OPINION
The plaintiffs in this case, Luciano and Pete Montoya, were seriously injured when the car which Luciano Montoya was driving left the road and rolled over several times. They brought suit against the dealer, Galles Chevrolet, and the manufacturer, General Motors Corporation, under a theory of strict products liability. Galles Chevrolet successfully moved against General Motors for indemnification in the event the plaintiffs won a judgment against Galles. The case was tried to a jury, which awarded damages to both plaintiffs in the total sum of $40,000.00. The trial judge then granted the defendants a judgment notwithstanding the verdict. The plaintiffs appeal. We reverse.
The standard for granting a judgment notwithstanding the verdict is the same as that for granting a directed verdict. Francis v. Johnson, 81 N.M. 648, 471 P.2d 682 (Ct.App.1970). The party who prevails in the jury’s verdict * * * is entitled to have the testimony considered in a light most favorable to him and is entitled to every inference of fact fairly deducible from the evidence * * *.’ ” Leonard Motor Company, Inc. v. Roberts Corporation, 85 N.M. 320, 512 P.2d 80 (1973).
Three issues must be considered on the question of whether the jury’s verdict should be reinstated: (1) the sufficiency of the evidence of a defect, (2) the inherent improbability of the plaintiffs’ theory of the origin of the defect, and (3) the inherent improbability of the plaintiffs’ theory of the accident.
I. The plaintiffs argue that the decision of the trial court was incorrect because there was sufficient evidence of a defect to allow the case to go to the jury. The issue is one of the quantum of proof required in a strict products liability case.
The evidence in this case focused on a broken axle. The plaintiffs contended that the broken axle caused the accident; the defendants contended that the axle was broken as a result of the accident.
The plaintiff, Luciano Montoya, purchased the car from Galles Chevrolet. The rear wheel assembly was in the same condition that it was in when Galles received it from General Motors. No mechanical work was done on the car from the time that it was purchased until the accident. The accident occurred three weeks after the automobile was purchased, when the car had been driven about 2,000 miles.
The plaintiffs testified that they were attempting to pass a truck when they heard a dragging sound and the car went out of control. They presented expert testimony that the axle had broken before the car crashed.
Mr. McCamey, one of the plaintiffs’ experts, had worked twenty years in the field of brakes, axles, and wheels, and had analyzed hundreds of axle fractures. In ex*585amining the axle, he conducted a magni-flux test from which he concluded that the fracture was torsional. He also concluded that the fracture was an unusual type and occurred at an unusual place for a torsional fracture. The plaintiffs’ other expert, Mr. Matuszeki, was a metallurgist who had twenty-two years of experience in that field. He had conducted examinations on bearings and misalignments on many axle shafts. A series of examinations had convinced him that the fracture was a torsional one. In addition he stated that the presence of a notch on the axle led him to conclude that a misalignment of the bearing containment hardware caused the notch which in turn caused the axle to fracture at that spot. On cross-examination, the witness further explained his theory to be that something had caused the axle to “freeze”, thus causing the twisting motion.
The standard of proof of a defect in strict liability cases is a well-litigated and debated area. See cases cited in Anno., Strict Products Liability-Proof of Defects, 51 A.L.R.3d 8 and discussion in 50 N.C.L. Rev. 417 (1972)
Carter Farms Co. v. Hoffman-LaRoche, Inc., 83 N.M. 383, 492 P.2d 1000 (Ct.App.1971) established that circumstantial evidence is sufficient to show the existence of a defect. In Carter Farms the court held that it was permissible for a jury to infer that a vaccine given sheep was defective where the sheep had been shown to be in good health prior to the injections and other causes of death had been shown to be improbable.
In automobile cases courts have varied widely in their requirements of proof of a defect. Several courts have held that testimony of the driver that the car was uncontrollable, coupled with evidence that an accident occurred and the car was being used properly, constitutes a sufficient basis upon which to infer a defective condition. See, e. g., Brownell v. White Motor Corp., 260 Or. 251, 490 P.2d 184, 51 A.L.R.3d 1 (1971); McCann v. Atlas Supply Co., 325 F.Supp. 701 (W.D.Pa.1971); Stewart v. Budget Rent-A-Car Corp., 52 Haw. 71, 470 P.2d 240 (1970); cf., Henningsen v. Bloomfield Motors, Inc., 32 N.J. 358, 161 A.2d 69 (1960). This trend has been roundly criticized. Freedman, “Defect” in the Product: The Necessary Basis for Products Liability in Tort and in Warranty, 33 Tenn.L.Rev. 323 (1966); See, 22 Me. L.Rev. 189 (1970). To be distinguished from this form of proof is expert testimony based on an examination of the automobile after the accident. See, e. g., Elmore v. American Motors Corp., 70 Cal.2d 578, 75 Cal.Rptr. 652, 451 P.2d 84 (1969); Chestnut v. Ford Motor Co., 445 F.2d 967 (4th Cir. 1971); Keeton, Manufacturer’s Liability, The Meaning of “Defect” in the Manufacture and Design of Products, 20 Syracuse L.Rev. 559, 564 (1969) (“This type of evidence is virtually always regarded as sufficient.”)
The evidence available in this case is of the second type; there is evidence that the car was defective based on the experts’ examination of the defective mechanism. The plaintiffs have demonstrated that their use was not improper, that the car went out of control, and that a broken axle caused the lack of .control. Carter Farms established a defect in drug cases can be demonstrated by reasonable inferences drawn from the facts. In the instant case, plaintiffs submitted direct evidence of a defect. Based upon Carter Farms, plaintiffs exceeded the standard of proof and presented sufficient evidence of a defect.
Although the defendants take issue with all plaintiffs’ evidence, we reiterate that on appeal from a judgment notwithstanding the verdict we must view the evidence in the light most favorable to the verdict. We conclude that the plaintiffs presented sufficient evidence of a defect to be allowed to go to a jury.
II. The defendants also argue that the plaintiffs’ theory of how the defect occurred is inherently improbable. The theo*586ry of inherent improbability is directed to the theory of misalignment and resulting malfunction of the bearing containment in the left axle assembly. The defendants argue that the notch observed by Mr. Ma-tuszeki was not present immediately after the accident; that there was no evidence of misalignment immediately after the accident; and, that even after welding the bearing rollers to the axle shaft, (an experiment offered by the defendants), the shaft still turned. None of this evidence is sufficient to make plaintiffs’ evidence inherently improbable. First, it cannot overcome plaintiffs’ experts’ testimony that the axle did in fact break while the car was being driven. Secondly, the evidence relied on by the defendants is all met by contradictory evidence of the plaintiffs, so that the resulting conflict was properly one for the jury.
III. The defendants also argue that the “physical facts” rule demonstrates that the plaintiffs’ theory of the case is inherently improbable.
The physical facts rule is found in Ortega v. Koury, 55 N.M. 142, 227 P.2d 941 (1951):
“Physical facts and conditions may point so unerringly to the truth as to leave no room for a contrary conclusion based on reason or common sense, and under such circumstances the physical facts are not affected by sworn testimony which in mere words conflicts with them. When the surrounding facts and circumstances make the story of a witness incredible, or when the testimony is inherently improbable, such evidence is not substantial.”
In the first place it should be noted that a great deal of the plaintiffs’ evidence was physical, so that this case does not present the traditional conflict between unsubstantiated oral testimony and compelling physical fact. See, Bolt v. Davis, 70 N.M. 449, 374 P.2d 648 (1962). The defendants amass a large amount of evidence to demonstrate that the collision was not caused by a broken axle. In doing so they indirectly propound the theory that the accident was caused by the plaintiffs having driven on the shoulder, swinging sharply to get back on the highway and thereby losing control. We cannot on appeal consider the likelihood of this theory because we must take all evidence in a light most favorable to the plaintiffs. The physical fact emerging from this theory which the defendants say contradicts the plaintiffs’ theory is that the truck driver saw the plaintiff turn the wheel and there was testimony that a car with a broken axle is impossible to control. In citing the testimony of the truck driver, the defendants are relying on the same sort of “non-physical” fact which he claims physical facts overcome. In any event the evidence that the plaintiff was trying to control the car does not demonstrate that he “steered” the car to the right side of the road since the plaintiffs’ expert testified that the car could go anywhere if it lost a wheel.
The other physical fact relied on by the defendants as showing inherent improbability was the presence of four skid marks on the highway. The defendants reason that when the axle broke the car’s corner would drop to the ground, leaving a gouge or scrape mark on the highway. However, the plaintiffs intoduced expert testimony that the tire might have been wedged in front of the fender, thus leaving skid marks on the ground. The defendants deride this explanation as impossible because one witness testified that the tire was unscathed. The only evidence that the tire should have had marks on it came from one of the plaintiffs’ witnesses, who said it was “possible” that there would be marks on the tire.
The physical evidence relied on by the defendants as' demonstrating the inherent improbability of the plaintiffs’ case can be explained by the plaintiffs as not inconsistent with their theory. Thus, the physical *587evidence does not contradict oral testimony, but presents “. . .a case of conflicting inferences to be drawn from the undisputed evidence. . . Wilson v. Wylie, 86 N.M. 9, 518 P.2d 1213 (Ct.App.1973).
We reverse and remand this cause with instructions to enter judgment for plaintiffs consistent with the jury verdict.
It is so ordered.
HENDLEY and SUTIN, JJ., concur.
WRIT OF CERTIORARI ■
The Supreme Court of the State of New Mexico having considered a Petition for Writ of Certiorari directed to the Court of Appeals, and having on the 15th day of January, 1976, entered its Order denying the Petition for Writ of Certiorari and returned the records in this cause to the Clerk of the Court of Appeals,
It is ordered That the Clerk of the Court of Appeals issue the Mandate in this cause forthwith.