CARCHMAN, P.J.A.D.

The narrow issue that we address on this appeal is whether the use of averaging of comparable sales by the trial judge in fixing the fair market value of the real property in issue represents an appropriate evaluation methodology and whether it fulfills a judge's fact-finding responsibility. Here, the trial judge, after excluding the high and low comparable sales presented by the expert witnesses in competing appraisals, averaged the values of the remaining comparables to arrive at a fair market value. We disapprove of the practice of averaging and conclude that it does not represent a reasoned and considered valuation technique. We reverse and remand.

The facts underlying this appeal are not in significant dispute. Plaintiffs Pansini Custom Design Associates, LLC and Roger Parkin are the owners of the property in Ocean City known as 801 Fourth Street, Lot 49, Block 303, consisting of a 100 x 130 foot (13,000 square feet) parcel improved with a single-family residence with a detached garage. The parcel was acquired on May 27, 1999, for $710,000 from Elizabeth Sheehan, who had resided on the property as her home. The property had been a former United States Life Saving Service and United States Coast Guard Life Saving Station until its retirement in 1940. The 19th century building is one of the last four lifesaving stations on the East Coast built by the United States Life Saving Services, the predecessor to the United States Coast Guard. Thereafter, additions and alterations followed as the premises was converted to a single family residence.

The property was designated as a historic structure under historic-preservation provisions of the Ocean City Zoning Ordinance. Originally located on the beach, because of changing tides and land accretion, the property is now two and one-half blocks from the beach and ocean. It is located in a developed mixed area, primarily a two-family residential zone. The property has been vacant since the sale by Sheehan to plaintiffs.

Because of the designation of the property as a historic structure, a developer must comply with the procedures set forth in Ocean City's zoning ordinance, specifically, the requirements of the historic-preservation ordinance, which controls development and requires a series of steps before a property owner can demolish a historic building. City of Ocean City Ordinance §§ 25-1800.1 to-.15.3. The procedures under the ordinance require an owner to first apply to the City's Historic Preservation Commission for a demolition permit. O.C. Ord. § 25-1800.7.1. The Commission is charged with the responsibility of reviewing zoning applications that involve demolition, relocation or changes to historic buildings. O.C. Ord. § 25-1800.8.3. If the Commission denies the application, an applicant may appeal to the zoning board for review.

If the zoning board affirms the denial, an applicant may then obtain relief by adhering to a six-month waiting period during which it must try to sell the property for "fair market value" to "any person or organization, governmental agency thereof or political subdivision or agency thereof, which gives reasonable assurance that it is willing to preserve the building, place or structure and the land pertaining thereto." O.C. Ord. § 25-1800.10.1(d). In an earlier opinion, we interpreted that provision to mean fair market value of the property in its present use as a historic site. Pansini Custom Design Assocs. v. City of Ocean City, No. A-5635-99T5, 2002 WL 549413 (App.Div. April 9, 2002) (slip op. at 6), certif. denied, 174 N.J. 189, 803 A.2d 1161 (2002). If at the end of the waiting period the property remains unsold, the applicant may then develop the property "as a matter of right." O.C. Ord. § 25-1800.10.1.

These procedures have been followed, as before plaintiffs purchased the site, Sheehan applied to the Commission to demolish the life saving station and build three duplex units. That application was denied. The Board of Adjustment reversed and granted the application, and the Law Division sustained the Board's decision. We reversed. See Citizens for Historic Preservation, Inc. v. Sheehan, Docket Nos. A-5106-03T2 and A-5845-03T2 (App.Div. July 26, 2005) (slip op. at 2), certif. denied, 185 N.J. 389, 886 A.2d 660 (2005). During that dispute, Pansini purchased the property from Sheehan and immediately began advertising its sale under the historic-preservation ordinance's notice provisions. Citizens for Historic Preservation, Inc. v. Sheehan, supra, (slip op. at 9).

Pansini advertised the property with a real estate broker for $1.1 million, which it deemed an appropriate price for a three-lot subdivision that could be developed with three duplexes. Defendant City of Ocean City disputed the price, claiming Pansini was required to advertise it for fair market value as a single-family home in a historic district. As noted earlier, we agreed with Ocean City, and in so doing, affirmed a trial court judgment concluding that Pansini did not comply with Ocean City's historic preservation ordinance. Pansini Custom Design Assocs. v. City of Ocean City, supra, (slip op. at 5).

Plaintiffs then filed this declaratory judgment action in the Law Division and sought to establish fair market value so that it could advertise the property for sale consistent with the City's ordinance. The trial judge held a two-day trial and heard three witnesses: an appraiser hired by Pansini, an appraiser hired by Ocean City and an appraiser hired by defendant Saving Our Station Coalition (SOS), objectors to plaintiffs proposed development.

In his decision, the judge criticized the appraisal methods used by all of the parties. He found that Pansini's appraiser improperly considered potential development of the site's excess property beyond its current use when evaluating fair market value and adjusting comparative sales. He also criticized defendants' two appraisers, noting that the comparable sales they chose occurred during a time when the real estate market in Ocean City was "flat at best."

Despite these criticisms of the comparable sales and the appraiser's adjustments, the judge averaged them to establish fair market value. More specifically, he averaged the three highest comparable sales used by defendants' appraisers and the three lowest comparable sales used by Pansini's appraiser. Using this methodology, he declared the fair market value to be $1,072,500. This appeal by SOS followed.

As we noted, the judge considered three appraisal reports. First, plaintiffs' appraiser, Michael P. Hedden, offered five comparables. While the actual sales prices of the comparables ranged from $507,500 to $995,500, Hedden adjusted the sales upward in various amounts, including an adjustment of $843,744 for a property that sold for $507,500. Converting the adjustments to percentages, Hedden adjusted the five properties respectively by, 25%, 40%, 72%, 73% and 166%. Based on the adjusted comparable sales, Hedden opined that the fair market value of the subject property was $1,400,000.

Robert M. Sapio, Ocean City's appraiser, opined that the property's fair market value was $900,000. He reached this conclusion by utilizing four comparable sales that were priced in a range of $675,000 to $771,600 and were respectively adjusted by 17% in two instances and by 20% and 31% as to the other two sales.

Finally, SOS produced Michael J. Lange as its appraiser. Lange used three comparables, ranging in sales price from $950,000 to $1,100,000. He adjusted those sales downward by 10%, 15% and 20% and concluded that the fair market value was $850,000. Unlike the two prior appraisers, Lange did not use sales that were in the historic zone, concluding that those available sales were too old. He instead utilized sales closer to the beach and adjusted accordingly.

As we have previously noted, ultimately, the judge, in reaching his valuation, utilized the three highest comparable sales used by Sapio and the three lowest sales used by Hedden, averaged them and declared the result to be the fair market value. A fact-finder's need for expert witnesses is well-defined. Expert testimony is required "where the fact[-]finder is not expected to have sufficient knowledge or experience and would have to speculate without the aid of expert testimony." Torres v. Schripps, Inc., 342 N.J.Super. 419, 430, 776 A.2d 915 (App.Div. 2001) (citing Kelly v. Berlin, 300 N.J.Super. 256, 268, 692 A.2d 552 (App.Div. 1997)). Expert testimony is generally required to determine the fair market value of real property, Jacobitti v. Jacobitti, 263 N.J.Super. 608, 613, 623 A.2d 794 (App.Div. 1993), aff'd, 135 N.J. 571, 641 A.2d 535 (1994), but the "fact[-]finder is not bound to accept the testimony of an expert witness," Torres, supra, 342 N.J.Super. at 431, 776 A.2d 915 (citing Johnson v. American Homestead Mortgage Corp., 306 N.J.Super. 429, 438, 703 A.2d 984 (App.Div. 1997)), and "may accept some of the expert's testimony and reject the rest." Id. at 430-51, 703 A.2d 984 (citing Todd v. Sheridan, 268 N.J.Super. 387, 401, 633 A.2d 1009 (App.Div. 1993)).

Where the judge has questioned or rejected the testimony of the expert, the judge, among other things, may appoint an independent expert. Id. at 436, 776 A.2d 915 (quoting Lavene v. Lavene, 148 N.J.Super. 267, 278, 372 A.2d 629 (App.Div.), certif. denied, 75 N.J. 28, 379 A.2d 259 (1977)). See also Alk Assocs., Inc. v. Multimodal Applied Systems, Inc., 276 N.J.Super. 310, 318, 647 A.2d 1359 (App.Div. 1994); Borodinsky v. Borodinsky, 162 N.J.Super. 437, 444, 393 A.2d 583 (App.Div. 1978); N.J.R.E. 614.

Ultimately, the fact-finder, here the judge, must weigh and evaluate the experts' opinions, including their credibility, to fulfill the judge's responsibility in reaching a reasoned, just and factually supported conclusion. In our view, averaging cedes this unique responsibility to a simple mathematical formula and is an unacceptable methodology for fulfilling one's role as a fact-finder.

While there is limited authority for the proposition that averaging is an inappropriate appraisal technique, the majority of reasoned decisions addressing the subject support this view.

In Wedgewood Knolls Condominium Ass'n v. West Paterson Borough, 11 N.J.Tax 514 (Tax 1991), the Tax Court considered an appraisal that averaged sale prices. In rejecting the appraisal technique, Judge Crabtree offered cogent policy considerations that are applicable here as well. The judge said:

The expert's methodology [of averaging] is without merit. To begin with, absent countervailing circumstances, the selling price of real property involved in a tax valuation proceeding is cogent evidence of its value. Hackensack Water Co. v. Div. of Tax Appeals, 2 N.J. 157, 65 A.2d 828 (1949); Niktan Realty Co. v. Passaic City, 1 N.J. Tax 393 (Tax Ct. 1980). The best evidence of the true value of each condominium unit involved in these appeals is the price for which it was sold, adjusted for time when appropriate. Second, real property must be valued for tax purposes as at October 1 of the pretax year. N.J.S.A. 54:4-23. The expert's method of valuation ignores that statutory requirement and fails to consider the effect of appreciation, as reflected in price changes, over the multi-year averaging period. Third, the expert's approach presupposes a fungibility within each category of unit model which has not been shown to exist. In addressing the use of sale price averaging in the related context of comparable sales, two noted authorities have observed:

Unless there are many adjusted sales prices and the appraiser can accept the assumption that each comparable sale has equal significance or validity (and therefore should be given weight), an arithmetic average should not be used. Certainly, averaging for three to four observations is extremely hazardous. Greatest emphasis is placed on the adjusted sales price figures for the comparable properties that are regarded as most nearly similar to the subject property. One guide is the amount of adjustment required (provided that they are reasonable in size). . . .

[Boyce and Kinnard, Appraising Real Property (1984) at 198.]

[ Id. at 522-23. See also Bloomfield Assocs. v. Town of Bloomfield, 12 N.J. Tax 501, 511 (Tax 1992).]

In Watnong Assocs., Inc. v. Twp. of Morris, 11 N.J. Tax 108, 117 n. 7 (Tax 1990), aff'd, 12 N.J. Tax 252 (App.Div. 1991), Judge Andrew referred to averaging as "a dubious technique" and added:

To treat all vacant land as fungible and to seek to reduce the disparity in prices by the elimination of range extremes does not constitute a reliable approach to the fair market value of the subject. What is required is an examination of the sales of comparable properties and the utilization of sound judgment with respect to the relevant value factors such as time, location, physical characteristics, utility and desirability.

Louisiana has adopted a similar view. In Greater Baton Rouge Consol Sewer Dist. v. Nelson, 144 So. 2d 186, 189 (La.Ct.App. 1962), the Louisiana Court of Appeals observed:

For obvious reasons any tendency to average appraisals solely as a matter of convenience and expedience to the courts must be rejected as utterly unwarranted and unjustified. To permit the averaging of appraisals which are not equally sound and logical will only lead to gross inequities and is, therefore, a method of solution to be discouraged and condemned at every opportunity. Adoption of such a procedure, as time saving as it may be to the courts, is no substitute for the obligation of the courts to dispense justice for justice's sake. Such a practice could only lead to the absurd result wherein ridiculous extremes would be resorted to by the contending parties to insure themselves of such protection as would thus be available under such a system:

[(Emphasis added).]

See also Mullins v. Page, 457 So.2d 64, 74 (La.Ct.App.), cert. denied 459 So.2d 538 (La. 1984); State ex rel Dept. of Highways v. Munson, 174 So.2d 923, 924 (La.Ct.App.), cert. denied 247 La. 1091, 176 So.2d 146 (1965).

The Pennsylvania courts have permitted averaging, see Williamson v. Williamson, 402 Pa.Super. 276, 586 A.2d 967, 974 (1991), Aletto v. Aletto, 371 Pa.Super. 230, 537 A.2d 1383, 1389 (1988), but in neither case did the court provide a reasoned explanation for permitting averaging. Similarly in Indiana, the court concluded, without reasons, that averaging appraisals was not an abuse of discretion and found that all of the real estate values arrived at by the trial judge were within the scope of evidence. Webb v. Schleutker, 891 N.E.2d 1144, 1152 (Ind.Ct.App. 2008).

One federal Bankruptcy Court commented that averaging is inappropriate if the form of the findings make it difficult to discern the court's basis for its conclusions. First Brandon Nat'l Bank v. Kerwin-White, 109 B.R. 626 (D.Vt. 1990) (citing Kelley v. Everglades Drainage Dist., 319 U.S. 415, 422, 63 S.Ct. 1141, 1145, 87 L.Ed. 1485, 1489 (1943) (stating that in order for the trial court's factual conclusion to withstand review "there must be findings . . . which are sufficient to indicate the factual basis for the ultimate conclusion")).

We are of the view that averaging, whether of appraisals or comparable sales, is not an appropriate methodology for assessing divergent values. The reasoned weight of authority provides sound policy reasons for such a conclusion. Properties are not fungible. Even with adjustments during the appraisal process, there are sufficient differences that must be weighed and considered by the fact-finder in addressing the ultimate issue in dispute. Here, the properties offered by the three appraisers were disparate in a host of factors, including time of sale; location, amenities and virtually all of the elements that enter into the appraisal process.

Finally, we view with caution the danger foreseen by those courts that have addressed this issue. To allow averaging will result in appraisals slanted to the extreme. Averaging will generate appraisals that will intentionally distort and skew the values to insure a high or low number without concern that the fact finder must resolve the issue with a careful analysis of data that may result in adoption of one appraisal figure over another.

In sum, we conclude that the averaging engaged in by the trial judge here so flawed the process that his finding of fair market value cannot be sustained.

We briefly address one additional issue-the adjustments to the comparable sales. As we previously noted, Hedden adjusted the comparable sales as high as 166%.

In ascertaining the value of a property, sales prices of comparable properties are critical to market analysis. See Inmar Assocs. Inc. v. Edison Twp., 2 N.J. Tax 59, 66 (Tax 1980). Variations between comparable sales and the subject property are anticipated and accommodated by making adjustments. Glenpointe Assocs. v. Teaneck Twp., 241 N.J.Super. 37, 48-49, 574 A.2d 459 (App.Div.), certif. denied, 122 N.J. 391, 585 A.2d 392 (1990). The differences are "dealt with by adjustments which recognize and explain the differences and then by relating the two properties to each other in a meaningful way so that an estimate of value can be determined." Id., at 49, 574 A.2d 459 (citing U.S. Life Realty Corp. v. Jackson Twp., 9 N.J. Tax 66, 72 (Tax 1987)).

"Adjustments that are too large [however,] suggest a lack of comparability between the concerned sales and the subject property and present a misleading indication of the subject property's value." 125 Monitor Street, LLC v. Jersey City, 21 N.J. Tax 232, 243 (Tax 2004), aff'd, 23 N.J. Tax 9 (App.Div. 2005). For example, in Pepperidge Tree Realty Corp. v. Kinnelon Borough, 21 N.J. Tax 57, 71 (Tax 2003), an appraiser presented four comparable sales, three of which had gross adjustments of 30% plus 10% for time, 30% plus 10% for time and 45% plus 11% for time. The Tax Court held that these high adjustment percentages raised doubts as to the true comparability of the underlying sales. Ibid. See also 125 Monitor Street, supra, 21 N.J. Tax at 242-14 (holding the same for adjustments of 20% plus 3% for time, 45% plus 10% for time, 35% plus 12% for time and 40% with no time adjustment).

Adjustments to sales of a large magnitude "vitiate comparability." Global Terminal Container Service v. City of Jersey City, 15 N.J. Tax 698, 704 (Tax 1996). In Global, where the appraiser made large gross adjustments including 50%, 50%, 75%, 50%, 65% and 55%, the court held that adjustments of that magnitude destroyed measures of comparability for the property. Id. at 702-04.

We deem Hedden's comparability and the adjustments of the magnitude that we have described (in one instance 166%) to raise significant questions as to the validity of the appraisal in the first instance. This should be addressed on remand.

Reversed and remanded for a new trial as to valuation.