I was in my late teens when I read the book and read it with the fascination a teenager has for such things. I certainly remember it better than I remember Jud

I was in my late teens when I read the book and read it with the fascination a teenager has for such things. I certainly remember it better than I remember Jude the Obscure despite writing several papers plus a term paper on that book for my AP English class. The particular story of the roommates stuck out precisely because it exemplified the theme of the book perfectly. I can't provide a citation because I don't remember the title or the author's name, but I can tell you that by expressing skepticism of the story you are in disagreement not just with an anecdote but with the book's overall message and consistent motif, which is not a detail I could have misremembered. You disagree with the author, in short, about whether the result of shooting someone with any particular caliber of handgun is consistent and predictable. The author felt strongly (based on the data he compiled) that it was a mistake to generalize about the effect of any type of gun or ammunition. The author seemed to fear after examining his data that the mere act of stating his conclusions on the relative merits of different types of handguns and ammunition would reinforce a more orderly and predictable vision of the world than the data justified. He said a lot of the appetite for research of the kind he was doing came from officers whose colleagues or subordinates were killed after placing a number of accurate shots on criminals who then lived to serve their sentences in comfort and good health, and he said he had no answers that would prevent that from happening again in the future, only recommendations that would make such incidents a little bit rarer. He took great pains to undermine the idea that the characterization of the real-world performance of any firearm or ammunition load is anything more than statistical, confounded by many other variables. He felt it was wrong to say about any kind of gun or ammunition that it penetrated well or consistently yielded a one-shot stop. He could say that a certain caliber of gun seemed to yield one-hit stops slightly more often than others, but regardless of the caliber you will still see a certain number of four-or-five-hit non-stops followed by the suspect emptying his gun at the officer and then drawing a knife. (I don't think the author put it exactly that way, but it's not snark or exaggeration on my part.) And of course any caliber of gun will yield a one-hit or even zero-hit stop a certain amount of the time, and the best-penetrating round will occasionally fail to penetrate when it matters. To be more concrete, in a gunfight between (say) a .45 and a .22 fired with equal accuracy, any trained person would want to be holding the .45, but the very clear-cut and undeniable advantage a .45 has over a .22 does not translate to a predictable outcome, any more than having two aces in hold 'em guarantees that you will have the best hand after the flop. You can be 100% certain that a .45 is better than a .22 in that situation, but you can't predict with 100% confidence that the guy with the .45 is going to end up better off after the random confounding variables are figured in. You would bet on the .45, of course, but sometimes you'd be wrong, even measured by the long-term effect of the wounds (which as I understand it is more predictable than the short-term effect of being shot.) And of course, as I pointed out before, he had excellent reason and opportunity to cherry-pick a perfect example that would surprise his readers and undermine their appetite for generalizations (which appetite many of them doubtless bought his book to feed.)