Comments on: That is why you fail

By: Darien

Darien — Tue, 02 Jun 2009 10:23:46 +0000

Well, I’d say they’re right to an extent — if their luck has been *extremely* bad, they’re likely to improve somewhat going forward, but only toward, you know, the mean — if you’re betting money on tosses of a fair coin, and over your first 100 bets you win 20, yes, you’re likely to see some better performance than that, but you are not likely to win 80 of your next 100 because the universe magically knows it needs to live up to your own standards of “fairness.”

When I used to work at the gas station many, many years ago, we had many scratch ticket players with such a poor understanding of probability as to boggle the mind. Now, one could argue that it takes a pretty poor grasp of probability to play scratch tickets in the first place, and one wouldn’t be wrong, but it’s also the case that they’re fairly cheap and the actualy mechanical process of scratching them off is entertaining. But I digress; we got some tickets in once that had stamped right on the face “1 in 4 chance of winning!”

You have no idea how many people bought four tickets, didn’t win anything, and then declared that they were going to sue us for “false advertising.” Which also, incidentally, doesn’t mean what people think it means.

By: Dave

Dave — Mon, 01 Jun 2009 15:58:25 +0000

I often hesitate to talk about regression to the mean because too often many people use it to reinforce their mistaken belief in the gambler's fallacy. You're using it correctly here, because we're discussing his entire run from start to finish--but too often people insist that they're "due" to hit because their luck has run so bad that regressision to the mean is totally about to work in their favor!

By: Darien

Darien — Mon, 01 Jun 2009 06:42:28 +0000

This bullshit is hanging on the wall of the office at my work. Just wanted to throw that out there. Oh, and you should probably also mention regression to the mean in there somewhere, too.