glenra comments on Closet survey #1 - Less Wrong
You are viewing a comment permalink. View the original post to see all comments and the full post content.
You are viewing a comment permalink. View the original post to see all comments and the full post content.
Comments (653)
So I'm not a mathematician but we note the outcomes of chance events all the time probably thousands to tens of thousands of times in your life depending on how much gaming you do. Given about 1000 low-likelihood events per person over their lifetime (which I'm basically making up, but I think its conservative) 1 in 100 million should experience 1 in 100 billion events, right? So basically there might be two other people with stories like yours living in the US. It is definitely a neat story, but I don't think its the kind of thing we should never have expected to happen. Its not like the quantum tunneling of macroscopic objects or anything.
1000 is extremely conservative. Every time you play any game with an element of chance - risk, backgammon, poker, scrabble, blackjack, or even just flipping a coin - the odds against you getting the exact sequence of outcomes you do get will be astronomical. So the limiting factor on how many unbelievable outcomes you perceive in a lifetime is how good you are at recognizing patterns as "unusual". Somebody who studied numerology or had "lucky numbers" or paid attention to "lucky streaks" would see them all the time.
In the case at hand, that same series of rolls would be just as unlikely if it had happened at the beginning of the game or in the middle or spread throughout the match and hadn't determined the outcome. Unless there was something special about this particular game that made its outcome matter - perhaps it was being televised, there was a million dollars bet on it, or it was otherwise your last chance to achieve some important outcome - the main thing that makes that sequence of rolls more noteworthy than any other sequence of rolls of equivalent length is selection bias, not degree of unlikeliness.
"the odds against you getting the exact sequence of outcomes you do get will be astronomical"
People notice and remember things they care about. Usually people care whether they win or lose, not the exact sequence of moves that produced the result. For an event to register as unusual a person must care about the outcome and recognize that the outcome is rare. The Risk game was special because I cared enough about the outcome to notice that I was losing, because the outcome (of losing) with 26 vs. 1 armies was incredibly unlikely, and because I could calculate the odds against such an outcome occurring due to chance.