His conclusion depends on there being a high likelihood of retrieving good, actionable intelligence through the use of torture. All the evidence I've seen indicates that this likelihood is extremely low compared to other methods of retrieving intelligence from a prisoner. I don't have a problem labeling torture too immoral to use, but morality can get low-contrast under extreme circumstances, while facts don't. The fact is that torture doesn't work well enough to employ. This makes it all the more aggravating when attempts are made to defend it on moral grounds.