OPINION
SIMANDLE, District Judge.I. INTRODUCTION
In this consolidated Multi-District Litigation class action suit, plaintiffs claim that they own or owned vehicles manufactured by defendant Ford Motor Company (“Ford”); that their vehicles came equipped with defective ignition switches manufactured by defendant United Technologies Corporation (“UTC”); that defendants knew that the switches were defective and nonetheless persisted in selling the affected cars and trucks; and that vehicles were damaged by fires caused by short circuits in the defective switches.
Presently before the Court is the renewed motion of the Snodgrass plaintiffs1 for class certification. This Court’s jurisdiction is based upon diversity of citizenship, and the laws of nearly 50 states give rise to the various causes of action asserted by members of the proposed classes and subclasses. The main issue for decision is whether plaintiffs’ amended class definition satisfies the prerequisites for class certification set forth in Fed.R.Civ.P. 23(b)(3). For the reasons discussed below, the Court concludes that even under plaintiffs’ revised proposed class definition, this case is not appropriate for class action treatment, and will deny plaintiffs’ application.
II. BACKGROUND
The background of this case is discussed in detail in this Court’s August 1997 Opinion and Order denying first plaintiffs’ motion for class certification. See generally In re Ford Motor Co. Ignition Switch Products Liability Litigation, 174 F.R.D. 332 (D.N.J.1997) *486(hereinafter “In re Ignition Switch”). The present opinion incorporates the factual and procedural discussion therein, and only a short background summary is needed here.
In brief, Snodgrass plaintiffs’ predecessors’ initial motion for class certification in this case2 sought to certify a class of “all persons or entities who purchased or who leased, other than for purposes of resale or leasing, one or more Ford vehicles in model years 1984 through 1992 and who owned Ford vehicles from model year 1993 with ignition switches manufactured before October 1992, and whose vehicles caught on fire as a result of a defective ignition switch.” CWilks Not. of Mot.) No personal injury or wrongful death claims were asserted.
By Order dated August 28, 1997, this Court denied plaintiffs’ motion for class certification without prejudice to later renewal of such motion. See In re Ignition Switch, 174 F.R.D. at 356. Among the reasons for the Court’s denial of plaintiffs’ proposed class certification were that, for the purposes of the proposed class (1) there were not sufficient common factual and legal issues, id. at 342-51, and (2) plaintiff failed to show that class action was a superior method of adjudicating the controversy, id. at 351-54. The Court left open the possibility, however, that “after further discovery ... the 158 model years could be grouped into a few risk categories if a common trial were to be held ... ”, id. at 345, and thus the dismissal was without prejudice to re-filing for certification upon curing these deficiencies and submission of a “suitable blueprint for trial”. Id. at 356. Plaintiffs now claim that they have cured the deficiencies which impaired the original certification motion, and once again move the Court to approve the plaintiffs’ request for class action treatment.
III. DISCUSSION
A. Class Action Certification Standards
1. Rule 23(a) Class Action Predicates
Rule 23(a) of the Federal Rules of Civil Procedure sets forth four general preconditions that putative class representatives must satisfy before any case is certified as a class action:
(1) the class is so numerous that joinder of all members is impracticable, (2) there are questions of law or fact common to the class, (3) the claims or defenses of the representative parties are typical of the claims or defenses of the class, and (4) the representative parties will fairly and adequately protect the interests of the class.
Fed.R.Civ.P. 23(a). Additionally, Rule 23(b) defines four different types of class actions, and the party seeking certification must demonstrate that the case falls within one of the Rule 23(b) categories. The plaintiffs bear the burden of proving that the proposed class action satisfies each of the requirements of Rule 23(a) and one of the prerequisites of Rule 23(b). See Baby Neal v. Casey, 43 F.3d 48, 55 (3d Cir.1994).
A motion for class certification should not turn on the court’s evaluation of the merits of the parties’ legal or factual claims. See Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 178, 94 S.Ct. 2140, 40 L.Ed.2d 732 (1974). The court may find it necessary, however, to analyze the elements of the parties’ substantive claims and review facts revealed in discovery in order to evaluate whether the requirements of Rule 23 have been satisfied. See Castano v. American Tobacco Co., 84 F.3d 734, 744 (5th Cir.1996) (citing Manual for Complex Litigation 3d § 30.11 (3d ed.1995)).
2. Rule 23(b)(3) Class Action
As in the first motion for class certification in this case, the parties’ arguments are directed to whether plaintiff has satisfied class certification requirements of Fed.R.Civ.P. 23(b)(3). In order for a class action to be certified under Rule 23(b)(3), the putative class representatives must show, in addition to the four general prerequisites of Rule 23(a), that:
*487questions of law or fact common to the members of the class predominate over any questions affecting only individual members, and that a class action is superi- or to other available methods for the fair and efficient adjudication of the controversy.
Fed.R.Civ.P. 23(b)(3). The Supreme Court has recently interpreted Rule 23(b)(3)’s requirements when it affirmed the Third Circuit’s determination that certification of a nationwide settlement class for alleged asbestos victims was improper. Amchem Prods., Inc. v. Windsor, 521 U.S. 591, 117 S.Ct. 2231, 138 L.Ed.2d 689 (1997), aff'g Georgine v. Amchem Prods., Inc., 83 F.3d 610 (3d Cir.1996). In that case, the Supreme Court reiterated that whether “questions of law or fact common to the members of the class predominate over any questions affecting only individual members” may be referred to as the predominance requirement. The predominance inquiry “trains on the legal or factual questions that qualify each class member’s case as a genuine controversy----” Amchem Prods., supra, 117 S.Ct. at 2249. “The Rule 23(b)(3) predominance inquiry tests whether proposed classes are sufficiently cohesive to warrant adjudication by representation.” Id. (citing 7A Charles Alan Wright, Arthur R. Miller & Mary Kay Kane, Federal Practice and Procedure § 1777, at 518-519 (2d ed.1986)).
B. Plaintiffs’ Renewed Motion for Class Certification
Plaintiffs’ renewed motion for class certification seeks to certify a truncated version of the original proposed class, and proposes two specific subclasses. These are (1) an “Implied Warranty (fitness for Ordinary Purposes) Subclass”, and (2) a “Deceptive Trade Practice Subclass”. (Snodgrass Mot. at 1-3.)3 (The full proposed class definition is set forth in the Appendix hereto.)
Plaintiffs’ proposed Implied Warranty subclass would include “All owners of Ford Class Vehicles that have burned or been damaged by fire because of a defective Fox ignition switch during the Class Period”. (Id. at 1.) Plaintiffs seek to restrict this class by excluding, inter alia, personal injury claims, purchasers of vehicles in at least seventeen states that require vertical privity, purchasers other than “consumers” in four states, and purchasers in Mississippi whose cars burned after 36,000 miles. (Id. at 2.)
The proposed Deceptive Trade Practices, group is even more complex. This subclass would also include “All owners of Ford Class Vehicles which have burned because of defective ignition switches during the Class Period”, but would restrict class membership to those plaintiffs who have in common similar state statutes that (1) prohibit unfair or deceptive trade practices generally, (id. at 3 (listing 30 states)), and/or (2) prohibit “conduct such as representing that goods have characteristics, uses, benefits, or qualities that they do not have or that goods are of a particular standard, quality, or grade if they are another” (id. (listing nine states)). Included in the second group are several sub-subclasses for plaintiffs from states that (a) require a showing of scienter (10 states), (b) require a showing of some “aggravating factor” other than scienter (2 states), and (c) require proof of justifiable individual reliance (3 states). (Id.) In addition to the specific requirements for membership, the deceptive trade practices subclass would exclude, inter alia, persons other than “consumers” residing in 22 states, and persons whose claims would be time-barred. (Id. at 4.)
For reasons discussed below, the court holds that common issues of law and fact do not predominate even these elaborately restricted proposed subclasses, that class action is not a superior means of litigating plaintiffs’ claims, and that class certification is not appropriate under Rule 23(b)(3). Accordingly, the court need not and does not reach the question whether plaintiffs have satisfied the threshold Rule 23(a) requirements.
C. Whether Individual Issues Predominate Even Under Plaintiffs’ Revised Class Definitions
The Supreme Court has noted that the predominance requirement is not met *488when there exist a greater number of questions peculiar to the several categories of class members, and to individuals within each category, where such uncommon questions are significant. Amchem Prods., 117 S.Ct. at 2250. In the context of the massive class certified by the district court in the Georgine .asbestos litigation, the Supreme Court held that the class cohesion that might emerge from the common issues of fact regarding the consequences of asbestos exposure was undermined by the disparities among class members in the means of exposures to different asbestos-containing products, for various lengths of time, resulting in disparities in injury ranging from no physical injury to lung cancer or mesothelioma, with complications of causation arising from cigarette smoking. Id. at 2250 (quoting the Third Circuit’s discussion in Georgine v. Amchem Products, Inc., 83 F.3d 610, 626 (3d Cir. 1996)).
That common issues must be shown to “predominate” does not mean that individual issues need be non-existent. All class members need not be identically situated upon all issues, so long as their claims are not in conflict with each other. Eisenberg v. Gagnon, 766 F.2d 770, 786 (3d Cir.), cert. denied, 474 U.S. 946, 106 S.Ct. 342, 88 L.Ed.2d 290 (1985); Hassine v. Jeffes, 846 F.2d 169, 176-77 (3d Cir.1988). The individual differences, however, must be of lesser overall significance and they must be manageable in a single class action, as discussed further below.
Here, plaintiffs have proposed a two-stage trial for adjudication of the vehicle owners’ claims. Stage One would involve a “class trial” during which competing experts would present classwide statistical and mechanical evidence tending to prove that the fires were caused by the faulty ignition switches. (PL Reply Br. at 10.) Plaintiffs argue that their evidence at this stage would establish a “re-buttable statistical presumption” of causation. (Id.) Plaintiffs recognize that statistical evidence of causation would include some “false positives” — fires that fall within the statistical boundaries but were not actually caused a faulty ignition switch. (Id.) Nevertheless, plaintiffs assert that Stage Two would account for such non-ignition fires. To this end, plaintiffs propose that, once the Stage One proofs are completed, Stage Two would allow defendants the opportunity to rebut this statistical proof.
During Stage Two, defendants would be permitted to show on a case-by-case basis that the ignition switches were not the cause of the fire at issue. (Id. at 11.) At this point, defendants would attempt to persuade the jury that certain individual fires were caused by something other than the defective switch. (Id.) Alternative causes could include the installation of after-market equipment or simple negligence on the part of the owner.
For the reasons next discussed, the Court holds that under plaintiffs’ proposed two-stage litigation plan, class-wide issues do not predominate over issues involving only individual plaintiffs. The Court’s holding results from the following four conclusions: (1) even under the revised proposed class definitions, there exist disparate legal standards that would complicate efforts to prove that common issues predominate the proposed subclasses; (2) the proposed subclasses are not amenable to plaintiffs’ proposed prima facie proof of causation through statistics; (3) the causation issues in this ease are inherently individual in nature and will depend on facts particular to each individual plaintiff; and (4) owing to the abundant legal and factual issues that would have to be determined on a case by case basis, en masse litigation is not a superior method for addressing the vehicle owners’ individual incident claims.
1. Whether Common Legal Issues Predominate
Were this Court to certify plaintiffs’ proposed subclasses, the Court first would have to resolve the knotty problem of which legal standards should be applied to each individual claimant. The Court also would have to consider individual legal defenses raised by defendants. The Court discusses these problems in turn.
a. Disparate Legal Standards
As this Court noted in its original opinion in this case, there are a multitude of different legal standards to be applied to *489individual claimants’ warranty, fraud, and consumer protection claims. In re Ignition Switch, 174 F.R.D. at 351. The Court now finds that, even under plaintiffs’ refined class definition focusing only on implied warranty and deceptive trade practices claims, such claims vary significantly from state to state, and that this issue thus tilts in favor of denying certification.
With respect to plaintiffs’ claims of breach of implied warranty, plaintiffs claim to have resolved the discrepancies between the states’ laws in this area by excluding from their proposed class “implied warranty privity states”. (PI. Br. at 3ÍM2.) However, even after excluding the residents of privity states from the proposed class, the Court still would have to wade through an analysis of the implied warranty laws applicable to each individual. Among the distinctions that would remain for resolution at trial are:
• Each relevant state’s definition of merchantability, and whether the alleged defect meets that threshold. Compare Nacci v. Volkswagen of Am., Inc., 325 A.2d 617, 620 (Del.Super.Ct.1974) (under Delaware law merchantability standard is whether ordinary manufacturer would pursue different design), with Venezia v. Miller Brewing Co., 626 F.2d 188, 190 (1st Cir.1980) (under Massachusetts law merchantability examines reasonable consumer expectations).
• The nature of an implied warranty claim under the relevant state’s law, i.e., whether the remedy lies in tort or contract. Compare Parrillo v. Giroux Co., 426 A.2d 1313, 1317 (R.I.1981) (contract' remedy), ivith First Nat’l Bank of Dwight v. Regent Sports Carp., 803 F.2d 1431, 1438 (7th Cir.1986) (applying Illinois law) (tort remedy).
• Whether the relevant State allows waiver of the Implied Warranty, and whether such waiver occurred. Compare Settlemires v. Jones, 736 So.2d 471 (Miss.Ct. App.1999) (no waivers of implied warranty), with Star-Shadow Prods., Inc. v. Super 8 Sync Sound Sys., 730 A.2d 1081 (R.I.1999) (waiver allowed even if phrase “implied warranty” not included in general waiver).
Plaintiffs have also alleged violations of deceptive trade practice statutes. A review of these state laws shows that there exist significant distinctions between the states’ consumer protection statutes. As a preliminary matter, it must be noted that the deceptive trade practice claims are contingent on the implied warranty claims. In other words, defendants cannot be held liable under the deceptive trade practice statutes unless plaintiffs succeed in proving that the defective ignition switched rendered their vehicles unmerchantable, and that defendants then knowingly sold the unmerchantable vehicles. (In re Ford Ignition Switch Products Liability Litigation (Snodgrass), Civ. Nos. 96-3125; 96-1814(JBS), Slip Op. at 19 (D.N.J. May 14, 1999).) In addition to the contingent nature of these claims, of the states included in plaintiffs’ proposed subclass, there exist many legal variations between the states’ consumer protection laws. Mindful of these variations, plaintiffs have attempted to screen out those states’ laws that require privity. Even after deleting the privity states, however, key distinctions remain.
The chief difficulty in finding a predominance of common legal issues within the proposed deceptive trade subclass is that the statutory protections vary greatly from state to state. For example, conduct actionable under one state’s laws may not be actionable under another’s. Compare N.J.S.A. § 56:8-2 (any unconscionable practice), with Cal. Civ. Code § 1770 (proscribing 23 specific activities). Moreover, despite plaintiffs’ attempt to limit the deceptive trade practice subclass only to “consumers”, problems remain. This is because some states have different definitions of the word “consumer”. Compare Ga. Code Ann. § 10-1-392(3) (consumer is a purchaser of goods for personal, family or household), with W.Va.Code § 46A-6-102 (“‘consumer transaction’ means a sale ... for a personal, family, household or agricultural purpose”) (emphasis added).
Based on the foregoing, the Court finds that there exist significant differences between the states’ causes of action for violation of implied warranty and deceptive trade practices. These distinctions lessen *490the predominance of common legal issues. Moreover, the practicalities of applying such varied state law would demand significant attention from this Court, not the least of which would be instructing the jury or juries consistent with the law of each relevant state. In sum, even under the revised class definitions, this Court finds that common legal issues do not predominate. Accordingly, the continuing legal distinctions present in this case favor denial of certification,
b. Available Legal Defenses to Vehicle Owners’ Individual Claims
There exist additional roadblocks to plaintiffs’ proposed classes even beyond the difficulty in applying the legal distinctions discussed above. One such barrier is the question of which states’ law should apply to particular claims. This Court has previously stated that the law to be applied to each individual claim is that of the plaintiffs, home state. In re Ignition Switch, 174 F.R.D. at 348. As defendants have pointed out, choice of law may prove to be a point of debate if a plaintiffs residence is in doubt, or if he bought his vehicle somewhere other than his home state.
Another knotty issue is whether plaintiffs were comparatively or contributorily negligent. These ignition switches are located in upon the vehicle’s steering column in the passenger compartment. The Court already has recognized that there are many causes of electrical fires, including after-market add-ons. . Id. at 347. The possibility that some of the potential plaintiffs had such accessories added decreases the likelihood that common issues predominate. Moreover, even if this court were to allow plaintiffs’ proposed two-step trial to proceed, defendants would have to be given the opportunity to cross-examine each plaintiff as to whether such accessories were in fact added, among other individual liability issues. Passenger compartment fires may likewise have many caused in the dashboard and steering column, or from materials within the front seat area having nothing to do with the ignition switch.
Furthermore, the calculation of damages for each plaintiffs loss, assuming liability, would require individualized attention at trial. As plaintiffs admit, not all vehicles within its proposed subclasses were completely destroyed by fire. Some were partially burned, and some were not burned at all, but weré instead temporarily filled with smoke. Moreover, some vehicles are worth more (or less) than the ascribed bluebook value. Accordingly, the damage involved in each claim is not common, and the amount of damages to be awarded depends upon the particular facts of each owners’ claim, occurring in vehicles spanning nine years and numerous makes and models.
Based on the foregoing, the Court finds that there are a number of important individual issues that would have to be resolved in each class member’s case should plaintiffs’ proposed subclasses be certified. The predominance of individual issues in this case militates against class treatment.
2. Whether Common Factual Issues Predominate
It is axiomatic that individual causation remains an prerequisite to class membership. Resolution of the “general causation” question of whether the subject switches are capable of causing the damage alleged by the vehicle owners does not show commonality under Rule 23(a)(2). See Arch v. American Tobacco Co., Inc., 175 F.R.D. 469, 488 (E.D.Pa.1997), aff'd by Barnes v. American Tobacco Co., Inc., 161 F.3d 127 (3d Cir.1998). The question is not whether the switches have the capacity to cause harm, but rather the highly individualistic inquiry of whether it did cause harm and to whom. Id. citing In re “Agent Orange” Product Liability Litigation, 818 F.2d 145, 164 (2d Cir.1987). In other words, there can be no inclusion of individuals in either of plaintiffs’ proposed subclass unless that person can show that the subject switches caused damage to their vehicle.4
*491A review of the history of this litigation demonstrates the need for individual causation investigation. The original lead plaintiff in this case, Michael Wilks, was dropped from the case after it came to light that he had himself set fire to his car. See In re Ford Ignition, 174 F.R.D. at 347. Likewise, plaintiff Doreen Giddings’s claim was voluntarily dismissed after an investigation revealed that the fire in her car was not related to the ignition switch. (Rept. of Defs.’ Statistical Expert Dr. Subbaiah Malladi, Ph.D., at 46.)
Furthermore, investigations by plaintiffs’ fire investigation expert, William Hagerty, led to the voluntary dismissal of two of the eight plaintiffs whose cars he investigated. (Hagerty Dep. at 46-47.) As shown by the events surrounding Mr. Hagert/s work for the plaintiffs, there is much uncertainty over whether the ignition switches themselves caused the fires, even among persons named as individual plaintiffs and class representatives. Other proposed causes include the installation of after-market equipment, which equipment appears to have increased the likelihood of fire.
Plaintiffs have not alleged that the subject switches always cause the type of harm alleged by the vehicle owners. Indeed, the very definition of plaintiffs’ revised proposed subclasses admits that individual causation remains a critical part of class membership. Under the revised class definition, individuals would not be included unless they are an owner of a vehicle that has been “burned or been damaged by fire because of a defective Fox ignition switch”. Thus, as recognized by plaintiffs, it is a threshold requirement for class membership that any fire or smoke damage must actually have been caused by one of the subject switches. As shown by a review the language of plaintiffs’ proposed subclass definitions, then, class membership is earned by demonstrating a common thread of causation of the particular vehicle fire linked to the ignition switch to the exclusion of other causes. Fundamentally, this raises a problem that one is unable to know who is a member of the class (and thus bound by the court’s processes) until a uniform basis of demonstrating causation emerges.
a. Plaintiffs’ Proffer of Statistical Evidence of Ignition Switch Failure Rates
This Court noted in its earlier opinion denying class certification that “there exist troublesome issues of causation that may require special ease-by-case measures rendering the [Snodgrass ] plaintiffs’ claims unsuitable for class certification.” In re Ford Ignition, 174 F.R.D. at 347. Mindful of this observation, plaintiffs now have attempted through use of statistics to answer the Court’s call for evidence showing causation on a subclass basis. The data and statistical analysis performed for these purposes by the Snodgrass experts and counsel reflect much hard work and expense in attempting to refine the analysis to show coherence of the class. For reasons now discussed, the Court finds that plaintiffs’ statistical evidence of causation at most may support a rebuttable presumption that the subject switches caused fires in a significant number of vehicles, but that this presumption still will require case-by-case litigation of whether each putative plaintiffs damages actually were caused by defendants’ switches. While statistical analysis may greatly aid the proof of individual incidents through the use of inferential evidence, the continued need for individual litigation of causation makes this case unsuitable for class action treatment.
b. Whether Plaintiffs Have a Legal Basis for Their Statistical Proffer
Even if the Court were to assume that the plaintiffs’ database was sufficiently reliable to create a presumption that the majority of the proposed subclass members owned vehicles with faulty ignition switches, such an assumption would not obviate the need for detailed claimant-specific investigations and jury trials in order to determine whether the subject switches actually caused the damage alleged. This is because rather than direct proof that the subject switches *492caused the fires, plaintiffs’ statistical proofs are geared only towards creating a “rebutta-ble presumption” of causation. (PL Br. at 11-22.)
Defendants assert that, even assuming that plaintiffs’ statistical presumptions are valid, this statistical analysis of Ford vehicles’ fire rates could at most suggest to a jury that a putative plaintiff — an owner of a given model/year Ford complaining of a steering column-area fire — might have experienced an ignition fire. Defendants also maintain that the fact that there is an increased statistical likelihood that an ignition switch caused a fire is not conclusive. Because the proffered statistical evidence cannot by itself establish individual causation, defendants argue, an analysis of the statistical likelihood of causation is an exercise in futility.
As further support for their argument that statistics cannot establish causation in this case, defendants cite the Third Circuit’s recent decision in Barnes v. American Tobacco, 161 F.3d 127, 135 (3d Cir.1998), cert. denied, 526 U.S. 1114, 119 S.Ct. 1760, 143 L.Ed.2d 791 (1999), for the proposition that even if plaintiffs are able to narrow the field of potential plaintiffs through statistics, the individual plaintiffs still would have to demonstrate that they suffered compensable damages on account of a Fox ignition switch.
In Barnes, a group of plaintiffs pursued a class action lawsuit against the tobacco companies on the theory that defendants intentionally enhanced the addictive characteristics of their products. Plaintiffs argued that a Surgeon General’s report conclusively established that cigarette smoking is a major cause of three different diseases, and that this evidence more than satisfied their burden of showing that addiction to tobacco caused them harm. Id. at 145. The court rejected this argument, stating that “plaintiffs cannot prove causation by merely showing that smoking cigarettes causes cancer and other diseases. They must demonstrate that defendants’ ... nicotine manipulation caused each individual plaintiff to have a significantly increased [risk of harm].” Id. The Barnes court agreed with the District Court’s decision to reject plaintiffs’ effort to resolve the issue of individual causation" by having each class member answer a questionnaire, stating that the questionnaire would at most establish a prima facie case of causation. Even after a showing of prima facie evidence of causation, the court held, defendants must be allowed the opportunity to cross-examine each individual plaintiff and other witnesses on the plaintiffs behalf, and offer expert testimony about each person’s specific circumstances. Id. at 145-46. The Third Circuit agreed with the District Court’s conclusion that, owing to the need for individual cross-examination of each individual plaintiff, plaintiffs were proposing an impossible litigation plan. Id. at 146, aff'g, Arch v. American Tobacco, Inc., 175 F.R.D. 469, 488 (E.D.Pa.1997).
Plaintiffs argue that statistical proof of causation has been approved by other courts, and that the holding in Barnes is inapposite and should be confined to the particular facts of that case.5 (Pl. Br. at 7 & n. 7 (citing cases).) The Court disagrees. The similarities between Barnes and this case are manifest. In both cases, the plaintiffs present indirect evidence of causation, and propose a system of rebuttal trials.
This Court finds that Barnes counsels that any proffer of prima facie proof of causation must be viewed with skepticism because it fails to resolve any individual claim without more evidence, and that a threshold require*493ment for certification is a realistic litigation plan for resolution of individual causation,
c. An Analysis of Plaintiffs’ Statistical Proffer
Plaintiffs claim to have created a database of specific model/years of Ford vehicles that have a far greater statistical likelihood of ignition switch fires than do most models. (Pl.Br. at 7.) This database is derived from an updated collection of vehicle owners’ complaints contained within Ford’s Master Owner Relations System (“MORS II”). The individual MORS II complaint entries were created by Ford customer service personnel taking down individual customers’ complaints' sent by way of letter or telephone, and include the model/year of the owner’s vehicle, and the owner’s stated complaint. These reports are terse, often unverified, and in-1 conclusive. For example, MORS II entry 19.03.52 involves a November 1995 complaint by an owner of a 1989 Ford Probe, and states only that:
Customer says:
• the vehicle caught on fire under the dashboard of the vehicle while he was at a stop sign
• The inside of the vehicle filled with smoke and the vehicle burst in to flames in the interior of the vehicle
• Under the hood of the vehicle is fine. Was not touched by fire
(Pl.Ex. 7, appended to PI. Supp. Br. Dated January 20, 2000.)
Beginning with a collection of 14,000 MORS complaints involving Ford vehicles with Fox ignition switches, Charles Adams & Associates — litigation consultants retained on behalf of plaintiffs — reviewed the collection for MORS complaints coded as “visible flame”, “smoke”, or “burnt” in the passenger cabin. (Affidavit of Charles Adams, Pl.Ex. A at K 9.) From this initial set of 14,000, Adams eliminated those complaints which he concluded could not involve ignition switch fires, because, for example, the report stated that the fire originated in the engine compartment or backseat area. (Adams Dep. at 73:20 to 74:5.) If the MORS entry mentioned an ignition switch, then Adams concluded that the alleged fire was “very likely” caused by one of the subject switches. (Id. at 79:2.) If the MORS report did not specify ■ the origin of the fire, but only said it started in the steering column, Adams classified such fires as “more probable than not as being ... ignition switch [related].” (Id. at 79:3-6.)
To this list of potential ignition switch fires derived from the MORS II reports, plaintiffs added all incidents in Ford’s “ASES” database6 reported as “alleged steering column fires”, together with the vehicles of insurance policyholders involved in the State Farm and CSAA subrogation claims, and the Snodgrass named plaintiffs. Apparently, the MORS II reports comprise approximately 47.3% of the total claims. (PI. Comments on MORS II Sample at 3.) Based on the data just described, the plaintiffs arrived at a figure which has been reduced through a series of eliminations to a putative class of 6,098. (See Pl.’s Second Stage Proceedings Proposals Dated Jan. 21, 2000 at 6 & n. 5.)
Plaintiffs have performed an elaborate statistical analysis of the above-described database aimed at creating a statistical presumption that any fire in a class vehicle was caused by one of the subject switches. In this analysis, plaintiffs’ expert Dr. Jack Moshman, a statistician, examined the Adams-created fire database and broke down the models/years of vehicles included therein, limiting the class to those vehicles to those model/years having statistically homogeneous fire rates of between 2.702 and 3.304 per 10,000 vehicles. (PI. Br. at 17.) Plaintiffs have further refined this large class to arrive at two subclasses: Subclass I, which includes seventeen models with an average fire rate of 2.386 per 10,000,7 and Subclass II, a three *494model group with a fire-rate of 6.210 per 10,000.8 (PI. Revised Proposed Class Definition at 6-7.) Under this analysis, therefore, the proposed subclasses of vehicle owners include those who own vehicles, which, according to Moshman, have roughly similar chances of catching on fire. (Id.) By plaintiffs’ view, Moshman’s analysis makes it possible to have two large subclasses of vehicle owners who have in common a similar statistical likelihood of ignition switch fires.
Moshman’s statistical analysis has been attacked by defendants, who claim that there exist no good grounds for plaintiffs’ conclusions, chiefly because Moshman’s report is based on untested allegations rather than actual ignition switch fires. (See Def. Surre-ply Br. at 5.) Defendants also have employed their own statistical expert, Dr. William E. Wecker, who states that the Adams-created database is unreliable and overstates the number of ignition switch-related incidents, and that Moshman’s analysis is therefore a parsing of non-probative data. (Wecker Aff. at H114-6, Def. Surreply Ex. 14.) Wecker also contradicts Moshman’s finding of statistical homogeneity within the subclasses, asserting that the reported incident rates vary greatly among the model/years. Further, by Wecker’s view, Moshman’s conclusions are undermined by the fact that he and plaintiffs’ counsel selected the model/years to be included in the subclasses in order to achieve the desired result, a technique which Wecker claims runs counter to accepted statistical norms. (Id. at H1Í15-17 & 24.) Wecker concludes that 'because Moshman’s tstudy is premised on unreliable data, and because his proposed subclasses do not include vehicles with statistically homogeneous fire rates, the worth of plaintiffs statistical proffer is doubtful at best. (Id. at 1117.)
Even beyond the concerns raised by defendants, this Court has concerns over the reliability of the Adams/MORS II database as a blueprint for defining classes and subclasses, as well as a basis for proving the homogeneity of the claims within a subclass. At the January 7, 2000 hearing in this case, defendants argued that the database included too many unverified and vague allegations to be reliable.9 Consequently, the Court questioned how many of the complaints included in the database were backed by substantiation of actual ignition switch fires, and asked the parties to submit a random sample of the MORS II reports so that the Court could form its own impressions as to the quality of the information contained in the MORS II reports. On January 14, 2000, the parties responded by submitting 72 randomly selected MORS II reports — the first report identified on every 10th page of plaintiffs’ database.
The randomly selected reports show that, of the 72 reports selected, 19 reports (26%) of the random subset do not even mention the words “ignition switch”. Instead, these reports are ambiguous and allege that fires started in various areas of the vehicle, such as the “inside the cab”, or within the steering column, the dashboard, or the instrument panel. (Def. Ltr. Dated Jan. 20, 2000 at 2.) One of the selected reports, that of Mr. Horn, states that the vehicle owner had replaced the ignition switch several times before the subject fire. Thus, it is unclear in Mr. Horn’s case whether his car was even equipped with a Fox switch when it caught fire. (Id.) Defendants argue that the large number of ambiguous reports shows that the MORS II entries are not probative of ignition switch fires and that the random sample refutes plaintiffs’ claim that the database is sufficiently reliable to serve as prima facie *495proof that the proposed subclasses share in common that they had an ignition switch fire.
Plaintiffs have responded to defendants’ critique of the random sample by contacting twelve of the nineteen individual vehicle owners (the others apparently were unavailable) whose reports Ford cited as ambiguous and non-probative. (PI. Response Dated Jan. 27, 2000 at 2-6.)- Plaintiffs’ counsel argues that in many instances, an interview of the vehicle owner led to additional information that, once reviewed by plaintiffs’ expert Mr. Hagerty, provides additional support for the theory that the alleged fire originated with the ignition switch. (Id. at pt. 2, 4-7.) In other instances, plaintiffs claim that once contacted, the complaining owner “verified” that the fire was caused by the ignition switch, either through his own observations or those of service personnel. (Id. at pt.‘ 2-4, 6, & 8-11.) Apparently, this verification is either based on lay observation or the statements of a. third party. Were this case to go to trial, therefore, such verification would likely be the subject of vigorous cross-examination.
Based on the foregoing, the Court finds that, while plaintiffs have arguably rebutted defendants’ challenge to several of the MORS II reports, there remain countless other entries that will be subjected to similar scrutiny. As exemplified by the debate on these contested reports in the 72-entry subset, further record development will be required of many, if not all of the MORS II reports. Defendants undoubtedly will refuse to concede that the subject switches caused the damages alleged in every ambiguously worded report, thus necessitating further interviews and expert witness review. In sum, it remains doubtful that all of the MORS II reports included in the database actually represent actual ignition switch fires.
The Moshman analysis of the incidents reported in the underlying database is thus not a suitable basis upon which subclasses for trial could be defined because there is not a reliable basis upon which to conclude that all members of the identified subclasses have in fact sustained an ignition switch fire. Based on the foregoing, the Court questions whether plaintiffs’ database is sufficiently reliable to serve even as a prima facie showing that certain vehicle owners have in common that they suffered ignition switch fires.10
3. Plaintiffs Have Not Shown That Class Action is Superior to other Available Methods for the Fair and Efficient Adjudication of This Controversy
This matter will not be certified for class action treatment for the additional reason that the Court is not satisfied that plaintiffs have set forth a workable plan for resolution of the abundant individual issues in this case. For reasons discussed above, the Court found that common issues of law and fact do not predominate over individual concerns in this matter. Moreover, this Court now finds that plaintiffs’ proposed a two-stage trial for litigating vehicle owners’ fire-related claims is unworkable.
As discussed above, plaintiffs have proposed a two-step litigation plan. In the first stage, the Court would hold a “class trial” wherein plaintiffs would present evidence tending to show defendants sold vehicles equipped with defective ignition switches, in violation of implied warranties and deceptive trade practice laws. In the second stage, defendants would be given the opportunity to address the issue of causation, and rebut the statically created presumption that individual plaintiffs’ vehicle were actually damaged by an ignition switch fire.
Defendants vigorously criticize plaintiffs’ proposed bifurcated trial, arguing that a method of “mini-trials” proposed for stage two would be inefficient for the Court and unfair to the defendants. The Court agrees.
*496In general, it is desirable to litigate similar, related claims in one forum. On the other hand, efficiency-based arguments on this point begin to lose their force as one contemplates the post-certification trial of plaintiffs’ claims. As discussed previously, there are numerous factual issues in this case that cannot be adjudicated on a class-wide basis, and each individual vehicle owner and his/her witnesses would be subject to cross-examination on the particular facts and circumstances surrounding that vehicle owner’s purported ignition switch fire. In addition, the second stage of the trial would necessarily involve individual consideration of each of the predictably numerous disputed claims. For example, even if the presumption of causation were available from plaintiffs’ successful pursuit of the first phase trials for the various classes and subclasses, individual proofs would have to tie that presumption to the facts of the individual claims, and defense experts, or perhaps fact witnesses, could be expected to testify that the fire did not start in the relevant area of the car, or that it had a separate cause. Thus, each mini-trial could be expected to involve the direct and cross-examination of at least three witnesses: the vehicle owner and a fact witness for plaintiff (such as the mechanic or dealer who diagnosed the’fire’s origin) to overcome problems of hearsay, and one or two experts. The geographical logistics of such a trial would be formidable, and perhaps prohibitive, for individual claimants from around the country to travel to this Court to try their claims.
Even assuming that the number of contested claims could — as- plaintiff suggests— through discovery be narrowed to 3,000, and that the individual “rebuttal trials” were to average just four hours each (a conservative estimate), the second phase of the trial would take approximately twelve years of judicial time.11 This would not serve the public well, nor would it provide relatively prompt remediation of the valid claims that exist.
Upon consideration of the realities of plaintiffs’ proposed system of mini-trials following the establishment of a prima facie case of causation, the Court holds that class wide adjudication is not a superior method of adjudication. Plaintiffs’ claims do not present a case where common issues predominate. Faced with the realities of plaintiffs’ proposed litigation plan, the Court finds that individual litigation of plaintiffs’ claims is a superior plan of action than the class action alternative facing the Court: a years-long trial process of rebutting a statistical presumption.12
CONCLUSION
The Court declines to certify these eases as class actions under Rule 23(b)(3). Certification under this provision is inappropriate here because plaintiffs have failed to demonstrate that, in these cases, “questions of law or fact common to the members of the class predominate over any questions affecting only individual members, and that a class action is superior to other available methods for the fair and efficient adjudication of the controversy.” Accordingly, plaintiffs’ petition for class certification is denied, and the representative plaintiffs’ class-based claims will be dismissed without prejudice to the named plaintiffs’ rights to pursue their individual claims remaining in this case.
APPENDIX
(Plaintiffs Proposed Class Definitions)
(a) Implied Warranty (Fitness for Ordinary Purposes) Subclass
All owners of Ford Class Vehicles that have burned or been damaged by fire because of a defective Fox ignition switch during the Class Period and who have notified Ford of the breach either individually or through the classwide notice of breach which this and previous Complaints and a demand letter dated January 16, 1998 constitute un*497der 15 U.S.C. § 2310(e) and/or UCC § 2-607(3)(a), with this breach of implied warranty claim asserted under 15 U.S.C. § 2310(d)(1) or alternatively, to any extent that the Magnuson-Moss Act is found inapplicable, under UCC § 2-314(2)(c), which provides that for goods to be merchantable they must be fit for the ordinary purposes for which such goods are used.
Excluded from this Subclass are:
(1) personal injury claims;
(2) persons who prior to final judgment have timely opted out or filed separate nonclass legal actions against Ford asserting similar claims;
(3) defendant Ford, including any parent, subsidiary, affiliate, or controlled person of Ford and its officers, directors, agents, or employees;
(4) purchasers of the Class Vehicles in states that at the time of final judgment require vertical privity to pursue such implied warranty claims, such states presently believed to include, e.g., (a) Alabama, Arizona, Hawaii, Idaho, New York, Florida, Wisconsin, Kentucky, Indiana, Washington, and Iowa (as to all owners) or (b) California, Georgia, Illinois, North Carolina, Vermont, and Texas (as to owners who were not the original purchasers and purchased after expiration of Ford’s written warranty);
(5) purchasers other than “consumers” in Kansas, Oregon, Rhode Island, and Connecticut; and
(6) purchasers in Mississippi whose Class Vehicles burned after 36,000 miles.
(b) Deceptive Trade Practice Subclass
All owners of Ford Class Vehicles which have burned because of defective ignition switches during the Class Period, with this claim asserted under the substantively similar state statutes (excluding Iowa, which has no private cause of action) that either (1) prohibit unfair or deceptive trade practices generally (e.g., Alaska, Connecticut, Florida, Georgia, Hawaii, Kentucky, Louisiana (La. Civ.Code Art. 2545 & 2534), Maine, Massachusetts, Missouri, Montana, Nebraska, New Hampshire, New Mexico, New York, North Carolina, Ohio, Rhode Island, South Carolina, Tennessee, Texas, Vermont, Washington, West Virginia, Wyoming, Arizona, Delaware, Illinois, New Jersey, and North Dakota) and/or (2) prohibit included but more specifically defined conduct such as representing that goods have characteristics, uses, benefits, or qualities that they do not have or that goods are of a particular standard, quality, or grade if they are of another (e.g., California, District of Columbia, Idaho, Maryland, Michigan, Minnesota, Mississippi, Pennsylvania, and Virginia), including the following additional subgroups: (a) those states that require a showing of scienter (e.g., Arkansas, Colorado, Indiana,* Kansas, Nevada, Oklahoma, Oregon, South Dakota, Utah, and Wyoming), (b) those states that require the showing of some “aggravating factor” other than scienter (e.g., North Carolina and New Jersey), and (c) those states that require proof of justifiable individual reliance, which can be rebuttably presumed from the materiality of the “fitness for ordinary purposes” misrepresentations (e.g., Arizona,** Georgia, and Wyoming).
Excluded from this Subclass are:
(1) the first three groups excluded from the Implied Warranty Fitness Subclass;
(2) persons other than “consumers” in states whose deceptive trade practices statutes are limited to consumers who purchase for personal, family, or household purposes (e.g., Alabama, California, District of Columbia, Georgia, Hawaii, Kansas (except sole proprietors are not excluded), Maine, Maryland (except agricultural purchasers (e.g., farmers) are not excluded), Michigan, Mississippi, Missouri, Montana, Ohio, Pennsylvania, Rhode Island, Tennessee, Texas (excluding only business entities with $25 million or more assets), Utah (except that purchases related to a business opportunity are not excluded), Virgi*498nia, West Virginia (except agricultural purchasers are not excluded), and Wyoming);
(3) persons whose claims would be time-barred, as of the date of the earliest class action Complaint filed as to this matter, under state deceptive trade practices statutes or other limitations doctrines that have repose provisions not permitting tolling of the limitations period due to fraudulent concealment, such as one year from discovery but no more than four years from the date of sale (e.g., Tennessee, New Hampshire, Ohio, and Wyoming), and
(4) persons who pursuant to this Court’s May 14, 1999 ruling have no “fitness for the ordinary purposes” predicate representations because their states require privity for implied warranty claims (see 1111 (a)(4) & (a)(6)).
The Ford Class Vehicles are defined by model and model year as follows:
Ford Class Vehicles by Sales, Ignition Fires, and Fire Rates per 10,000 vehicles
Subclass I
MODEL YEARS SALES FIRES * FIRE RATE*
Bronco 1984-91 399,707 179 4.478
GrMar** 1984-89 750,946 307 4.080
F-Series 1985, 87-91 3,129,236 1104 3.528
CrVic** 1984-89 823,121 238 2.891
Escort 1984, 86-90 2,024,269 542 2.678
Thunderbird,*** 1984-93 1,095,268 294 2.680
Tempo **** 1985-93 1,904,246 433 2.270
Cougar*** 1984-93 859,159 172 2.002
Topaz**** 1984-93 684,423 134 1.958
Mark WI 1985-89 125,517 15 1.195
Explorer 1991-93 736,862 82 1.113
Bronco 11 1988 147,779 16 1.083
Continental 1984-87 92,297 9 0.975
Lynx 1986 74,565 7 0.939
Econoline 1985-88, 92-93 1,013,267 94 0.928
Ranger 1985-93 2,178,028 190 0.872
Unknown
TOTAL 16,038,690 3,826 2.386
Subclass II
MODEL YEARS SALES FIRES FIRE RATE
TownCar 1984-89 707,896 564 7.967
Aerostar 1986-91 1,011,336 632 6.249
Mustang 1986-93 1,000,532 493 4.927
TOTAL 2,719,764 6.210
. Originally, there were multiple cases proceeding under the caption of In re Ford Motor Co. Ignition Switch Products Liability Litigation. The first of these, known as the "MDL cases”, involved a nationwide consolidation of plaintiffs who own Ford vehicles of the subject 1984-93 model years, but whose vehicles experienced no problem related to the ignition switch. By Stipulation and Order dated August 31, 1999, the MDL plaintiffs' cases were dismissed. Accordingly, the present Opinion affects only the putative plaintiffs who allegedly suffered smoke or fire damage to their vehicles within the Snod-grass case.
. This group of plaintiffs originally was known collectively as the "Wilks plaintiffs”, a nomination which was changed in the superseding Third ' Amended Complaint after it came to light that the original lead plaintiff, Michael Wilks, purposefully set fire to the interior of his Ford motor vehicle.
. After briefing was complete on this class certification motion, the Court nonetheless granted plaintiffs' motion to refine and clarify the proposed class definition. Because this definition is lengthy, it is set forth in full in the Appendix to this Opinion.
. It is conceivable that in another class action suit a group of plaintiffs might be allowed to circumvent basic individual causation requirements simply by showing that the subject defective product always caused a certain type of harm. One example might involve a class of plaintiffs alleging that they all suffered lung damage uniquely caused by asbestos fibers. No such singular harm is alleged here. As discussed further below, plaintiffs proffer statistical evidence *491that — if preliminary factual assertions of the vehicle owners are credited — ignition switch fires are a very rare event among the class vehicles, but that such fires show up with somewhat greater incidence among certain makes and models.
. Plaintiffs point to Ford Motor Co. v. Sheldon, 965 S.W.2d 65, 69, 72-73 (Tex.Ct.App.1998), and R.J. Reynolds Tobacco Co. v. Engle, 672 So.2d 39 (Fla.Ct.App.), rev. denied, 682 So.2d 1100 (Fla. 1996), as persuasive authority for the certification of a class in this case under Rule 23(b)(3). Sheldon is factually and legally distinguishable from this case because it did not involve the_ use of statistics to create a rebuttable presumption of causation. In any event, the lower court’s opinion in Sheldon was recently reversed by the Texas Supreme Court. See Ford Motor Company v. Sheldon, 22 S.W.3d 444 (Tex.2000) (rejecting proposed plan of mini-hearings to determine individual causation). The Court also finds that the three-page opinion in RJ. Reynolds offers no real assistance to the determination of Rule 23(b)(3) certification because it is devoid of a thorough analysis of the requirements which must be satisfied before a class is certified.
. The ASES database is a collection of Ford ignition switch fires compiled by Ford’s Automotive Safety Office and the Office of General Counsel.
. This proposed subclass would include owners of the following model/years: 1984-91 Ford Broncos, 1984-89 Mercury Grand Marquis, 1985 & 1987-91 Ford F-Series pickup trucks, 1984-89 Ford Crown Victorias, 1984 & 1986-90 Ford Escorts, 1984-93 Ford Thunderbirds, 1985-93 Ford Tempos, 1984-93 Mercury Cougars, 1984-93 Mercury Topaz, 1985-93 Lincoln Mark VIIs, 1991-93 Ford Explorers, 1988 Ford Bronco IIs, *4941984- 87 Lincoln Continentals, 1986 Mercury Lynx, 1985-88 & 92-93 Ford Econoline Vans, 1985- 93 Ford Ranger pickup trucks.
. The models included are 1984-89 Lincoln Towncars, 1986-91 Ford Aerostar vans, and 1986- 93 Ford Mustangs.
. Defendants also argued at this hearing that a large percentage of the MORS II reports were generated between February and June 1996, a period during which there was substantial media coverage of Ford's ignition switch recall and the filing of plaintiffs’ class action lawsuit. Accordingly, defendants argue that the reliability of the MORS II therefore may be tainted by the sensationalism surrounding the media reports, in which persons experiencing any type of fire would tend to self-report the fire as having an ignition switch origin.
. Examination of plaintiffs’ class definitions also shows that the class period is nowhere defined. Accordingly, plaintiffs' proposed classes and subclasses do not set forth when the putative plaintiffs allegedly suffered harm caused by faulty ignition switches. Without a time frame, it is conceivable that plaintiffs could continue to be added up to the day of trial. This uncertainty as to the temporal breadth of the class period would only make it more difficult to ascertain who are proper members of the proposed class, and to apply the proposed class definitions. The impact of time in individual cases is also crucial due to the statute of limitations defenses raised.
. This calculation is based on this Court’s average of 1000 hours of bench time per year. This lengthy trial would also necessarily preclude resolution of most of the Court’s normal caseload of 400 civil and 80 criminal cases filed per year during this period.
. The Court finds further support for the conclusion that individual litigation is a superior method of adjudication in that there already have been hundreds of individual ignition switch claim suits filed nationwide.
Indiana is excluded from this Subclass pursuant to H (b)(4) below.
Arizona is excluded from this Subclass pursuant to 11 (b)(4) below.
These are the fires/rates for U.S. vehicles only, as derived from Plaintiffs’ Database (Exh. D to Plaintiffs’ Reply Brief). The fire reports there, and in Tables 3 and 4 to the Second Affidavit of Jack Moshman (Plaintiffs’ Reply Brief Exh. C), also include Canadian vehicles, so the U.S. only fires/rates will differ from the U.S. and Canadian combined fires/rates.
The Grand Marquis and Crown Victorias are essentially the same vehicles with different model names.
The Thunderbirds and Cougars are essentially the same vehicles with different model names.
'j'gjnpo ancj T0paz are essentially the same vehicles with different model names.