Dissenting Opinion by
MEREDITH, J.1. The patient’s greatly increased likelihood of death established probable causation.
As the majority opinion recognizes, this is not a case in which the plaintiff sought to recover for a “loss of chance” of survival for a patient whose probability of survival was less than 50% at the time of the alleged negligence. At the time of the alleged negligence of the defendants in this case, assuming all facts in a light most favorable to the plaintiff, as we are obligated to do at this stage in the litigation, the patient’s likelihood of survival was rather good: 80%. Stated another way, at the time of the defendants’ alleged negligence, the patient’s odds of successful treatment were 4-to-l, and the patient’s statistical likelihood of death was only 20%, or one chance in five. As a consequence of the delay in treatment caused by the defendants’ alleged negligence, the patient’s statistical likelihood of death more than doubled, to approximately 50%, before treatment was begun. Under such circumstances, a rational trier of fact could conclude that the defendants’ negligence was the probable cause of the patient’s death. To conclude otherwise requires a finding that, at the time of the missed diagnosis, the patient was among the 20% who were going to die anyway. From a probability stand*699point, however, it is much more probable (by a ratio of 4-to-l) that the patient was among the 80% who would survive if properly treated. I do not understand how one could rationally conclude that the delay that reduced the patient’s odds of survival from 4-to-l to merely 1-to-l was not a probable cause of her ultimate 100% death. Accordingly, I would reverse the judgment of the circuit court.
Unlike the plaintiffs who asserted the claims in Weimer and Fennell even though it was undisputed that their likelihood of death exceeded 50% before any act of medical negligence, the plaintiff in this case offers proof that, at the time of the defendants’ alleged negligence, the patient’s chance of survival was 4 times greater than her likelihood of death. Unlike Weimer and Fennell, liability in this case does not depend upon the recognition of an action for loss of a chance. The traditional standard of causation, as expressed in the Maryland Pattern Jury Instructions, MPJI-Cv 19:10, can be met. The pattern instruction on causation states:
For the plaintiff to recover damages, the defendant’s negligence must be a cause of the plaintiffs injury. [There may be more than one cause of an injury, that is, several negligent acts may work together. Each person whose negligent act is a cause of an injury is responsible.]
We do not need to resort to statistics and estimates to know that the patient in this case died. Notwithstanding the evidence that, at the time her treatment was eventually begun, there was a statistical likelihood of survival of 50% for patients whose cancer had progressed to a stage similar to Ms. Schaefer’s, we now know with certainty that she was among the 50% who would ultimately die. The pertinent causation question for the trier of fact is whether it is probable that she was, at the time of the defendants’ alleged negligence, already in the group who would die regardless of treatment. Statistical probability does help us answer that question: because 80% of all patients would survive, it is statistically much more likely than not that Ms. Schaefer was in the 80% group who would survive (rather than the 20% group who would not) if treatment had begun at the time of the defendants’ alleged negli*700gence. Common sense permits a rational conclusion that the delay occasioned by the defendants’ alleged negligence most likely caused Ms. Schaefer to shift from the group comprised of the 80% of all patients who would survive into the group who ultimately would not.
This common sense conclusion is supported by mathematics. Professor Lars Noah, of the University of Florida Levin College of Law, has written a law review article that points out the mathematical blunders lawyers and judges frequently make when analyzing statistics regarding the likelihood of patient survival. Lars Noah, An Inventory of Mathematical Blunders in Applying the Loss-of-ar-Chance Doctrine, 24 Rev. Litig. 369 (2005). In the article, Professor Noah points out the arithmetical fallacy in the analysis that led to the entry of summary judgment for the defendants in this case. Professor Noah explains that it is not arithmetically correct to require (as the majority opinion does) that a plaintiff demonstrate that the defendants’ negligence resulted in more than a 50% loss of a chance of survival. On the contrary, he asserts that the correct analysis would focus upon the patient’s increased risk of death and permit recovery when the attributable risk ratio was over 50%. Professor Noah explains that, if the patient’s likelihood of death more than doubles between the date of the missed diagnosis and the date treatment begins, and the patient in fact dies, then the delay was a probable cause — not merely a possible cause — of the patient’s death. In contrast to the analysis set forth in the majority opinion, Professor Noah observes, id. at 393-97:
Some courts ... demand[ ] that the plaintiff demonstrate that the defendant’s negligence resulted in more than a 50% loss of a chance of survival.
Under the latter view, a patient who experiences a drop [in likelihood of survival] from 60% to 40% (20 percentage points) fails on causation grounds, as does a drop from 60% to 15% (45 percentage points), but a patient who drops from 60% to 5% (55 percentage points) would prevail. In fact, this approach results in an error ... insofar as the patient who experiences the 45 percentage point drop *701(from 60% to 15%) also should satisfy traditional causation requirements because the alleged negligence more than doubled the mortality rate from 40% to 85% (calculated by subtracting each of the survival estimates from 100%). Although it amounts to the same 45 percentage point change, here an increase rather than a decrease, these numbers give a “relative risk” (a.k.a. “rate ratio”) of 2.13 (.85/.40), where 1.0 functions as the baseline, or an “attributable risk” (a.k.a. “attributable fraction”) of 53% ((.85-.40)/ .85), which means that the defendant’s negligence probably caused the ultimate injury.
In this vein, one might say that courts make a mistake of looking through the wrong end of the telescope. Some courts have endorsed an “increased risk” theory as the basis for recognizing loss-of-a-chance claims, but they nonetheless still tend to frame the statistical information in terms of reductions in the probability of survival caused by the malpractice. Instead of asking about the loss of a chance for survival, courts should focus on the flip-side question framed as the increased risk of morbidity and mortality.
[I]f a patient would have enjoyed an 85% chance of survival, which then drops to 68% due to the negligent failure to diagnose, the courts mistakenly view it as a 17 percentage point or perhaps a 20% ((.85-.68)/.85) loss of a chance of survival. If converted into risk estimates, these numbers actually would satisfy the traditional causation standard, at least insofar as the patient’s odds of dying have more than doubled, from 15% to 32% ((.32-.15)/.32 = 53.1%). This attributable risk calculation does not exclude the possibility that the patient would have died in any event, but, as courts increasingly require in tort cases, proof of a doubling in the relative risk after exposure may help to establish causation by a preponderance of the evidence.
(Bold emphasis added; footnotes omitted.)
Applying the relative risk analysis urged by Professor Noah to the present case would produce the following result. The *702patient’s risk of death rose from 20% at the time of the missed diagnosis to 50% by the time the cancer was diagnosed and treatment was begun; in other words, the risk of death “more than doubled.” Inserting these numbers into Professor Noah’s equations produces the following: (.50-.20)/.50 = 60.0%. That is, the delay attributable to the defendants’ missed diagnosis increased the patient’s “attributable risk” of death by 60%, which is, obviously, more than 50%, and therefore, sufficient to attribute causation to the physicians who negligently missed the diagnosis.
As in the last example quoted from Professor Noah’s article, the patient’s odds of dying in this case more than doubled (from 20% to 50%) between the time of the missed diagnosis and the beginning of treatment. As Professor Noah states, this “means that the defendants’] negligence probably caused the ultimate injury.” Id. at 394.
Although the majority opinion purports to base its affirmance of the judgment in favor of the defendants upon an arithmetical analysis (concluding that “the alleged malpractice diminished Ms. Schaefer’s chance of survival by, at most, thirty percent”), the majority refuses to recognize that the more appropriate question is whether the alleged malpractice increased the likelihood of the death that ultimately occurred by more than 100% during the period of delay. From a probability standpoint, the rational trier of fact could rationally conclude that the delay was the probable cause of the patient’s death if the delay caused the statistical risk of death to more than double.
The majority opinion dismisses this analysis as focusing upon a “mere possibility” of causation. On the contrary, this analysis focuses directly upon probability and only permits recovery for wrongful death when the evidence establishes that the delay was a probable causé of the patient’s death. The example used by the majority opinion in its attempt to illustrate a fallacy in Professor Noah’s formula appears, at first blush, to expose a flaw in the theory. But upon examination, the example confirms that focusing upon likelihood of *703mortality is the proper approach. The majority opinion states:
Apparently, under Professor Noah’s approach, recovery would be allowed if a decedent’s chance of survival decreased (due to defendant’s negligence) from ninety-eight to ninety-five percent because the decedent’s “chance of morbidity” would have “more than doubled” from two to five percent. Adoption of such an approach would allow recovery for wrongful death based upon the mere possibility that prior to the malpractice the decedent was not in the two percent of the population that would have died absent the negligence.
Putting aside the extreme nature of the hypothetical, and the fact that few such cases will arise because there is such a small chance of mortality in any event (that is to say, even after the delay in treatment, 95% of the patients survived and the other 5% who died were not all victims of negligent treatment), if we focus upon the group of patients who did, in fact die, it is rational to attribute causation to the delay that more than doubled their risk of death. In this example, at the time of the alleged negligence, only two patients out of 100 would die regardless of treatment. And, indeed, it is possible that the plaintiffs decedent was one of those two. From a statistical analysis of the patient population, however, it is extremely unlikely that plaintiffs decedent was one of those two because 98% of all patients with this condition survive if properly treated at that point time. The example further assumes that, as a consequence of medical malpractice, treatment is delayed for a period of time, and that, as a direct consequence of the delay, by the time treatment begins for plaintiffs decedent, five out of 100 patients will die. We know with 100% certainty that plaintiffs decedent was one of those five patients who would eventually die. If we focus on the five patients who eventually die regardless of treatment, we know that, from a statistical standpoint, only two of those five patients would have died if treatment had been initiated at the time of the defendants’ medical malpractice; the other three patients would have survived if treatment had been initiated at *704the time of the malpractice. Again, it is possible that decedent’s patient was among the two who would eventually die regardless of treatment. But it is more probable that plaintiffs decedent was among the three who would have otherwise survived simply because that is a larger group. Focusing upon probability, it is more likely than not — ie., probable— that plaintiffs decedent in this example would have survived had the treatment not been delayed by the defendants’ malpractice.
Returning to the case before the Court, the majority opinion assumes that the patient’s chances of survival were as high as 80% at the time of the alleged malpractice that resulted in a delay in treatment, and as low as 50% at the point in time that Ms. Schaefer’s treatment was begun. In the population of patients with a similar medical condition, 20 out of 100 would die regardless of whether treatment was initiated at the time of the alleged malpractice. As a direct consequence of the delay caused by the alleged malpractice, in the population of patients with a condition similar to Ms. Schaefer’s worsened state as of the time her treatment was belatedly begun, 50 out of 100 would die despite being properly treated. We know that Ms. Schaefer was one of those 50 patients who would ultimately die. From a statistical standpoint, any given patient, including Ms. Schaefer, was more likely to be among the 30 patients whose condition progressed to the point of being unsurvivable during the period of delay, with only a 40% likelihood (ie., 20 out of 50) that she was among the 20 patients who were already terminally ill at the time of the defendants’ alleged negligence.
We should not analyze this issue by looking through the wrong end of the telescope, which is what Professor Noah concludes the analysis employed in the majority opinion does. In this case, our common sense tells us that the delay probably caused the patient’s cancer to progress to a point that precluded her survival. Although resort to mathematical formulas should not even be necessary, in this case at least, as Professor Noah has made clear, the arithmetic, when properly *705employed, confirms our common sense conclusion regarding causation.
The numbers do not preclude a finding that the delay caused by the defendants’ alleged negligence was a probable cause of Ms. Schaefer’s eventual death. Consequently, the numbers do not support granting the appellees’ motions for summary judgment.
2. The sham affidavit rule
I also disagree that the affidavits of Dr. Hutchins and Dr. Shmookler were properly stricken as “sham” affidavits. Such a ruling suggests that these expert witnesses outright lied at the time they affirmed the statements made in the affidavits. That inference is based upon an alleged “contradiction” between the deposition testimony and the affidavits. As the commentators state in the Maryland Rules Commentary, the recently adopted Rule 2-501(e) “sets up the presumption that an affidavit or other sworn testimony that materially conflicts with prior sworn testimony of that person is a ‘sham’ and may be stricken.” Paul V. Niemeyer & Linda M. Schuett, Maryland Rules Commentary at Supp. 50 (3d ed.2003, 2006 supp.). Cf. Pittman v. Atl. Realty Co., 359 Md. 513, 539, 754 A.2d 1030 (2000) (holding, prior to adoption of current Rule 2-501(e), that it is improper for motion court to decide credibility of affidavit testimony).
Although there is undeniably a difference between the expert witnesses’ deposition testimony and the affidavits, the distinction is not so irreconcilable that, as a matter of law, the court should refuse to give any credence whatsoever to the statements made in the affidavits. Whether the statements were untimely and should have been disregarded because they were in flagrant conflict with the scheduling order in the case is a different issue. Cf. Rodriguez v. Clarke, 400 Md. 39, 69-70, 926 A.2d 736 (2007) (“The Clarkes’ preterition reflected by their sparse expert witness designation, elusive answers to interrogatories, and failure to communicate, warrant preclusion of their experts — the sanctions were proportionate to the *706discovery abuse”). But that was not the basis relied upon by either the circuit court or the majority opinion. In my view, the supplemental opinions expressed by the experts in their affidavits in this case did not contradict the opinions they had previously expressed. The affidavits were not irreconcilably at odds with the opinions expressed by them at their depositions, and the affidavits should not have been considered mere shams subject to being stricken under Rule 2-501(e).
An appropriate test for assessing whether an affidavit should be considered a sham would be this: if the statement made during the deposition of the witness, when fairly considered in context and in light of any explanatory statements given during the deposition, is assumed to be the true testimony of the witness, would the statement in the affidavit necessarily be false? If so, the affidavit should not be given effect by the court unless the witness can satisfy the court that the provisions of Rule 2 — 501(e)(2) apply and thereby satisfactorily explain the contradiction.
In this case, however, neither expert expressed an opinion during the deposition and then purported to express a contradictory opinion by way of affidavit. The experts were asked whether they intended to express an opinion on causation, and each answered that he did not so intend as of the date of the deposition. Changing one’s mind as to whether to make any statement about a subject is different from expressing one opinion and then expressing the opposite opinion. And, in most cases, the scope of subject areas as to which an expert will be asked to testify is within the knowledge and control of counsel rather than the witness.
The allegedly contradictory deposition testimony, in each instance, related to the witness’s present state of mind with respect to his intended future conduct. In Dr. Hutchins’s case, he answered “no” when asked: “Are you going to be rendering an opinion within reasonable medical probability as to Ms. Schaefer’s cause of death?” Dr. Shmookler’s allegedly contradictory testimony was similar: “I’m not going to be going into that.”
*707The situation is similar to one in which a person testifies on the date of his deposition, “I am not going to eat tomorrow,” but then nevertheless eats something the following day. If the witness then gave an affidavit swearing that he ate the day after the deposition, such testimony does not contradict the fact that he intended otherwise at the time of his earlier testimony. It may not be expected by the other party — which might or might not present a basis to preclude the supplemental testimony on the topic — but the affidavit testimony is not so inherently incredible or such an obvious sham that it should be treated as false and stricken on the basis of Rule 2-501(e). See generally Pittman, supra, 359 Md. 513, 754 A.2d 1030, in which Judge Rodowsky sets out many sound reasons for exercising judicial restraint in connection with a sham affidavit rule.
Whether the supplemental opinions of these experts came too late in the discovery process for the court to consider them is a different question from whether they contradicted the opinion testimony previously given by the witnesses. But Rule 2-501 (e) should not be the basis of excluding affidavits that contain supplemental opinions of experts.