dissenting.
Assuming, as the court does, that only-two pills match the description given by Galvin’s mother of the small, round, white, cross-scored pill that she took, Galvin has proffered evidence from which a reasonable jury could find that she was more probably exposed to diethylstilbestrol (“DES”) manufactured by Lilly than the alternative manufactured by Marsh Parker. The events underlying Galvin’s complaint occurred more than four decades ago. Hence, it is hardly surprising that evidence such as pharmacy business records from 1964-65 is unavailable to demonstrate with certainty which manufacturer produced the DES giving rise to Galvin’s claims for relief. But Galvin need not produce at trial evidence of substantial certainty, and the court errs by holding her to a standard of proof greater than is required under the relevant State law, to wit: “more probable that the event was caused by the defendant than that it was not.”
In a universe of two pills, any evidence tending to implicate the Lilly pill suffices to make that pill the more probable cause of Galvin’s injuries. Galvin has proffered multiple pieces of evidence that, when viewed together as a mosaic, support the reasonable inference that among the small, round, white, cross-scored pills, her mother’s DES pill was more probably manufactured by Lilly than by Marsh Parker. In response to Lilly’s alternative theory that the Marsh Parker pill, which also matches Galvin’s mother’s description, may have caused Galvin’s injuries, the court initially acknowledges that, as the non-moving party, Galvin “need not disprove that alternative.” Op. at 1031. This, however, is what the court requires her to do. Thus, having rejected the probative value of each piece of her evidence viewed in isolation, the court observes that “Galvin simply has not provided the evidence — market share data for the Marsh Parker pill in a relevant market — necessary for a reasonable juror to conclude that a Lilly pill was probably the cause of her injury.” Op. at 1035. But Galvin is not relying solely on statistical evidence. See Op. at 1034-36. Once that flaw in the court’s analysis is removed the “mosaic” that emerges from Galvin’s evidence, according her as we must the benefit of all reasonable inferences, suffices to defeat Lilly’s claim that it is entitled to judgment as a matter of law, for at the summary judgment stage Galvin is not required to “produce evidence in a form that would be admissible at trial.” Celotex Corp. v. Catrett, 477 U.S. 317, 324, 106 S.Ct. 2548, 91 L.Ed.2d 265 (1986).
I.
“Under Rule 56(c), summary judgment is proper ‘if the pleadings, depositions, answers to interrogatories, and admissions on file, together with the affidavits, if any, show that there is no genuine issue as to any material fact and that the moving party is entitled to a judgment as a matter of law.’ ” Id. at 322, 106 S.Ct. 2548 (quoting Fed. R. Civ. P. 56(c)). In other words, the “mosaic” approach, in contrast with the court’s breaking up of Galvin’s evidence under two theories of analysis, see Op. at 1032, is consistent with the Supreme Court’s instruction on summary judgment. Thus,
the plain language of Rule 56(c) mandates the entry of summary judgment, after adequate time for discovery and upon motion, against a party who fails to make a showing sufficient to establish the existence of an element essential to that party’s case, and on which that party will bear the burden of proof at trial.
Id. Galvin’s proffer of evidence is not so deficient. At the summary judgment *1038stage, “[t]he court’s function is not to try-disputed issues of fact, but only to ascertain whether such an issue is present, and any doubt on that score is to be resolved against the movant.” Abraham v. Graphic Arts Int’l Union, 660 F.2d 811, 814 (D.C.Cir.1981) (footnote omitted). Where any such dispute exists after giving the non-movant “the most favorable view of the record,” Exxon Corp. v. FTC, 663 F.2d 120, 126 (D.C.Cir.1980); see Celotex, 477 U.S. at 330 n. 2, 106 S.Ct. 2548, summary judgment must be denied because “[t]rial by affidavit is no substitute for trial by jury which so long has been the hallmark of ‘even handed justice.’ ” Poller v. CBS, Inc., 368 U.S. 464, 473, 82 S.Ct. 486, 7 L.Ed.2d 458 (1962). At the summary judgment stage, Galvin’s evidence “is to be believed, and all justifiable inferences are to be drawn in [her] favor.” Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 255, 106 S.Ct. 2505, 91 L.Ed.2d 202 (1986). Additionally, she is “not require[d] ... to discredit every conceivable alternative theory of causation.” Shields v. Eli Lilly & Co., 895 F.2d 1463, 1465 (D.C.Cir.1990).
To determine whether a reasonable juror could find in Galvin’s favor based on her proffered evidence, the court must apply the substantive law of Kansas. See Erie R.R. Co. v. Tompkins, 304 U.S. 64, 78, 58 S.Ct. 817, 82 L.Ed. 1188 (1938). To prevail under Kansas law, a plaintiff must “‘introduce evidence from which reasonable persons may conclude that it is more probable that the event was caused by the defendant than that it was not.’ ” Yount v. Deibert, 282 Kan. 619, 147 P.3d 1065, 1073 (2006) (quoting Peossee & Keeton on Toets § 41, at 269-70 (5th ed.1984)). Kansas law does not hold the plaintiff “to a highly detailed burden of proof,” id. at 1072, but simply requires “evidence which affords a reasonable basis for the conclusion that it is more likely than not that the conduct of the defendant was a cause in fact of the result,” id. (quoting PROSSER & Keeton § 41, at 269-70). Thus, “triers of fact are allowed to draw upon ordinary human experience with regard to the probabilities of the case.” Id. at 1073. To create a material issue of disputed fact, “circumstantial evidence in a civil case need not rise to that degree of certainty which will exclude every reasonable conclusion other than the conclusion sought to be established.” Len-herr v. NRM Corp., 504 F.Supp. 165, 168 (D.Kan.1980). At trial, a jury is entitled to make a reasonable inference from the evidence even if “some other inference equally reasonable might be drawn therefrom.” Arterburn v. St. Joseph Hosp. & Rehabilitation Ctr., 220 Kan. 57, 551 P.2d 886, 890 (1976). In a products liability case, the evidence “must justify an inference of probability as distinguished from mere possibility.” Mays v. Ciba-Geigy Corp., 233 Kan. 38, 661 P.2d 348, 360 (1983).
II.
The court declines to consider Lilly’s contention that other manufacturers of DES may have made a pill that matches Galvin’s mother’s description because “Gal-vin has not eliminated the Marsh Parker pill from consideration.” Op. at 1032. Galvin is not required to do so in order to defeat summary judgment, see Part I, although the court repeatedly suggests that she must, see Op. at 1032, 1033-34, 1035. Assuming, as the court does, that only two DES pills meet the description given by Galvin’s mother, Op. at 1032, Lilly has not offered any evidence suggesting that Gal-vin was probably exposed to the Marsh Parker pill, and the mosaic based on the evidence proffered by Galvin suffices to show that it is more probable that she was exposed to the Lilly pill. Even assuming no individual piece of evidence proffered by Galvin would alone suffice to meet her burden, she has offered evidence that begins with her mother’s identification of the pill and her hospital records, continues *1039with affidavits from several pharmacists— including an expert — concerning Lilly’s nationwide agreements and stocking practices, and with Lilly’s Warehousing and Distribution Service Agreement, and concludes with the druggists’ reference books — the Physician’s Desk Reference (“PDR”) and the American Druggist Blue Book (“Blue Book”) — and a national market-share matrix. The conclusion that Galvin has met her burden follows without deciding whether the district court properly denied Galvin’s request to file supplemental affidavits, see Op. at 1030, or whether this court errs by allowing Lilly to avoid the result of a disavowing affidavit, see Op. at 1032 n. 1.
Galvin’s mother gave a four-point description of the pill she took in 1964-1965 and hospital notes indicate that she was prescribed “Stilb 25 mg,” which supports the reasonable inference that her prescription also indicated a generic name only. The initial affidavit of pharmacist Bill Wal-trip provides information about the business practices with regard to DES at the drug store in Pittsburg, Kansas from which Galvin’s mother obtained her DES pills; no records exist for the pharmacy from 1964-65. Waltrip states that, “[i]n the 1960’s, Ash Drug Store only stocked and dispensed the [Lilly] brand of diethyl-stilbestrol.” Waltrip asserts, “[b]ased on [his] personal observation of the habit and custom” of the drug store in the 1960’s, that the Lilly brand of DES would have been dispensed to a “woman who came into the pharmacy with a prescription for ‘DES’, ‘stilbestrol’, or ‘diethylstilbestrol.’ ” He also states that the prescribing obstetrician for Galvin’s mother practiced in Ft. Scott, Kansas, and that the Ash Drug Store filled and dispensed many of his patients’ prescriptions. Finally, Waltrip notes that the drug store ordered DES from Pennington Wholesale Drug located in Joplin, Missouri, which was a Lilly wholesaler.
Although Waltrip did not begin working at the Ash Drug Store until 1967, his affidavit establishes that at that time (1) Lilly DES was available in Kansas through a Lilly wholesaler, (2) it was sold at the same drug store where Galvin’s mother’s prescription was filled, and (3) no other DES was available at that drug store. The court rejects the probative value of Waltrip’s testimony about stocking practices on the ground that it is adopting “a neutral posture, inferring neither change or continuity at the drugstore” since 1964-65. Op. at 1036; see id. at 1036 n. 4 (quoting Fed. R. Evid. 406). On summary judgment, however, the court must give the non-moving party the benefit of all reasonable inferences. See Anderson, 477 U.S. at 255, 106 S.Ct. 2505. A reasonable jury could infer, when Galvin’s other evidence is considered, that Waltrip’s affidavit makes it more probable that Lilly DES was sold in the same drug store in 1964-65. The two-year period affects the strength of the inference that may reasonably be drawn; it does not eliminate its relevance. See Fed. R. Evid. 401 advisory committee’s note.
Bolstering the probative force of Wal-trip’s affidavit is the Lilly Warehousing and Distribution Service Agreement, which states that the wholesaler shall promote Lilly products and not give preference to any other brand when no brand is specified. Galvin’s mother’s hospital records indicate that she was prescribed “Stilb 25 mg” without specifying a brand. Op. at 1035. Additionally, the affidavits of Harold B. Sparr and Philip J. Cafferty are relevant because of their personal familiarity with Lilly’s literature, its nationwide wholesale strategy in the 1950s and 1960s, and the nature of its agreements with wholesalers throughout the United States. Both Sparr and Cafferty state that no manufacturer other than Lilly had a round, white, cross-scored DES pill with*1040out any other markings. Their evidence further bolsters the probative force of Waltrip’s affidavit.
Sparr states that he was familiar with Lilly publications and its marketing and stocking practices in the 1950s and 1960s. He also states that he reviewed the sworn statements of over 105 pharmacists regarding the prevalence and availability of Lilly DES products in their stores in the 1950s and 1960s. Sparr is a registered pharmacist in Massachusetts, New York, and California; he holds a Masters in Health Care Management from Pacific Western University and has taught pharmacy at the Massachusetts College of Pharmacy and Northeastern; he has over forty-seven years of experience as a practicing pharmacist, beginning in the mid-1950s; he has served as President of the Massachusetts Board of Registration in Pharmacy and was named Pharmacist of the Year in 1995 by the Massachusetts Pharmacy Association; and he was twice a delegate to the United States Pharmacopoeia.
According to Sparr, in the 1950s and 1960s “Lilly was the leading pharmaceutical manufacturer in America [] with top market popularity because of its reputation, quality, control, efficiency of inventory and distribution through wholesalers.” He notes, among other things, that at that time “Lilly was the only major drug house that employed licensed pharmacists as de-tailmen” to restock drug stores with Lilly products and also employed unique stocking practices, enabling retail pharmacists to save money. Even though Sparr’s pharmaceutical practice and teaching experiences were in New England, his familiarity with Lilly literature that was distributed nationwide and its nationwide stocking practices and its unique detailing practices lends support to the reasonable inference that the Ash Drug Store in Pittsburg, Kansas — which, according to Waltrip, obtained its DES from a Lilly wholesaler in 1967- — dispensed the Lilly DES pill to Gal-vin’s mother. The affidavit of Philip Caf-ferty, a pharmacist since 1961, is to the same effect, noting also that for nineteen years, beginning in 1965, he was employed by Lilly as a professional representative or detailman and district manager in New England.
While it is possible, in the sense that anything is possible, that Lilly functioned in a different manner in Kansas in 1964-65, Sparr and Cafferty are familiar with Lilly’s nationwide practices and top popularity in the relevant years. The court rejects their evidence because Sparr’s and Cafferty’s experiences were “almost exclusively limited to New England.” Op. at 1033. This overlooks not only that Lilly also relies on the testimony of a pharmacist who worked solely in Massachusetts, Op. at 1032 n. 1, but also that Sparr and Cafferty have undisputed knowledge of Lilly’s nationwide practices when Galvin’s mother was taking DES. By proffering this evidence, Galvin has offered further support for the reasonable inference that a jury could draw that it was more probable than not that her mother took Lilly’s DES pill.
Other support for the probative force of the inference about 1964-65 from Waltrip’s affidavit is found in three survey-type references. The first two references are national catalogues listing brands of drugs. The Physician’s Desk Reference (“PDR”), which Galvin asserts (and Lilly does not contest, a point overlooked by the court, Op. at 1034 & n. 2), is the “Bible of [djrugs” that was “on the [d]esk of [ejvery [djoctor in America,” Appellant’s Br. at 17, lists Lilly as the only brand of DES in the Twentieth Edition published in 1965. A second reference, proffered by Lilly, is the 1964-65 edition of the American Druggist Blue Book, which includes the catalogues *1041of seventy-four manufacturers and describes itself as listing the latest products and price changes. The Blue Book also does not list Marsh Parker among the brands of DES.
The court dismisses the probative value of the PDR on the ground that Galvin concedes that it is not an exhaustive list of DES manufacturers and imposes on Galvin a unique burden to show the compilation methodologies used in the PDR and the Blue Book even though Lilly does not dispute Galvin’s characterization of the PDR and Lilly proffered the Blue Book. See Op. at 1034 & n. 2. The court does not address the fact that Marsh Parker does not appear in the Blue Book other than to impose a compilation burden, Op. at 1034 n. 2, instead finding that the listing of Marsh Parker in the Red Book, also proffered by Lilly, is conclusive. See Op. at 1033-34. No doubt the non-exhaustive nature of the listing means that the PDR is not complete proof. But Galvin, not Lilly, is entitled on summary judgment to all reasonable inferences and a reasonable juror could find that it is more probable that the more prominent and more widely available drug would have been listed than the less prominent drug. From the absence of Marsh Parker from these two listings a reasonable juror could infer that Marsh Parker was not a major producer of DES during the relevant period. Although the Red Book for 1964 and 1965 lists both Lilly and Marsh Parker, a reasonable juror could accept that Marsh Parker was on the market in 1964-65 and still conclude that it is more probable than not that Galvin was exposed to Lilly DES. To the degree that the Red Book supports the contrary inference — that Galvin was exposed to Marsh Parker DES — under Kansas law, Galvin’s burden at trial does not require her to exclude other reasonable inferences, see Arterburn, 551 P.2d at 890, much less offer mathematical proof of causation, see Yount, 147 P.3d at 1072-73.
The third reference is a national market share matrix developed in connection with litigation in New York. See Hymoivitz v. Eli Lilly & Co., 73 N.Y.2d 487, 541 N.Y.S.2d 941, 539 N.E.2d 1069, 1078 (1989). The matrix indicates that Lilly had a significant percentage — approximately 28% — of the national market in 1964-65. It does not list Marsh Parker at all. That summary judgment was granted to some drug manufacturers and that Gal-vin proffered only one page of the matrix, see Op. at 1033, are red herrings, for Lilly was not among the manufacturers granted summary judgment and the single page is where Marsh Parker would have been listed in alphabetical order. The court also dismisses the probative value of this evidence because it shows that Lilly had less than 30% of the market. Op. at 1035. Contrary to the court’s implication, Op. at 1034-35, Galvin is not relying on statistical evidence “by itself’ to defeat summary judgment and, consequently, the court’s expression of doubt about the probativeness of statistical evidence, see Op. at 1034-35, misses the mark. But even accepting Lilly’s bald assertion that the matrix leaves 44% of the market share unaccounted for, a reasonable juror could doubt, in view of Galvin’s evidence, that Marsh Parker was a dominant market force accounting for more than 28% of the remaining 44% but was somehow excluded from the PDR and the Blue Book and the matrix. Contrary to the court’s conclusion, a reasonable juror need not believe that these reference materials have no probative value at all. Rather, they lend support to the reasonable inference that in 1964-65 Lilly DES was more prevalent than Marsh Parker DES, thus making it more probable that Galvin was exposed to Lilly’s pill than Marsh Parker’s.
In reaching a contrary conclusion about the probative force of the mosaic of evidence that Galvin has proffered, the court *1042appears to suggest that Lilly is entitled to summary judgment because Galvin may not have proven that it is certain that she was exposed to Lilly’s product or that it is significantly more probable that she was exposed to Lilly’s product than to Marsh Parker’s. See Op. at 1032, 1032, 1033-34, 1035. This is not her burden as the non-moving party. See Anderson, 477 U.S. at 248, 106 S.Ct. 2505; see also Yount, 147 P.3d at 1072-73. At this stage of the proceedings, Galvin need only adduce evidence from which a reasonable juror could conclude that it is more probable — even if only slightly more probable — that she was exposed to Lilly’s pill even if there is another reasonable inference. She has done so. The court’s contrary conclusion stems from its failure to adhere to the summary judgment standard, see, e.g., Op. at 1034 n. 2, 1036 n. 5, where the non-movant’s evidence is to be taken as true and receive all reasonable inferences, see Anderson, 477 U.S. at 255, 106 S.Ct. 2505, and Kansas law, which requires only more-probable-than-not proof of causation.
To the extent some courts insist on “direct” evidence of causation, see, e.g., Smith v. Rapid Transit, Inc., 317 Mass. 469, 58 N.E.2d 754 (1945), while others permit circumstantial evidence to establish liability based on probabilities, see, e.g., Kramer v. Weedhopper of Utah, Inc., 141 Ill.App.3d 217, 95 Ill.Dec. 631, 490 N.E.2d 104, 107-08 (1986); cf. Kaminsky v. Hertz Corp., 94 Mich.App. 356, 288 N.W.2d 426 (1979), we understand that “direct” evidence is not necessarily more accurate than statistical evidence of probabilities, both of which rely on a degree of generality. See FREDERICK Schauee, Profiles, Probabilities, and Stereotypes 93-97 (2003); Michael J. Saks & Robert F. Kidd, Human Information Processing and Adjudication: Trial by Heuristics, 15 Law & Soc’y Rev. 123, 151-56 (1980-81); cf. Ronald J. Allen, A Reconceptualization of Civil Trials, 66 B.U.L. Rev. 401, 414-15 (1986); Daniel Shaviro, Statistical-Probability Evidence and the Appearance of Justice, 103 Harv. L. Rev. 530, 538-39, 542-43 (1989). As one jurist has put it, “it is now generally recognized ... that since all evidence is probabilistic — there are no metaphysical certainties — evidence should not be excluded merely because its accuracy can be expressed in explicitly probabilistic terms.” Richard A. Posner, An Economic Approach to the Law of Evidence, 51 Stan. L. Rev. 1477, 1508 (1999), quoted in Jonathan J. Koehler, When Do Courts Think Base Rate Statistics Are Relevant?, 42 Jurimetrics J. 373, 373 & n. 1 (2002).
Whatever may be the merits of the broader debate over whether courts should allow verdicts on the basis of statistical evidence, Galvin’s mosaic of evidence suffices to meet her burden of proof and thus to defeat Lilly’s motion for summary judgment. Although Lilly maintains that even if Galvin was not exposed to the Marsh Parker pill, she has not demonstrated that she was exposed to Lilly’s pill because other manufacturers of DES in 1964-65 may have produced a pill that matches her mother’s description, Lilly’s alternative argument cannot prevail. The court suggests that mediating this contention would require consideration of a complex burden-shifting scheme in which the plaintiff is burdened with disproving all remaining alternatives. See Op. at 1032. However, Galvin does not bear the burden of proving a certainty. See, e.g., Shields, 895 F.2d at 1466. Galvin’s mother offered a fairly precise description of the pill she took that provides four data points — it was small; it was white; it was round; and it was cross-scored. The court acknowledges that Gal-vin’s mother was shown a photo array in which she identified the Lilly pill but deems significant the fact that the Lilly pill was the only pill among the twenty shown that matched her description. See Op. at 1032. Galvin is, however, entitled to *1043all reasonable inferences. The court fails to recognize that the photo array suggests that the four-point composite description given by Galvin’s mother is uncommon. The only question this court must ask is whether a reasonable juror could infer that there are probably no additional DES pills that match this description. That Lilly could find one other matching pill — the Marsh Parker DES pill — would not make it unreasonable for a juror to infer that Lilly had probably found the only other matching pill.
Accordingly, because the court has improperly raised the standard for defeating summary judgment to require Galvin to produce evidence from which a reasonable juror could conclude that it was significantly more probable that Galvin was exposed to the Lilly DES, I respectfully dissent and would reverse in view of Gal-vin’s Rule 56(e) evidence and reasonable inferences in her favor as the non-moving party, and remand the case to the district court.