# Is criminal profiling basically cold reading? (with a special call out to Ianzin) **URL:** \ **Category:** Great Debates\ **Created:** [March 6, 2011, 8:08am UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670 "2011-03-06T08:08:23Z")\ **Posts on this page:** 14\ **Page:** 14
**Author:** ![pravnik](https://avatars.discourse-cdn.com/v4/letter/p/8edcca/32.png) [@pravnik](https://boards.straightdope.com/u/pravnik)\ **Post date:** [March 14, 2011, 3:26pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/261 "2011-03-14T15:26:28Z")
> [@Diogenes\_the\_Cynic](#): > > I still stand by that. It has no investigative utility and has never solved a crime. Your own cites don’t even go that far. From the Gladwell piece: > [@](#): > > In the mid-nineties, the British Home Office analyzed a hundred and eighty-four crimes, to see how many times profiles led to the arrest of a criminal. **The profile worked in five of those cases.** That’s just 2.7 per cent. From the Muller paper provided by Czarcasm: > [@](#): > > Pinizzotto (cited in Wilson et al., 1997), for example, found that > from192 requests for profiles, only **17% actually were used to help > identify the suspect.** More positively, 77% of the respondents > reported that the profile had helped them to focus their investigation. The criticism being leveled towards criminal profiling isn’t that it has never, ever helped solve a single case ever, it’s that there isn’t enough data that it has done so in a statistically significant number of cases so as to prove itself truly useful. If a detective throws a dart at a corkboard covered with photographs of suspects, investigates the suspect that the dart hits who subsequently turns out to be the killer, the dart technique was successful in that one instance. The criticism isn’t that it never works, it’s that limited or anecdotal success doesn’t prove anything - if the next 999 times the dart technique is tried it doesn’t work, the one time it did work can be judged an outlier and the technique can be regarded as not useful on the whole. ---
**Author:** ![Diogenes\_the\_Cynic](https://avatars.discourse-cdn.com/v4/letter/d/e99b99/32.png) [@Diogenes\_the\_Cynic](https://boards.straightdope.com/u/Diogenes_the_Cynic)\ **Post date:** [March 14, 2011, 3:37pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/262 "2011-03-14T15:37:47Z")
That’s not very specific about what is meant by “it worked,” or “it helped.” That just sounds like confirmation bias to me. ---
**Author:** ![pravnik](https://avatars.discourse-cdn.com/v4/letter/p/8edcca/32.png) [@pravnik](https://boards.straightdope.com/u/pravnik)\ **Post date:** [March 14, 2011, 5:17pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/263 "2011-03-14T17:17:22Z")
> [@Diogenes\_the\_Cynic](#): > > That’s not very specific about what is meant by “it worked,” or “it helped.” That just sounds like confirmation bias to me. So when your cite supports your position we should take it as truth, but when it doesn’t we should disregard it? Confirmation bias indeed. :dubious: Okay, here’s a specific, anecdotal, proves-nothing-standing-alone example of one of the 2.7% British cases where profiling was used successfully: John Duffy. > [@](#): > > Eventually, police enlisted the help of Prof. David Canter, an expert in behavioural science and professor of applied psychology at Surrey University. Canter built up a profile based on witness statements and crime scene reports, mostly geographical information. Canter was able to guess that the killer lived in the Kilburn/ Cricklewood area of north-west London, was married, childless, and had an unhappy married life. > > The profile proved to be amazingly accurate in 13 out of 17 points. Canter’s technique was summed up as “the criminal leaves evidence of his personality through his actions in relation to a crime, and any person’s behavior exhibits characteristics unique to that person as well as patterns and consistencies which are typical of the sub-group to which he or she belongs”. > > **When the psychological profile was run alongside the Operation Hart database, it came up with a computer match for John Duffy**. He was immediately arrested at his mother’s house where police also obtained enough forensic evidence to prosecute him. At his trial in 1988, he was sentenced to seven life sentences. Case Study 2: The Railway Killer & Prof. David Canter From BBC news: > [@](#): > > An expert in behavioural science, his first significant success was in helping police catch the so-called Railway Rapist John Duffy who is now serving a life sentence for five rapes and three murders. **Professor Canter drew up an accurate profile of the killer, enabling detectives to eliminate almost 2,000 potential suspects.** [Story](http://news.bbc.co.uk/2/hi/uk_news/288542.stm) For good measure, from article by Muller that Czarcasm linked to: > [@](#): > > Although he observes that police are often very resistant to anything new or to any outsiders telling them how they should do their jobs, **Canter’s advice proved very helpful to the police in catching John Duffy.** [Muller article](http://www.uwinnipeg.ca/academic/ddl/viol_cr/files/readings/reading22.pdf) ---
**Author:** ![Diogenes\_the\_Cynic](https://avatars.discourse-cdn.com/v4/letter/d/e99b99/32.png) [@Diogenes\_the\_Cynic](https://boards.straightdope.com/u/Diogenes_the_Cynic)\ **Post date:** [March 14, 2011, 5:25pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/264 "2011-03-14T17:25:11Z")
Like I said, confirmation bias. What about the all the other times when the profiles are wrong? Just because people sometimes guess right doesn’t mean that guessing is a useful investigative technique. ---
**Author:** ![pravnik](https://avatars.discourse-cdn.com/v4/letter/p/8edcca/32.png) [@pravnik](https://boards.straightdope.com/u/pravnik)\ **Post date:** [March 14, 2011, 5:40pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/265 "2011-03-14T17:40:42Z")
> [@Diogenes\_the\_Cynic](#): > > Like I said, confirmation bias. What about the all the other times when the profiles are wrong? Just because people sometimes guess right doesn’t mean that guessing is a useful investigative technique. :smack: Did you already forget what I said about the dartboard? _The entire point is that profiling has been successful, even if that limited success doesn’t statistically prove it’s usefulness!_ You wanted just one example of any time it has ever successfully been used to solve a crime, and now you have one. It is in no way fatal to your position that profiling overall is not useful. It does not hurt your argument in any way if you acknowledge that you have been given exactly what you asked for. ---
**Author:** ![Musicat](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/musicat/32/20189_2.png) [@Musicat](https://boards.straightdope.com/u/Musicat)\ **Post date:** [March 14, 2011, 5:43pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/266 "2011-03-14T17:43:21Z")
> [@pravnik](#): > > From BBC news: > > > [@](#): > > > > Professor Canter drew up an accurate profile of the killer, enabling detectives to eliminate almost 2,000 potential suspects. > > > > He was even able to pinpoint the area of north west London in which Duffy lived. Without knowing more about the case – and that story doesn’t tell any more – we don’t know how the Professor drew up the profile. You assume that he got nothing to start, but he may have been fed copious data that anyone else could have used to draw the same conclusions. It’s similar to a trick used by mentalists, to make you think the conclusions came from thin air instead of solid data. And you don’t know how many times the Professor drew up other profiles that turned out to be wrong. This is just selective data mining. ---
**Author:** ![pravnik](https://avatars.discourse-cdn.com/v4/letter/p/8edcca/32.png) [@pravnik](https://boards.straightdope.com/u/pravnik)\ **Post date:** [March 14, 2011, 5:56pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/267 "2011-03-14T17:56:07Z")
> [@Musicat](#): > > Without knowing more about the case – and that story doesn’t tell any more – we don’t know how the Professor drew up the profile. You assume that he got nothing to start, but he may have been fed copious data that anyone else could have used to draw the same conclusions. It’s similar to a trick used by mentalists, to make you think the conclusions came from thin air instead of solid data. > > And you don’t know how many times the Professor drew up other profiles that turned out to be wrong. This is just selective data mining. Please see the post directly above yours. I am in no way intending to offer a single example where profiling appeared to work as a defense of the practice as a whole. I am merely pointing out that none of the cites offered state that it has never, ever worked even by accident, just that there’s no data that it works often enough to be considered useful overall. ---
**Author:** ![Diogenes\_the\_Cynic](https://avatars.discourse-cdn.com/v4/letter/d/e99b99/32.png) [@Diogenes\_the\_Cynic](https://boards.straightdope.com/u/Diogenes_the_Cynic)\ **Post date:** [March 14, 2011, 5:58pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/268 "2011-03-14T17:58:27Z")
> [@pravnik](#): > > :smack: Did you already forget what I said about the dartboard? \*The entire point is that profiling has been successful This has not been shown to be the case any more than lucky guessing. ---
**Author:** ![Musicat](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/musicat/32/20189_2.png) [@Musicat](https://boards.straightdope.com/u/Musicat)\ **Post date:** [March 14, 2011, 6:13pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/269 "2011-03-14T18:13:42Z")
> [@pravnik](#): > > Please see the post directly above yours. I am in no way intending to offer a single example where profiling appeared to work as a defense of the practice as a whole. I am merely pointing out that none of the cites offered state that it has never, ever worked even by accident, just that there’s no data that it works often enough to be considered useful overall. I think we might have a definitional difference to “worked.” Just because someone created a description of a criminal and that criminal was caught doesn’t mean that criminal profiling “works.” You haven’t provided any data that shows it works. You are just assuming that a coincidence is something more. If I am sick, take a pill, then get better, did the pill cause it? By your logic, it worked. Should everyone take that pill when they are sick, based on that one example? > [@pravnik](#): > > The entire point is that profiling has been successful… You haven’t shown that. All you have is one possible coincidence in a sea of non-coincidences. ---
**Author:** ![pravnik](https://avatars.discourse-cdn.com/v4/letter/p/8edcca/32.png) [@pravnik](https://boards.straightdope.com/u/pravnik)\ **Post date:** [March 14, 2011, 6:29pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/270 "2011-03-14T18:29:41Z")
> [@Musicat](#): > > I think we might have a definitional difference to “worked.” Just because someone created a description of a criminal and that criminal was caught doesn’t mean that criminal profiling “works.” You haven’t provided any data that shows it works. You are just assuming that a coincidence is something more. You’re missing my point entirely. I am not claiming in the slightest that the profile here _wasn’t_ a lucky guess, or a coincidence, or anything better than a dart thrown at a dartboard, which is why I chose that particular analogy. There’s no way of proving that from a single anecdote. I am merely pointing out that the assertion that it has never led to a single arrest isn’t exactly true - even its harshest critics admit that it has helped lead to arrests of guilty subjects, just not often enough to statistically matter or prove it to be a useful technique. The request was made for a single case where it appeared to work, so I provided one. If the argument is now that a single case proves nothing, sure, I already admitted that. ---
**Author:** ![Musicat](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/musicat/32/20189_2.png) [@Musicat](https://boards.straightdope.com/u/Musicat)\ **Post date:** [March 14, 2011, 7:20pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/271 "2011-03-14T19:20:56Z")
> [@pravnik](#): > > The request was made for a single case where it appeared to work, so I provided one. Weasel word: “appeared.” > [@](#): > > If the argument is now that a single case proves nothing, sure, I already admitted that. A single case proves nothing, not even that it “works.” All you’ve shown is…nothing. > [@](#): > > even its harshest critics admit that it has helped lead to arrests of guilty subjects… Weasel words: “Helped lead to arrests…” Too subjective, and too tainted by opinion. How much help? How did it help? By stating the obvious? Is that helping? You got nothing here. Criminal profiling may not be a textbook case of cold reading, but it’s not a heck of a lot more than that. ---
**Author:** ![pravnik](https://avatars.discourse-cdn.com/v4/letter/p/8edcca/32.png) [@pravnik](https://boards.straightdope.com/u/pravnik)\ **Post date:** [March 14, 2011, 7:42pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/272 "2011-03-14T19:42:00Z")
> [@Musicat](#): > > Weasel word: “appeared.” Yes, appeared to work. Since it’s just an anecdote, that’s all anyone can say about it, but an anecdote was all that was asked for. Going back to your pill analogy, if somebody said “Nobody’s ever taken that pill and gotten better! Tell me one time that that’s ever happened!”, you could reasonably point out “well, that’s not entirely true…to be fair, I took it once and subsequently got better, not that that one instance proves anything.” You’d probably be mystifyed if people kept insisting “that one instance doesn’t _prove anything!_” despite your lack of protestaions to the contrary. > [@Musicat](#): > > A single case proves nothing, not even that it “works.” All you’ve shown is…nothing. I already admitted that, several times. We can’t debate if you agree with me. > [@Musicat](#): > > Weasel words: “Helped lead to arrests…” Too subjective, and too tainted by opinion. How much help? How did it help? By stating the obvious? Is that helping? That’s not my position. I’m paraphrasing Gladwell and Muller, both critcs of profiling. > [@](#): > > You got nothing here. I don’t claim that I have anything. ---
**Author:** ![Musicat](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/musicat/32/20189_2.png) [@Musicat](https://boards.straightdope.com/u/Musicat)\ **Post date:** [March 14, 2011, 10:03pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/273 "2011-03-14T22:03:54Z")
> [@pravnik](#): > > I don’t claim that I have anything. Fair enough. Can I buy you a beer? ---
**Author:** ![pravnik](https://avatars.discourse-cdn.com/v4/letter/p/8edcca/32.png) [@pravnik](https://boards.straightdope.com/u/pravnik)\ **Post date:** [March 14, 2011, 10:10pm UTC](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670/274 "2011-03-14T22:10:06Z")
> [@Musicat](#): > > Fair enough. Can I buy you a beer? Right at 5 pm! It’s a sign! 😃 Anytime, next is on me. [Previous page](https://boards.straightdope.com/t/is-criminal-profiling-basically-cold-reading-with-a-special-call-out-to-ianzin/573670.md?page=13)