Chess World Champions are sometimes notoriously superstitious, you can still rely on the consistency of their chess moves.
No, you can’t. In 2006, world chess champion Vladimir Kramnik accidentally left himself open to mate in one when playing against computer program Deep Fritz (http://www.chessbase.com/newsdetail.asp?newsid=3509). Even the very best individual humans are all subject to simple mistakes of types that computers simply don’t ever make.
The original question was not whether humans make mistakes (they do in every area, this is undisputed) but whether irrationality in one domain makes more unreliable in others.
No, the original question was whether we should be surprised when humans make mistakes, and what influences the probability of them doing so. The occasional grandmaster bluder shows that even for extremely smart humans within their field of expertise, the human mind effectively has a noise floor—ie, some minimum small probability of making stupid random decisions. Computers, on the other hand, have a much lower noise floor (and can be engineered to make it arbitrarily low).
You shouldn’t be surprised that a chess world champion has made a mistake over the course of their entire career. However, given a specific turn, you should be surprised if the world champion made a mistake in that turn. That is, given any turn, you can rely on their making a good move on that turn. You can’t rely with perfect confidence, of course, but that wasn’t the claim.
No, you can’t. In 2006, world chess champion Vladimir Kramnik accidentally left himself open to mate in one when playing against computer program Deep Fritz (http://www.chessbase.com/newsdetail.asp?newsid=3509). Even the very best individual humans are all subject to simple mistakes of types that computers simply don’t ever make.
This is irrelevant. Human players make mistakes. The question is whether being superstitious makes them make more mistakes.
It’s not just chess—here’s two 9dan go players, one of them misthinking and killing his own group: http://www.youtube.com/watch?v=qt1FvPxmmfE
Such spectacular mistakes are not entirely unknown in go, even in top level title matches.
In pro-level shogi it’s even worse, as illegal moves (which are instant lose) are supposedly not at all uncommon.
The original question was not whether humans make mistakes (they do in every area, this is undisputed) but whether irrationality in one domain makes more unreliable in others.
No, the original question was whether we should be surprised when humans make mistakes, and what influences the probability of them doing so. The occasional grandmaster bluder shows that even for extremely smart humans within their field of expertise, the human mind effectively has a noise floor—ie, some minimum small probability of making stupid random decisions. Computers, on the other hand, have a much lower noise floor (and can be engineered to make it arbitrarily low).
You shouldn’t be surprised that a chess world champion has made a mistake over the course of their entire career. However, given a specific turn, you should be surprised if the world champion made a mistake in that turn. That is, given any turn, you can rely on their making a good move on that turn. You can’t rely with perfect confidence, of course, but that wasn’t the claim.
Even chess computers can blunder, it seems.