Interrogation

[ravenfish]

Originally Posted by fula farbrorn

Where Torture can work well is things like "give us your computer password" as you can verify it in real time.

Of course, there's the risk the "password" they give you is actually the string of characters that triggers the deletion of everything on the drive...

I have seen it alleged (though I will not vouch for the accuracy), that, under torture, IRA members sometimes "broke down" and revealed times and places the British should investigate- that happened to be prearranged ambushes prepared for just such a contingency. In general, the number of people who, under torture, confessed to hexing their neighbors and flying around on broomsticks should give a fair idea of the method's ability to extract accurate information.

[WingedKagouti]

Originally Posted by ravenfish

In general, the number of people who, under torture, confessed to hexing their neighbors and flying around on broomsticks should give a fair idea of the method's ability to extract accurate information.

There was a lot of issues with the witch hunts, one of which was some of the witch hunters being paid for how many "witches" were caught and executed (sometimes the money were extracted from the accused/their family). That kind of policy tends to make some people go for a volume approach with as much emphasis on getting a confession as possible.

"The pain doesn't stop until you confess."

[Sam Baughn]

Originally Posted by David Johnston2

To be specific the way in which torture is ineffective is that it tends to produce results that reflect the interrogator's initial bias as the subject tries to produce whatever the interrogator wants to hear regardless of whether or not that's the truth.

Thing is, that's a huge problem with literally every form of extracting information from people. People confess to crimes they didn't commit all the time under normal police interrogation. Anthropologists have to put a lot of effort into sorting out plausible accounts of cultural traditions from things made up by their interview subjects. Even anonymous surveys get a lot of false data. People will literally ruin their lives by admitting to terrible things they never did just because they feel that someone wants to hear it. Add sleep deprivation, drugs, legal threats, political pressure, etc. and you're likely to get results just as bad as those obtained by beating it out of them.

It can be educational to study how the Inquisition used torture. Contrary to their reputation, they were extremely wary of false confessions and didn't convict people based on the statement of a single subject under duress. Instead, they used torture carefully, after a long period of trying to get information by other means, and always checked the information they got either against provable facts (e.g. physical evidence) or by cross-referencing multiple independent statements. Despite not being in any rush and using other methods first they often found information they would otherwise not have done during torture sessions. It seems that most inquisitors were quite reluctant to actually use torture (especially since their methods were rather more dangerous than stress positions and waterboarding and they didn't want to kill prisoners who may have information) but they kept using it for centuries because it gave them results.

I think +6 is far too generous, but I don't think it is realistically useless either. At the very least, it should offset penalties for haste or allow you to claim the same benefit as taking extra time.

[ravenfish]

I think it would be useful to distinguish between rolls to achieve a desired effect (get the prisoner to give an essentially prewritten confession, or just break down and start babbling) and rolls to extract useful information. There is no denying that torture is exceeding effective at the former (that is part of the reason the Inquisition et. al. continued to use it in light of the knowledge of its drawbacks- not only was it nice to be able to say, when faced with doubters, that the sinner had confessed from his own lips, but confession might make it easier on him in the next life, so torturing him until he confessed was really doing him a huge favor, and the Church has always been in favor of charity), its utility in the latter is considerably less.

[Eukie]

Originally Posted by ravenfish

There is no denying that torture is exceeding effective at the former

Statistics from the use of torture to extract confessions in France before the 20th Century actually show a success rate of merely 6-30%.

[ravenfish]

Originally Posted by Eukie

Statistics from the use of torture to extract confessions in France before the 20th Century actually show a success rate of merely 6-30%.

Blimey. That's fairly impressive, in a disturbing kind of way.

[GWJ]

Torture is actually good (I mean "with high chance of succeeding") way to get people talk. And just it - without arguing about quality of this "testimony". It can be truth, it can be total ********, but he'll talk SOMETHING, sooner or later.

[malloyd]

Originally Posted by Refplace

Given that a failed roll could result in a believable lie maybe the +6 is too high.

Thinking about it a bit, I think the most reasonable way of handling that might be to go ahead and allow that +6, or maybe even +8, but only to offset penalties for taking less time. Letting you make an unmodified Interrogation roll in 40% (or 20%) of the usual time seems like it models the major advantage of torture (it's fast) without giving you an unwarranted increase in the chance the answer you get is true.