Traditional tort law resolves questions of causation with an unforgiving arithmetic. A plaintiff who cannot show that the defendant's negligence more likely than not caused her injury - that it was responsible for the harm with a probability exceeding fifty percent - recovers nothing; a plaintiff who clears that threshold recovers in full. This all-or-nothing rule works tolerably in the ordinary case, where the causal link between wrong and injury is either reasonably clear or plainly absent. It produces troubling results, however, in a recurring class of medical-malpractice disputes. Consider a patient whose cancer, had it been diagnosed promptly, carried a forty-percent chance of survival, but whose physician negligently failed to detect it, after which the patient died. Under the conventional rule the claim fails at the threshold: because the patient would probably have died even with competent care, the negligence cannot be shown to be the more-likely-than-not cause of death. The physician's carelessness, however real, is in effect immunized - and immunized precisely in the cases where a patient's prospects were already grim, so that the worse a patient's odds, the less accountability the law imposes on those who diminish them further. To address this asymmetry, many jurisdictions have adopted the "loss of chance" doctrine, which reconceives the injury. On this view the compensable harm is not the death itself but the destruction of the patient's statistical prospect of survival - a distinct interest, valued proportionally. The negligent physician who reduced a forty-percent chance to zero is liable not for the whole of the death but for forty percent of the associated damages. The doctrine thus severs recovery from the fifty-percent threshold and scales it to the magnitude of the chance destroyed. Critics raise two objections. The first is conceptual: a mere probability of survival, they contend, is not the sort of concrete interest the law has traditionally protected, and to treat a statistical artifact as an injury is to compensate phantom harms. The second is practical: once lost chances are actionable, they warn, liability will metastasize, since almost any negligence marginally alters some probability, and courts will be flooded with speculative claims resting on contested statistics. Neither objection is decisive. The conceptual complaint proves too much, for the law routinely assigns value to probabilistic interests - a lost business opportunity, the contingent expectancy of an heir - and a patient's chance of survival is no more ethereal than these. The practical worry has more substance, but it counsels careful limitation rather than wholesale rejection: courts can confine the doctrine to settings, such as established treatment relationships, where reliable statistical baselines exist and the defendant's duty is clear. What the critics establish, at most, is that the doctrine requires disciplined boundaries - not that the traditional rule, with its perverse indifference to the fate of the already-imperiled, is the better of the two regimes.