experiments are those that allow the jury to discover new evidence by delving into areas not examined during trial. The distinction between proper and improper jury conduct turns on this difference. The jury may weigh and evaluate the evidence it has received. It is entitled to scrutinize that evidence, subjecting it to careful consideration by testing all reasonable inferences. It may reexamine the evidence in a slightly different context as long as that evaluation is within the ‘ "scope and purview of the evidence." ’ [Citation.] What the jury cannot do is conduct a new investigation going beyond the evidence admitted." ( Collins , at p. 249, 110 Cal.Rptr.3d 384, 232 P.3d 32.)

The propriety of the jurors' rocking the boat thus turns on whether they were merely scrutinizing the evidence admitted on a question joined at trial, or instead "invad[ing] new fields" that lay outside the "scope and purview of the evidence." ( Higgins, supra , 159 Cal. at p. 657, 115 P. 313.) The question of the boat's stability was already contested, and evidence had been submitted on this very question. The boat itself was admitted into evidence. Rocking the boat to get some rough sense of its stability did not expand the issues in the case or amount to a taking of new evidence on a previously unexamined question — although, to be clear, the information to be gleaned from assessing the boat's stability on land, on a trailer, as opposed to in water, was likely minimal.

Peterson invokes a number of cases finding impermissible experimentation, but these cases generally involved experiments with items outside the scope of the evidence in the case. In Smoketree-Lake Murray, Ltd. v. Mills Concrete Construction Co. (1991) 234 Cal.App.3d 1724, 1745–1749, 286 Cal.Rptr. 435, a juror used kitty litter and crayons to model the pouring of concrete. In People v. Castro (1986) 184 Cal.App.3d 849, 852–854, 229 Cal.Rptr. 280, a juror used an entirely different pair of binoculars to determine whether a witness who used binoculars could have seen what he testified to seeing. In People v. Conkling (1896) 111 Cal. 616, 627–628, 44 P. 314, jurors used a rifle other than the murder weapon to conduct tests on the distance at which powder marks would show. And in Bell v. State of California (1998) 63 Cal.App.4th 919, 930–933, 74 Cal.Rptr.2d 541, after the plaintiff claimed he had been falsely arrested, placed in an awkward hold, and forced to walk, a juror tested whether one could walk in the manner described by trying to recreate the hold using different people of different size, strength, and so on. Peterson also relies on Wilson v. U.S. (9th Cir. 1902) 116 F. 484, cited with approval in Higgins, supra , 159 Cal. at pages 657 to 658, 115 P. 313, but that case holds only that it is misconduct to investigate questions on which no evidence at all has been submitted. That is not the case here.

This case shares far more in common with People v. Cumpian (1991) 1 Cal.App.4th 307, 1 Cal.Rptr.2d 861, in which an issue in the case was how difficult it would have been for the defendant to remove a duffel bag he had strapped to his body. The actual duffel bag was admitted into evidence and several jurors put it on as the defendant had described and then tried to remove it. ( Id. at pp. 310–311, 1 Cal.Rptr.2d 861.) This was not misconduct: "To prohibit jurors from analyzing exhibits in light of proffered testimony would obviate any reason for sending physical evidence into the jury room in the first instance." ( Id. at p. 316, 1 Cal.Rptr.2d 861.)

"Nothing requires that the jury's deliberations be entirely verbal, and we would expect a conscientious jury to closely examine the testimony of the witnesses, no less so when that testimony takes the form of a physical act." ( People v. Cooper (1979) 95 Cal.App.3d 844, 854, 157 Cal.Rptr. 348.) When physical exhibits are admitted into evidence and supplied to the jury, they may examine and manipulate the exhibits to assess propositions placed at issue, and upon which evidence has been submitted, during the trial. (E.g., People v. Baldine (2001) 94 Cal.App.4th 773, 777–780, 114 Cal.Rptr.2d 570 [jurors could test whether police scanner, admitted into evidence, worked, after the defendant testified it did not]; People v. Bogle (1995) 41 Cal.App.4th 770, 778–781, 48 Cal.Rptr.2d 739 [permissible for jurors to test whether the defendant's keys, admitted into evidence, opened a safe, also admitted into evidence]; see People v. Singh (2012) 206 Cal.App.4th 366, 373, 141 Cal.Rptr.3d 714 ["Manipulation of an exhibit in evidence does not constitute receipt of new evidence" and is "a legitimate part of deliberations"].)

When jurors tried to rock Peterson's boat to assess its stability, they did no more than manipulate a physical exhibit admitted into evidence at trial. Their movements did not result in the impermissible receipt of extrinsic evidence. Moreover, and in any event, the court's cautionary instruction helpfully ensured jurors would consider material differences between the setting in which they were permitted to examine the physical evidence — on land, on a trailer — and the setting in which the boat's stability was at issue — on the San Francisco Bay, under the weather conditions of December 24, 2002.

Because there was no misconduct, it was not an abuse of discretion to deny Peterson's motion for a mistrial, nor were Peterson's rights to trial by an impartial jury infringed. G. Other Juror Misconduct Issues

1. Dismissal of Juror No. 5

During the guilt phase trial, the court excused a juror for discussing the case with others, contrary to the court's admonition. Peterson contends this was error or, in the alternative, that the court should have excused other jurors as well, and that the decision to excuse one juror and not others violated his statutory and constitutional rights to trial by an impartial jury. ( U.S. Const., 5th & 6th Amends.; Pen. Code, § 1089.) The trial court did not abuse its discretion, and Peterson was not denied an impartial jury.

a. Background

Three weeks into trial, the court's bailiff received information from multiple jurors that Juror No. 5 was discussing the evidence with others over the objections of jurors who had asked him to stop and in contravention of the court's daily admonitions to the jury to not discuss the case until after closing argument. To investigate the allegations, the court questioned every member of the jury individually, in chambers and under oath, beginning with Juror No. 5.

Specific written allegations from one juror included claims that Juror No. 5 had commented on the homemade boat anchor introduced as a prosecution exhibit, testimony from prosecution witness Detective Brocchini, Laci's pregnancy weight gain, the sufficiency of reports prepared by the Modesto Police Department, deficiencies in the prosecution's presentation, and how the juror was being portrayed in the media. When called in to court to testify, Juror No. 5 initially denied the allegations were true but then agreed that "general conversations" about some of these topics might have occurred. Describing the anchor conversation, he said that a juror had wondered about the anchor's weight, and in response he had discussed his own experience using anchors when fishing. Juror No. 5 denied commenting on Detective Brocchini's testimony, the prosecution's presentation, or Laci's weight but said comments about her weight had been made by others. His response to being called a "loose cannon" and other names by commentators on a cable network covering the trial, Court TV, was "keep them coming."²² It was other jurors, not Juror No. 5, who had discussed the need for detail and accuracy in official filed reports.

22 Media coverage of the Peterson trial was sufficiently intense that perceived interactions (both verbal and nonverbal) between various jurors and members of the defense and prosecution as they entered and exited the courtroom became the subject of ongoing comment in the press and on television.

Juror No. 8, who had submitted the confidential letter to the court, was sworn and supplemented its contents. He affirmed that Juror No. 5 spoke constantly about the facts and issues in the case, even after the court's admonishment that the jurors not do so. He reported that the previous day, Juror No. 5 had discussed the anchor introduced into evidence and had said it could not anchor a boat as big as Peterson's because of the strength of the currents in the San Francisco Bay. Juror No. 5 had said that in his opinion Detective Brocchini's testimony raised many questions. After evidence was introduced that Laci's weight had increased from 126 pounds to 153 pounds during pregnancy, Juror No. 5 had opined that she might have been more than eight months pregnant. He had said, based on his work experience filing reports, that the Modesto Police Department should have done a better job. Juror No. 5 had commented on "[m]ore than one occasion" that the prosecution's presentation left "a little to be desired." Finally, Juror No. 5 had never said he personally watched Court TV, but when his girlfriend relayed that he was being described by them as "a loose cannon," he had said, "Well, I sort of pride myself on that." Juror No. 8 had twice personally confronted Juror No. 5 to ask him to stop but had given up because it hadn't worked, despite Juror No. 5's statements that "[i]f anyone has a problem with this, they should be man enough to come up to [me]."

Every juror was questioned. Some jurors had not heard any of what Juror No. 8 reported, but others confirmed either comments Juror No. 5 had made or more generally that these topics had been discussed. Alternate Juror No. 2 admitted having asked about the anchor's weight and being assured it could be examined during deliberations. Jurors No. 7 and No. 9 confirmed a conversation about fishing and anchors had taken place, although they could not be certain of the source. Juror No. 4 affirmed that there had been discussion concerning the weight of the anchor, though he too was unsure of the source. Other jurors were able to specifically attribute comments to Juror No. 5: Juror No. 6 said Juror No. 5 had discussed how currents can pull a boat's anchor, and Alternate Juror No. 6 heard Juror No. 5 say the anchor in the case was smaller than anticipated, too small to anchor a boat like Peterson's.

Concerning Detective Brocchini, Juror No. 6 said someone had commented that Detective Brocchini was "[g]etting a reaming" on cross-examination, and Juror No. 4 said Juror No. 5 had asked him if he got anything out of Detective Brocchini's testimony. Juror No. 6 said Juror No. 5 had made remarks about the Modesto Police Department, though he could not recall the content, and had said the prosecution did not seem organized. Juror No. 3 likewise reported that Juror No. 5 had commented on the prosecutors' presentation. Alternate Jurors No. 5 and No. 6 heard discussions of the prosecution's and defense's presentations, though they could not say whether Juror No. 5 had participated. Juror No. 2 and others told Juror No. 5 he should not be discussing the case, while Alternate Juror No. 3 overheard another juror remind Juror No. 5 they shouldn't discuss the case after a comment he made. Lastly, most jurors had heard Juror No. 5 discuss his portrayal as a loose cannon in the media and how he was being negatively portrayed on Court TV.

In light of this testimony, defense counsel urged that Juror No. 5 be retained but the entire panel given a stern lecture. In the alternative, if Juror No. 5 were to be excused, others who had talked with him about the case should be excused too. The prosecution argued that Juror No. 5 had repeatedly ignored the court's instructions not to discuss the case and should be removed.

Over defense objection, the court discharged Juror No. 5. It concluded that Juror No. 8 was more credible than Juror No. 5, that Juror No. 5 had disregarded the instruction not to discuss the facts of the case, and that he likely would continue to do so. In the court's opinion, based on the testimony received, Juror No. 5 was "a total cancer [on] this jury" who could not be allowed to remain.

b. Discussion

Under Penal Code section 1089, "[i]f at any time, whether before or after the final submission of the case to the jury, a juror dies or becomes ill, or upon other good cause shown to the court is found to be unable to perform his or her duty, ... the court may order the juror to be discharged and draw the name of an alternate" to replace the discharged juror. A failure to follow the court's instructions is misconduct and a basis for dismissal. ( People v. Williams (2015) 61 Cal.4th 1244, 1262, 192 Cal.Rptr.3d 266, 355 P.3d 444 ; People v. Linton (2013) 56 Cal.4th 1146, 1194, 158 Cal.Rptr.3d 521, 302 P.3d 927.) This extends to the obligation not to discuss a case prematurely. Courts are required to instruct jurors not to discuss any aspect of a case amongst themselves before beginning deliberations ( Pen. Code, § 1122 ),²³ and the court followed that directive here, instructing the jury before opening statements and at each adjournment to refrain from discussing the case. (See CALJIC No. 0.50 ; CALCRIM No. 101.²⁴ ) "A juror's violation of these directions constitutes serious misconduct." ( Williams , at p. 1262, 192 Cal.Rptr.3d 266, 355 P.3d 444 ; see People v. Sandoval (2015) 62 Cal.4th 394, 437, 196 Cal.Rptr.3d 424, 363 P.3d 41 ; People v. Weatherton (2014) 59 Cal.4th 589, 599 & fn. 10, 174 Cal.Rptr.3d 45, 328 P.3d 38 ; People v. Ledesma (2006) 39 Cal.4th 641, 743, 47 Cal.Rptr.3d 326, 140 P.3d 657 ; People v. Daniels (1991) 52 Cal.3d 815, 863–866, 277 Cal.Rptr. 122, 802 P.2d 906.)

23 Under Penal Code section 1122, subdivision (a)(1), a court must admonish the jury before opening statements not to "converse among themselves, or with anyone else ... on any subject connected with the trial." Under subdivision (b), the court must repeat this admonishment at every adjournment until the case is submitted to the jury.

24 After the jurors were sworn, they were instructed, in the language of CALJIC No. 0.50, that "[y]ou must not converse among yourselves or with anyone else on any subject connected with this trial" except when deliberating.

" ‘The ... ultimate decision whether to retain or discharge a juror ... rests within the sound discretion of the trial court.’ " ( People v. Sattiewhite (2014) 59 Cal.4th 446, 486, 174 Cal.Rptr.3d 1, 328 P.3d 1 ; see People v. Williams, supra , 61 Cal.4th at p. 1262, 192 Cal.Rptr.3d 266, 355 P.3d 444.) " ‘In determining whether juror misconduct occurred, "[w]e accept the trial court's credibility determinations and findings on questions of historical fact if supported by substantial evidence." ’ " ( People v. Linton, supra , 56 Cal.4th at p. 1194, 158 Cal.Rptr.3d 521, 302 P.3d 927 ; see People v. Nesler (1997) 16 Cal.4th 561, 582, 66 Cal.Rptr.2d 454, 941 P.2d 87.) We will uphold the trial court's decision if the record supports the basis for that decision as a " ‘demonstrable reality.’ " ( Williams , at p. 1262, 192 Cal.Rptr.3d 266, 355 P.3d 444.) This means simply that the record must reveal the reason for the court's decision to discharge a juror and in turn substantial evidence must support that reason. ( People v. Duff (2014) 58 Cal.4th 527, 560, 167 Cal.Rptr.3d 615, 317 P.3d 1148.) So long as it does, " ‘the court's action will be upheld on appeal.’ " ( Sattiewhite , at p. 486, 174 Cal.Rptr.3d 1, 328 P.3d 1.)

Here, Juror No. 8 testified that Juror No. 5 violated the directive not to discuss any aspect of the case on multiple occasions and in connection with multiple topics. Other jurors corroborated aspects of that testimony. Although Juror No. 5 largely denied discussing the case with others, the court credited Juror No. 8's testimony over that of Juror No. 5. The court concluded, after hearing from every juror, that Juror No. 5 had violated the instruction not to discuss the case and could not be trusted to refrain from doing so in the future, and on that basis discharged Juror No. 5. Substantial evidence — specifically, the testimony of the many other jurors who heard Juror No. 5 discuss aspects of the case — supports that conclusion.

Peterson acknowledges Juror No. 5 engaged in misconduct by disregarding the court's admonition not to discuss the case. He nevertheless urges that the court's decision to excuse the juror was an abuse of discretion. We disagree.

Peterson's argument depends on crediting Juror No. 5's own report minimizing the significance of his actions. It disregards both the contrary testimony of Juror No. 8 and the portions of other jurors' statements that corroborated that testimony. We are not free to do the same. The trial court made an express credibility finding, siding with Juror No. 8's version of events over Juror No. 5's version: "I have the testimony of Juror Number 8, and I'm more inclined to believe Juror Number 8 than I am to believe Juror Number 5." That finding rested in part on the court's observation of these jurors as they testified: "I'm satisfied by watching [Juror Number 5's] demeanor, and watching the demeanor of Juror Number 8 and some of these other [jurors]." The court was able to observe matters not evident from the cold appellate record — and what the record does show is generally supportive of the court's conclusion. Here, for example, the prosecutor commented that when Juror No. 5 was confronted with the accusations, the hesitation he gave before issuing a denial was "the longest pause I've ever seen." The court did not disagree. Such considerations may play a central role in evaluating credibility and underlie the requirement that appellate courts defer to such assessments when they find any support in the record. (See People v. Williams, supra , 61 Cal.4th at p. 1262, 192 Cal.Rptr.3d 266, 355 P.3d 444 ["We defer to the trial court's credibility assessments ‘based, as they are, on firsthand observations unavailable to us on appeal’ "].) We accept the court's determination that Juror No. 5 was being less than fully truthful, a determination the court could rely on in deciding to excuse the juror. (See id. at pp. 1261–1263, 192 Cal.Rptr.3d 266, 355 P.3d 444.) People v. Wilson (2008) 44 Cal.4th 758, 80 Cal.Rptr.3d 211, 187 P.3d 1041, on which Peterson heavily relies, is distinguishable. There, a juror on one occasion made "solitary and fleeting comments" to another juror. ( Id. at p. 839, 80 Cal.Rptr.3d 211, 187 P.3d 1041.) We determined this violation of the court's admonition "was a trivial one: one, possibly two sentences, spoken in rhetorical fashion and not in an obvious attempt to persuade anyone." ( Id. at pp. 839–840, 80 Cal.Rptr.3d 211, 187 P.3d 1041.) Only one other juror even heard the remark, and did not respond. Nor did the substance of the remarks — " ‘this is what happens when you have no authority figure’ " ( id. at p. 836, 80 Cal.Rptr.3d 211, 187 P.3d 1041 ) — suggest prejudgment of the appropriate penalty, as the trial court had found ( id. at pp. 840–841, 80 Cal.Rptr.3d 211, 187 P.3d 1041 ). Here, in contrast, the court received testimony under oath that in the first three weeks of trial, Juror No. 5 had discussed aloud the effectiveness of the prosecution's presentation of evidence, the cross-examination of one of its witnesses, inferences to be drawn from Laci's weight gain, the usefulness of a boat anchor admitted into evidence, and other case-related topics.

More analogous is People v. Williams, supra , 61 Cal.4th 1244, 192 Cal.Rptr.3d 266, 355 P.3d 444, where a juror was overheard offering her opinion as to which witnesses were telling the truth. Although only one other juror recalled the statement, the court could credit that juror, disbelieve the first juror's denial, and discharge her based on her willingness to prejudge matters, discuss the prejudgment aloud, and conceal her misconduct. ( Id. at pp. 1260–1263, 192 Cal.Rptr.3d 266, 355 P.3d 444.) Here, as there, the court could credit Juror No. 8's testimony over that of Juror No. 5 and conclude Juror No. 5 could not be trusted going forward.

In the alternative, Peterson argues that if it was permissible for the court to discharge Juror No. 5, then it was an abuse of discretion not to simultaneously excuse other jurors Juror No. 5 talked to, including Jurors No. 4 and No. 6 and Alternate Jurors No. 2 and No. 6. But Juror No. 5 was different from these others in at least two material respects. First, the court heard testimony from Juror No. 8, whom it credited, that Juror No. 5 was "the leader of the clique" of jurors who talked about the case and the one who "usually starts the conversation." In other words, Juror No. 5 was not simply a participant, but an instigator. Second, Juror No. 5 was alone among the jurors in denying participation in conversations in a way the court found less than credible. Based on these considerations, the trial court concluded "this guy is not following the Court's admonitions" and, going forward, "[h]e's not about to follow the Court's admonitions." The court viewed Juror No. 5, specifically, as "a cancer in that jury room" whom it could not trust to follow the court's instructions and someone who, if left on the jury, seemed likely to supply grounds for a new trial motion down the road. Where, as here, the record supplies evidence that a juror cannot be trusted to follow the court's instructions going forward, the court may discharge the juror. (See People v. Williams, supra , 61 Cal.4th at pp. 1262–1263, 192 Cal.Rptr.3d 266, 355 P.3d 444 ; People v. Daniels, supra , 52 Cal.3d at p. 865, 277 Cal.Rptr. 122, 802 P.2d 906 ["a judge may reasonably conclude that a juror who has violated instructions to refrain from discussing the case ... cannot be counted on to follow instructions in the future"].)

Peterson's brief, undeveloped claims of federal constitutional error arising from the dismissal of Juror No. 5 rest on the principle, established in Witherspoon, supra , 391 U.S. 510, 88 S.Ct. 1770 and Adams v. Texas, supra , 448 U.S. 38, 100 S.Ct. 2521, that the right to trial by an impartial jury may be compromised when a state selectively culls jurors able to consider the facts and faithfully apply the law. But Adams and Witherspoon both involved the selective removal of prospective jurors on an entirely different basis — namely, their generally unfavorable views of the death penalty. Neither those cases nor any other authorities establish the proposition that the constitutional right to a jury trial constrains a court from removing sitting jurors who fail or refuse to follow the court's instructions.

2. Failure to Adequately Investigate Allegations of Misconduct by Juror No. 8

Before the start of the penalty phase trial, the court received information that Juror No. 8 had discussed the case with others and the jury had predetermined Peterson should be sentenced to die. The court held an evidentiary hearing, concluded the allegations were unfounded, and denied Peterson's motion for a mistrial. Peterson now argues that the hearing preceding that ruling was inadequate. Because Peterson made no objection, the argument is forfeited. On the merits, Peterson has not shown an abuse of discretion.

a. Background

After the conclusion of the guilt phase trial and before the beginning of the penalty phase trial, the San Mateo District Attorney's Office received a call from a local attorney concerning Juror No. 8. The tipster relayed conversations she had had with a neighbor, Gino Gonzalez, who was a bartender. Gonzalez reportedly said Juror No. 8 frequented his bar, and Gonzalez had learned that the jury kept secret notebooks and had already decided to impose the death penalty. The attorney acknowledged that her report involved " ‘multiple [levels of] hearsay.’ "

At the court's request, the investigator who received the call followed up directly with Gonzalez. Gonzalez said the attorney's report was " ‘ridiculous and not true in any sense.’ " Gonzalez knew Juror No. 8 and that he was on a jury in an unspecified high-profile case, but had never heard Juror No. 8 discuss which case it was or anything about the case. The court held a hearing and had the tipster attorney testify. She described the circumstances and affirmed the content of her conversations with Gonzalez but acknowledged that she did not know whether Gonzalez had gotten his information directly from Juror No. 8 or from Juror No. 8's girlfriend, who worked with Gonzalez. She added that Gonzalez reported serving beer to Juror No. 8 in the morning, after the juror came in following a night shift and before he went to court.

Gonzalez appeared through a lawyer, who represented that Gonzalez would invoke the Fifth Amendment if called to testify unless granted immunity. Gonzalez's attorney also represented that, if questions were limited to whether Gonzalez had spoken with Juror No. 8, he would answer, and would say, consistent with the investigator's report, that Juror No. 8 had never revealed any more than that he was a juror in a high-profile case. The court elected not to call Gonzalez, and instead to question each juror and alternate concerning whether they had discussed or predetermined their penalty verdict. Peterson did not object.

When called, every juror denied discussing the penalty to be imposed or reaching a premature decision on the question. Juror No. 8 additionally denied drinking before coming to court, conversing with Gonzalez or anyone else at his bar about the case, or saying to anyone that the jury kept secret notebooks.

Based on this testimony, the court concluded the jury had not predetermined the penalty verdict and denied Peterson's motion for a mistrial.

b. Discussion

"A trial court learning of grounds for dismissal [of a juror] ‘has an affirmative obligation to investigate.’ [Citation.] However, ‘[b]oth the scope of any investigation and the ultimate decision whether to discharge a given juror are committed to the sound discretion of the trial court.’ " ( People v. Duff, supra , 58 Cal.4th at p. 560, 167 Cal.Rptr.3d 615, 317 P.3d 1148 ; see People v. Manibusan (2013) 58 Cal.4th 40, 53–54, 165 Cal.Rptr.3d 1, 314 P.3d 1.) Hearsay evidence of "alleged jury misconduct ordinarily is insufficient to establish an abuse of discretion in either denying the motion [for a mistrial based on misconduct] or declining to conduct an evidentiary hearing." ( Manibusan , at p. 55, 165 Cal.Rptr.3d 1, 314 P.3d 1.) Here, the court had before it only the tipster attorney's testimony about hearsay statements by Gonzalez to her and Gonzalez's hearsay denial of those statements. It could have called Gonzalez to testify but chose not to given Gonzalez's insistence on invoking the Fifth Amendment absent a grant of immunity from prosecution. Instead, the court elected to go directly to the jury and ask each juror whether discussions or prejudgment of the penalty had occurred. If Peterson disagreed with this course of action, it was incumbent on him to object at the time and give the court the opportunity to correct any perceived error.

Peterson did not. Quite to the contrary, defense counsel asserted that, if granted immunity and called to testify, Gonzalez would surely just say any statements attributed to him by the attorney were untrue. Counsel thus backed down from an earlier request that the parties be sent to the presiding judge and afforded an opportunity to seek immunity for Gonzalez, apparently concluding that such proceedings would be unhelpful in light of Gonzalez's anticipated testimony. Indeed, when the court obtained from Gonzalez's counsel an agreement to appear and answer questions limited to whether Gonzalez and Juror No. 8 had ever spoken about the case and the content of those discussions, defense counsel talked the court out of this approach. He argued that if Gonzalez were called, Peterson would be entitled to cross-examine Gonzalez more broadly, and thus the court could not call Gonzalez and agree to limit any questioning. Instead, counsel proposed that the court bring in each juror for questioning, the precise approach the court adopted. Peterson's argument that the evidentiary hearing was unduly limited because Gonzalez was not called is forfeited. (E.g., People v. Huggins (2006) 38 Cal.4th 175, 238, 41 Cal.Rptr.3d 593, 131 P.3d 995.)

The contention is also without merit. The court heard from every juror and alternate, each of whom denied discussion or prejudgment of the penalty phase verdict had occurred. The court could credit that testimony. It had before it a representation from Gonzalez's counsel that Gonzalez would deny ever having discussed the case with Juror No. 8 and would otherwise invoke the Fifth Amendment, a representation that defense counsel expressly agreed was surely true. In these circumstances, there was no abuse of discretion in failing to require Gonzalez to go through putting these statements on the record.

H. Penalty Phase Issues

Peterson raises procedural and evidentiary challenges to the conduct of the penalty phase trial and contends California's death penalty scheme is unconstitutional. Because we reverse the penalty verdict based on errors in jury selection, we need not address these claims. (See People v. Armstrong, supra , 6 Cal.5th at p. 800, 243 Cal.Rptr.3d 105, 433 P.3d 987.) III. DISPOSITION

We affirm the judgment as to guilt, reverse the judgment as to the sentence of death, and remand the matter for a new penalty determination.

We Concur:

CANTIL-SAKAUYE, C.J.

CHIN, J.

CORRIGAN, J.

LIU, J.

CUÉLLAR, J.

GROBAN, J.