Posner, Chief Judge.

The plaintiffs in this legal malpractice suit appeal from its dismissal on the defendants' motion for summary judgment, raising a novel issue concerning the law of legal malpractice. The issue, which arises when as in this case the plaintiff is complaining that his lawyer booted a procedural entitlement, such as the right to a jury trial, is whether the plaintiff must show that his lawyer's negligence not only caused him to lose but brought about an unjust result — the wrong party won. The plaintiffs are the Jones Motor Company, a trucker, and its insurer. The defendants are lawyers who represented Jones in a personal injury lawsuit brought against it by Elston Cannon. Federal jurisdiction is based on diversity of citizenship; and the applicable law, the parties agree, is Illinois's common law of malpractice.

The underlying suit had been filed in a state court in St. Clair County and assigned to a judge who we are told, and accept for purposes of deciding this appeal, has the reputation of favoring plaintiffs in personal injury suits. Jones's lawyers negligently failed to make a timely effective request for a jury because they failed to accompany the request with payment of the fee for a jury trial. As a result the case was tried to the judge, who entered a judgment of $2.8 million for the plaintiff; the suit was then settled for $2.5 million. In the present case, the malpractice case, Jones tendered the opinion of an experienced lawyer in St. Clair County that had the case been tried to a jury, the verdict would have been in the neighborhood of $500,000. Jones and its insurer, which paid a part of the $2.5 million settlement, are suing for the $2 million difference.

There is some underbrush to clear out of the way before we get to the principal issue. To begin with there is the defendants' "contingent cross-appeal," which challenges the district judge's ruling that the insurer has standing to bring a malpractice suit even though it was an excess rather than primary insurer and had no duty to defend Jones, only to indemnify it if Jones's liability exceeded the excess threshold, as it did. The cross-appeal (which is contingent because the ruling is of no consequence if we affirm the judgment for the defendants) is improper, because it does not seek an alteration of the judgment. Massachusetts Mutual Life Ins. Co. v. Ludwig, 426 U.S. 479 (1976) (per curiam); Singletary v. Continental Illinois National Bank Trust Co., 9 F.3d 1236, 1240 (7th Cir. 1993); In re Sims, 994 F.2d 210, 214 (5th Cir. 1993). The defendants want us to affirm, not modify or reverse, the judgment, which dismissed the suit with prejudice. They want to advance an alternative ground of affirmance with regard to the insurer that was rejected by an interlocutory order by the district court, but since they do not want to disturb the judgment they can advance their alternative ground, and in doing so attack the order that rejected it, simply by arguing the ground in their appellees' brief.

The cross-appeal must thus be dismissed irrespective of the failure of the contingency to materialize, but we can still consider the alternative ground for affirmance, as it has been fully briefed. The insurer had no contract with the defendants (Jones's lawyers in the suit by Cannon), as it might have had if it had had a duty to defend its insured and in fulfillment of that duty had hired the defendants to handle the defense of Cannon's suit. Therefore, the defendants argue, the insurer cannot complain that they violated a duty to it. This amounts to saying that Illinois requires privity of contract in legal malpractice cases — and it does, unless the plaintiff was an intended beneficiary of the lawyer's contract, Pelham v. Griesheimer, 440 N.E.2d 96, 99-100 (Ill. 1982); Jewish Hospital v. Boatmen's National Bank, 633 N.E.2d 1267, 1275 (Ill.App. 1994), which the insurer plaintiff was not. But this argument does not touch a suit by an insurer not claiming a breach of duty to itself but instead suing as a subrogee. Jones settled with Cannon, and the insurance company picked up a part of the tab and wants to get it back. The insurance company was in effect an assignee, and an assignee is in privity with the other party to its assignor's contract; and while it is true that Illinois refuses to permit the assignment of malpractice claims, its refusal is bottomed on grounds — having to do with fear of the "merchandising" and "commercialization" of the lawyer-client relationship, Brocato v. Prairie State Farmers Ins. Ass'n, 520 N.E.2d 1200 (Ill.App. 1988) — that have little if any applicability to the subrogation of a claim of legal malpractice, since the relation that gives rise to subrogation preexists the malpractice and the malpractice claim. But as the cases are divided on the propriety of allowing a subrogee to sue on such a claim, compare, e.g., American Centennial Ins. Co. v. Canal Ins. Co., 843 S.W.2d 480, 485 (Tex. 1992) (pointing out that the insured client may have little incentive to sue and so lawyer negligence will go unpunished), with American Continental Ins. Co. v. Weber Rose, P.S.C., 997 S.W.2d 12 (Ky.App. 1998) (expressing concern that the lawyer's loyalty may be divided between insured and insurer), and the Illinois courts have not spoken to the issue and its resolution is inessential to the decision of this appeal, we shall forgo ruling on it.

The plaintiffs complain not only about the defendants' failure to obtain a jury trial for them in Cannon's case but also of another procedural bobble, as a result of which Jones's expert witness was not permitted to testify. This complaint is so little developed in the plaintiff's opening brief in this court that it must be deemed waived. E.g., JTC Petroleum Co. v. Piasa Motor Fuels, Inc., 190 F.3d 775, 780-81 (7th Cir. 1999); Kerr v. Farrey, 95 F.3d 472, 481 (7th Cir. 1996); Karibian v. Columbia University, 14 F.3d 773, 777 n. 1 (2d Cir. 1994). In any event the judge admitted the expert's reports into evidence, and it is unclear what if anything the expert's live testimony would have added. Remember that this was a bench trial; oral testimony by an expert witness is less essential to enlightening a judge, who is accustomed to reading complex documents, than it is to enlightening a jury.

We come to the most important issue, which is whether, and if so when, the loss of a procedural advantage can give rise to a malpractice suit even if the advantage was not essential to the protection of the client's substantive rights. Through the defendants' negligence Jones and its insurer lost their right to a jury trial and were forced to submit to a bench trial — which means they got a trial before an authorized tribunal. They allege no error in the conduct of the trial by the judge whom they did not want to try the case, and they did not appeal from the judgment that he rendered, large as it was. The plaintiffs thus got a fair trial and there is no basis for supposing that the judgment was excessive, albeit it may have been higher than it would have been had Jones's lawyers not thrown away their client's right to a jury trial. Some Illinois cases say or imply that you cannot get a judgment for malpractice against a lawyer unless you can show that you had a meritorious claim (or defense, when the client had been a defendant rather than a plaintiff), e.g., Lucey v. Law Offices of Pretzel Stouffer, Chartered, 703 N.E.2d 473, 476 (Ill.App. 1998); Moore v. Owens, 698 N.E.2d 707, 709 (Ill.App. 1998); Serafin v. Seith, 672 N.E.2d 302, 309-10 (Ill.App. 1996), and Jones's lawyers argue correctly that their client had no entitlement not to be mulcted by a judgment of $2.8 million.

But we think the real thrust of these cases is that a malpractice plaintiff cannot prevail merely by showing that his claim which his lawyer booted, though baseless, had some nuisance value. Campbell v. Magana, 184 Cal.App.2d 751, 753-54 (1960). Imagine a situation in which a class action is brought and is thrown out as a result of a negligent mistake by the lawyer for the class, who is then sued for tens or hundreds of millions of dollars in another class action, in which it is argued that although the suit was frivolous it is well known that frivolous class actions can sometimes extort sizeable settlements from the defendants, as we pointed out in In re Rhone-Poulenc Rorer Inc., 51 F.3d 1293, 1298 (7th Cir. 1995). To impose malpractice liability for booting a nuisance suit would — like deeming a plaintiff who obtains a nuisance settlement a prevailing party for purposes of entitlement to an award of attorneys' fees, which courts also refuse to do, see, e.g., Fletcher v. City of Fort Wayne, 162 F.3d 975, 976 (7th Cir. 1998) — simply encourage nuisance suits, of which we have enough already.

But to say that the plaintiff's loss of a nuisance suit is not a ground for the plaintiff's suing his lawyer is not the same thing as saying that the plaintiff must prove that had it not been for his lawyer's negligence he would have won the suit for sure. Take the classic case of legal malpractice in litigation — failure to file suit before the statute of limitations expires. If the suit thus aborted had only nuisance value, then, as we have just said, the lawyer's negligence would not support a malpractice suit. But if as with most suits the probability of a successful outcome was less than 100 percent, the plaintiff in the malpractice suit could not "prove" that he would have won. The outcome of the suit (had there been no malpractice) might have turned on which of two witnesses the jury would have believed, and if there were a reasonable probability that the defendant would have won the swearing contest then it could not be said that a loss of the case by the plaintiff could only have reflected injustice. Such possibilities do not defeat malpractice liability. Nicolet Instrument Corp. v. Lindquist Vennum, 34 F.3d 453, 455 (7th Cir. 1994); 2 Ronald E. Mallen Jeffrey M. Smith, Legal Malpractice § 19.3, p. 600 (4th ed. 1996).

There is a difference, however, between saying that a claim can be meritorious without its being certain to prevail at trial and saying that one of the parties would have done better than the other, had it not been for the negligence of his lawyer, regardless of the relative merits of the parties' positions. And that is (at most) this case. Although the judge who tried the case against Jones may have a reputation of being more liberal in personal-injury suits than the average jury in his county, it is impossible to infer from this that the $2.8 million judgment that he rendered against Jones was too high; the average jury verdict in such a case might be too low. Of course if this judge were prejudiced against motor carriers, or litigants named Jones, or defendants in personal injury cases, there would be a basis for inferring that the negligence of Jones's lawyer had cost Jones a shot to which he was entitled at a lower damages award, albeit an entitlement not certain to be enforced even by a jury; but of this there is no evidence.

Yet this analysis is not satisfactory either. We must ask why Illinois allows a defendant in a civil suit to elect to be tried by a jury even though the plaintiff would prefer a bench trial. The answer must be that each party is deemed entitled to seek the "protection" of the jury against being tried by a judge. "Developments in the Law: The Civil Jury," 110 Harv. L. Rev. 1408, 1429-32 (1997). That entitlement, a real legal entitlement and not just a tactical opportunity to obtain a more favorable tribunal, was worth something to Jones, and it was kicked away by the defendants' negligence. The only reason for treating it differently from other entitlements, such as the entitlement to introduce evidence or to enforce a substantive right, is practical; it is the difficulty of valuing its loss. See 3 Mallen Smith, supra, § 29.35, p. 729. The difficulty becomes impossibility in a case (which is not this case, however) in which, at the time the right to a jury trial is forfeited, the identity of the judge who will try the case in lieu of the jury is not known. There is variance among judges as well as among juries, and it is very hard to say that the average jury is likely to be more favorable to a defendant than the average judge. Hence the foreseeable loss in such a case would be extremely hard to estimate.

What is true, but not helpful to Jones, is that a defendant who has a very weak case — a case he deserves to lose — will prefer a jury trial. Gerald D. Gay et al., "Noisy Juries and the Choice of Trial Mode in a Sequential Signalling Game: Theory and Evidence," 20 RAND J. Econ. 196 (1989). The reason is that there is greater variance (implying less accuracy) in jury verdicts and therefore a greater chance that a weak case will convince a jury than that it will convince a judge. The fact that Jones wanted a jury instead of a judge would ordinarily signal a weak rather than a strong case, but the inference is countered here by the argument that even with a strong case a defendant might prefer a jury to a judge known to favor plaintiffs in personal-injury suits.

Partly because the precise issue has never arisen before, so far as the parties' research or our own discloses, either in Illinois or in any other jurisdiction, we hesitate to rule out the possibility of convincing an Illinois court to allow a malpractice suit to go forward on the basis of an argument that the plaintiff lost a procedural entitlement even though it was not an entitlement necessary to avert an unjust outcome. But given the uncertainty of harm we think the plaintiff in such a case must do more than the plaintiffs have done here to show that they can prove damages to a reasonable certainty. Some degree of speculation is permissible in computing damages, e.g., Story Parchment Co. v. Paterson Parchment Paper Co., 282 U.S. 555, 563-66 (1931); Mid-America Tablewares, Inc. v. Mogi Trading Co., 100 F.3d 1353, 1365-66 (7th Cir. 1996); Olympia Equipment Leasing Co. v. Western Union Telegraph Co., 797 F.2d 370, 383 (7th Cir. 1986); Starceski v. Westinghouse Electric Corp., 54 F.3d 1089, 1100-01 (3d Cir. 1995), because reasonable doubts as to remedy ought to be resolved against the wrongdoer; but there are limits. Although there is plenty of evidence that any personal-injury case assigned to the judge who presided at Cannon v. Jones Motor Group, Inc. would want a jury rather than this judge to determine damages, there is no credible evidence of what a jury might have awarded. The principal evidence is the opinion of the lawyer who thought Cannon's case worth to a jury in the range of $500,000, but this was offered as a bare conclusion without data of actual verdicts in St. Clair County in comparable cases from which some reasonable confidence interval, some range in which any jury verdict would be quite likely to lie, might have been computed. No reasonable trier of fact could have been allowed to award damages to Jones and its insurer on the basis of such unsubstantiated expert testimony. Zaz~ Designs v. L'Oréal, S.A., 979 F.2d 499, 505-06 (7th Cir. 1992); Schiller Schmidt, Inc. v. Nordisco Corp., 969 F.2d 410, 415-16 (7th Cir. 1992).

The plaintiffs argue that the way to compute damages in this case is simply to try the malpractice claim to a jury. That is a bad suggestion quite apart from the fact that a jury in a federal district court is not drawn from the same pool as the jury in a state court. The suggestion overlooks the fact that given the variance among juries, it would be necessary to try the malpractice claim a number of times in order to get a sense of the average performance of a jury in this case, and it is the difference between the judge's judgment and the judgment that Jones could have expected from a jury, which would be an average jury performance, that is the measure of what Jones lost as a result of its lawyers' negligence. So the suit was rightly dismissed after all.