The subject of these complaints are contiguous industrial zoned properties situated at the intersection of Ridgedale Avenue and East Frederick Place and known as block 2104, lots 2, 8 and 9. The taxpayers appeal the 1974-1975 assessments on lot 2 and the 1975-1976 assessments on lots 8 and 9. All assessments were affirmed by the Morris County Tax Board. In addition to the question of valuation, to be resolved is whether there existed discrimination in assessments as alleged by the taxpayers. The assessment history follows.
Lot 2 Lot 8 Lot 9
Land $ 56,200 $ 78,000 $ 46,100
Improvements 532,600 314,000 -
Total $588,800 $392,000 $ 46,100
*328Lot 2, which contains 5.62 acres, is located at the corner of Ridgedale Avenue and East Frederick Place and is improved with a three story (and penthouse) structure which was constructed in 1948 at a cost of $959,222. It is a concrete and steel structure with brick/masonry and glass block window exterior. On the interior it has 38,150 square feet of concrete floors surfaced with tile, glazed block and sheetrock walls, acoustic tile ceilings, tiled rest rooms, automatic passenger elevator, 150 ton air conditioning system, gas heating and sprinkler system. There is a truck dock. Since its construction the building has been used for various purposes including offices, laboratory and light industrial uses.
Lot 8, which contains 7.79 acres, fronts on East Frederick Place immediately to the rear of lot 9 which separates lot 8 from lot 2, is improved with four buildings and a railroad siding. Building A, the largest of the four structures, is a two story brick, steel and concrete structure built in 1948 and contains 36,562 square feet with brick interior walls and concrete floors. It has 13 inch thick wall partitions, ceiling heights of 20 and 40 feet, sprinklers, multi-paned industrial type windows, 10,000 pound capacity freight elevator, minimal toilet facilities, truck and rail loading facilities. It has been devoted to heavy industry and warehouse uses.
Building B, also constructed in 1948, contains two stories and is constructed of brick, steel and concrete. Having a gross floor area of 13,676 square feet, the structure has concrete floors and roof, sprinklers, freight elevators, 20 foot ceilings, rest room, truck door and truck loading platform. It has been devoted to manufacturing and warehouse uses.
Buildings C and D, also of 1948 vintage, are of brick, concrete and steel construction and contain 2,048 square feet and 2,312 square feet of floor space respectively. Used as a change house, building C contains lockers, showers and lavatories. Building D is a storage building.
Lot 9 is vacant and contains 4.6 acres of land. This lot was termed marginal because of its elevation, shape and wet condi*329tion. The Whippany River forms the southerly boundary of lots 8 and 9.
These structures, as well as additional improved and unimproved properties, were formerly owned and utilized by the Flintkote Company which had been engaged in the manufacture of asbestos products, a use which suggests the need, at least with respect to the buildings on lot 8, for substantial construction. The testimony indicated that Flintkote abandoned the premises in approximately 1970 (leaving only a caretaker in charge of the building on lot 2) and in 1972 the premises and more were purchased by Whippany Associates, an investment group, with Flintkote taking back a purchase money mortgage which mortgage was subsequently sold to the Marine Midland Bank. Apparently, in an effort to develop it into an industrial park, Whippany Associates subdivided the land into various lots and constructed East Frederick Place. Its efforts to sell the premises being to no avail Whippany Associates, in or about 1975, conveyed the premises in question to Marine Midland Realty Credit Corp. in lieu of foreclosure of the mortgage. Marine Midland, in January 1976, sold lot 2 for a consideration of $560,000 and, in March 1977, lots 8 and 9 were sold for $320,000.
The taxpayers’ expert, while utilizing the income approach, primarily relied upon the prices in the Marine Midland conveyances to establish valué. The township contends those sales are suspect and unuseable and not representative of transactions between willing parties, neither being under a compulsion to buy or sell. In fact, in computing his Township of Hanover sales ratio study, the Director of Division of Taxation found the sale of lots 8 and 9 to be unuseable. No reasons were given for the Director’s decision.
A real estate broker who was involved in the two transactions found that the sales were negotiated at arm’s length. While nothing was presented that would suggest otherwise, under the circumstances here present it is obvious that the prices did not, in fact, represent the true value of the premises *330as of the critical assessing date. It is well settled that the sales price of the subject is not controlling as to value but is merely evidential. McCrory Stores Corp. v. Asbury Park, 89 N.J.Super. 234, 235, 214 A.2d 526 (App.Div. 1965); Rek Investment Co. v. Newark, 80 N.J.Super. 552, 559, 194 A.2d 368 (App.Div. 1963). “The Division (Tax Court) [is] thus required to weigh and appraise the factors surrounding the sale to determine whether there were special circumstances which had a tendency to increase or depress the sales price of the property without affecting its true value, or whether, . . . the sale[s] price was controlling”. Id. pg. 560, 194 A.2d pg. 373.
I find the circumstances surrounding these sales indicate the prices were depressed. A bank is generally not in the business of renting and managing real estate holdings and when Marine took title to the properties in lieu of foreclosure, it behooved it to sell the properties in order to cover the mortgage loan as rapidly as possible.
Indeed, on cross examination the broker acknowledged that it was Marine’s policy to liquidate holdings which it had received in satisfaction of mortgage loans. Being aware of these circumstances, a prospective purchaser would naturally offer a price somewhat less than true value. I find the seller here was under a greater economic compulsion to sell than would be the ideal hypothetical “willing seller” and accordingly will assign the sales price no weight in my determination.
I likewise will assign no weight to the taxpayers’ capitalization of income analysis which allegedly was founded on economic rents. However, no actual comparable rents were offered and, in fact, the taxpayers’ expert admitted that he relied only on his general experience in formulating his net rental amounts. Obviously, he was forced to adopt that position as neither he nor the township’s expert could discover comparable rentals of buildings of this type. As earlier noted the subject buildings were vacant as of the assessing date. The court cannot rely on an expert’s unsupported opinion without any foundation to verify its accuracy and reliability. See In Re: Appeal of City of *331East Orange, 103 N.J.Super. 109, 111, 246 A.2d 722 (App.Div. 1968).
Both experts cited a comparable sale of improved property from Whippany Associates to Frederick Realty Corporation in July 1974. Neither relied heavily on that sale and this Court also assigns it little weight. I note that the comparable sales approach, as a general rule, is not readily applicable to industrial/commercial buildings because of their diversity in design, materials, size and age.
The township’s expert appraiser relied on the reproduction cost approach and trending of original cost to establish improvement value. In the comparable sales approach in valuing the land, he submitted 19 sales but did not describe them in detail. The record is silent as to the topography, condition, location or circumstances of the sales. However this Court will rely on the sale of block 2104, lot 6 in March of 1977. We are able to determine from a map in evidence that this vacant 8.02 acre parcel, which sold for $21,200 per acre, is located opposite the subject lots 8 and 9 on Frederick Place. Lot 6 is comparable to the subjects in size and location. After adjusting for time, I find the land value of lot 8 to be $155,800 based on $20,000 per acre. Lot 9 is described by both experts as being marginal due to its low, wet nature and its shape. Adjusting for this marginal condition and time, I find the true value of lot 9 is $69,000 at $15,000 per acre. However, lot 2, primarily because of its location, is found to have a greater value which I establish at $22,500 per acre for a total value of $126,450.
As to the improvement value, I find the reproduction cost approach supported by trending original costs the most reliable indicator of value in this case. New Brunswick v. State Div. of Tax Appeals, 39 N.J. 537, 543, 189 A2d 702 (1963); In re: Erie RR System, 19 N.J. 110, 115 A.2d 89 (1955). Relying on the Real Property Appraisal Manual for New Jersey Assessors, (2nd ed.) the witness placed the lot 2 improvement in the 25.07 class and calculated the 1954 cost at $761,267. An application of the cost conversion factor for 1974 of 280% indicates a $55 per square *332foot reproduction cost. This sum finds further support in trending the original cost for 1948 to 1975 utilizing the Dodge Building Cost Calculator and the Historical Local Cost Index of 250%. The witness then deducted 60% for depreciation and a reasonable restoration cost of $150,000 resulting in a total improvement value of $689,300.
The improvements on lot 8 were classified as 42.3 industrial buildings pursuant to the Manual. The expert calculated a reproduction cost in excess of $40 per square foot but settled on $40 per square foot unit price which is supported by trending the original cost. The unit price includes the loading facilities, additional site improvements and takes into account the differing building heights. Deducting 70% for depreciation and $275,-000 for restoration, which sum finds ample support in the record, he found the true value of improvements on lot 8 to be $380,200.
I find the reproduction cost of the lot 2 improvement to be $55 per square foot or $2,098,250 ($55 X 38,150 sq. ft.). The $55 figure is supported by the witness’ competent analysis, the Appraisal Manual and trending the original cost. The depreciation rate for a 28 year old building is 20% according to the Manual. An additional 20% is deducted for the dilapidated condition of the premises. Since single story modern buildings are favored for laboratory or light industrial use, the Court will apply a 25% factor for functional obsolescence to this three story building. “Functional obsolescence refers to the loss in value resulting from conditions within the building itself * * * Bostian v. Franklin State Bank, 167 N.J.Super. 564, 572-573, 401 A.2d 549, 553 (App.Div. 1979).
Applying the total deductions of 65% to the reproduction cost of $2,098,250 we find a net value of $734,387. After deducting the $150,000 cost of restoration the true value of the improvements on lot 2 is found to be $584,387.
On the basis of the expert’s testimony and utilization of the Appraisers Manual and the Dodge Building Cost Index, I find the reproduction cost of $40 per square foot or a total of $2,184,000 (54,600 sq. ft. X $40) on the lot 8 improvements. *333Again 40% will be deducted for physical depreciation based on a 28 year effective age and the deteriorated condition of the improvements. The buildings on lot 8 were described as containing bird infestation, broken windows, disconnected heat, vandalized electrical system, frozen sprinkler system, leaking roof, and unuseable railroad siding. A further deduction of 32% comprising a factor for both functional and economic obsolescence will be applied taking into consideration the heavy, explosion proof construction and design which is considered outdated and the high vacancy rate for the years under appeal. See Anaconda Company v. City of Perth Amboy, 157 N.J.Super. 42, 384 A.2d 531 (App.Div. 1978). Applying the total rate of 72% to the reproduction costs leaves a net to the improvements of $611,520. An additional $275,000 for the cost of restoration will be deducted. The Court notes this figure is actually in excess of the taxpayer’s restoration estimate of $186,500. In summary, I find the true value of improvements on lot 8 to be $336,520.
The taxpayers allege discrimination in assessments and seek relief in accordance with In Re: Appeal of Kents, 2124 Atlantic Avenue, Inc., 34 N.J. 21, 166 A.2d 763 (1961).
The N. J. Supreme Court has addressed the issue of discrimination in two recent cases. In Tri-Terminal Corporation v. Boro, of Edgewater, 68 N.J. 405, 346 A.2d 396 (1975), the Court noted three separate attacks which may be leveled against an assessment in order to prove discrimination. First, the taxpayer may challenge the relative fairness of a revaluation of his property compared to that of other taxpayers generally in the taxing district. Second, the taxpayer may prove the absence of a common level by showing a wide range in ratios of assessments to sales reflected in individual sales within the municipality. Lastly, a discrimination claim may be found where a taxpayer proves his property did not rise in value compared to other properties, such as to render its assessment ratio substantially higher than others generally. In Tri-Terminal, the taxpayer did not advance the first and third arguments but attempted to prove lack of a common level. The Court found a common level *334and further found that the premises did rise in value together with all properties in the borough.
Piscataway Associates v. Twp. of Piscataway, 73 N.J. 546, 376 A.2d 527 (1977) dealt only with the common level issue. In finding an absence of a common level, the Court noted most of the sales ratios fell in the 40%-100% range.
Initially I note the taxpayer does not challenge the efficacy of the revaluation which presumptively establishes a common level. That common level may be eroded by the passage of time if no revaluation is conducted due to the uneven effect of economic forces on the various properties in the taxing district. All property in Hanover is assessed at 1961 values which were carried forward unchanged in subsequent years. That 1961 common level would ordinarily become eroded over a 13 year period but the facts in this case simply do not indicate such a trend. The statistical study produced by the taxpayer clearly indicates the existence of a common level in Hanover from 1974 through 1976. The study contains a sufficient number of useable sales (125-1974,127-1975,115-1977) which can be considered a reliable representation of all property within Hanover. The statistical analysis which is a summary of the New Jersey Division of Taxation sales ratio studies for Hanover Township indicates general coefficients of deviation of 10.05%, 11.18% and 15.23% for the three tax years respectively. The coefficient is a most instrumental statistic in determining the existence of a common level. Since all ratios of assessments to market prices are not identical it is useful to consider whether the ratios cluster around one focal point. In the study, that point is the average (mean) ratio. The variances from the mean of each individual ratio (assessment to sale) are also averaged to form a coefficient of deviation. Thus the coefficient is a statistic which measures the degree of departure of all the ratios (assessment to sale) from the average ratio. If the coefficient is small the ratios are concentrated near the mean and if the coefficient is large the ratios are scattered widely about the mean. Carrying *335this analysis one step further, a small coefficient indicates the existence of a common level which is the average ratio. In the alternative, where the coefficient is large, no common level exists and the average ratio becomes inconsequential. In Tri-Terminal Corp. v. Boro, of Edgewater, supra, 68 N.J. at 413, 346 A.2d 396, the Court, relying on the Handbook for New Jersey Assessors (revised 1965, page 203, by Bureau of Government Research Rutgers State University), noted that a coefficient of deviation under 20% was acceptable. Therefore, we may infer a coefficient under 20% is sufficiently small to indicate the existence of a common level.1 Since the coefficients of deviation in the subject case are all well below the 20% mark, I find the existence of a common level at the average weighted ratio of 44.9%, 43.34% and 40.14% for 1974, 1975 and 1976 respectively.
Even though a discrimination claim based on the absence of a common level has been rejected the taxpayer may prevail if it is shown that its property did not rise in value compared to other properties in the taxing district. I find the evidence clearly indicates the subject’s rise in value was considerably less than other properties. The extremely poor physical condition of the premises has limited the rise in value. The increase in market values of other properties from the 1961 valuation at 100% of true value is documented in the average assessment to sales ratios for the tax years (44.90%, 43.34% and 40.14%). The ratios of assessment to true value are 84.5%, 79.6% and 66.8% for the subject lots 2, 8 and 9 respectively. Thus the plaintiffs are being taxed at a much higher rate than all others in the taxing district and are therefore being discriminated against. “It is fundamental that a taxpayer is entitled to ‘treatment commensurate with that given his fellow taxpayers within the municipality’ and that if it is not accorded, he is entitled to a judicial or *336quasi-judicial remedy.” In Re: Appeal of Kents, supra, 34 N.J. at 25, 166 A. at 763; Tri-Terminal Corporation v. Boro, of Edgewater, supra, 68 N.J. at 409, 346 A.2d at 399; Piscataway Associates, Inc. v. Twp. of Piscataway, supra, 73 N.J. at 556, 376 A.2d 527.
It has been determined the weighted average ratio constitutes a common level in this case. That average ratio will be applied in order to reduce the true value to the common level of assessments in Hanover. The weighted average ratio for all classes of property is utilized because the reduction should be to the common level of all real property rather than to the level of industrial property. New Brunswick v. State of New Jersey, Director, Div. of Taxation, 39 N.J. 537, 542, 189 A.2d 702 (1963). The Court once again notes that there are a sufficient number of useable sales to consider the study reliable; thus, there is no need to consider the average ratios of other years.
Judgments shall be entered as follows.
N.J.S.A. 54:1-35A fixes a common level range which is plus/minus 15% of the average ratio found by the Director, Division of Taxation. This Act was effective for the tax year 1978.