re: #67 freetoken

Hmmm… it actually works the opposite.

Think of tossing a coin. I have only about a 50% chance of accurately telling you what that flip’s outcome will be.

However, if you flip a coin every minute for the next year, I can pretty assuredly tell you that almost half of them will be heads.

Hmm, wait a sec. Been a while since I took a class in probability, but the problem with teh coin toss and the gambler’s fallacy is that every individual coin flip always has a 50% chance of coming up heads (say). People watching you flip a (fair, nonweighted) coin tend to mistakenly think that because it came up heads 7 times in a row, it’s more likely to come up tails on the 8th toss— but the 8th flip has just the same odds of coming up heads as ever, namely, 50%.

There is a point where with a really large number of coin tosses you do get a levelling out, and can reasonably state that the results will be 50-50 — but it’s an incredibly large number. I don’t think the one per minute for a whole year is close enough to it. Not sure though.

I’m being a pedant here though, and in practice it works as a good guess that it would be 50-50, and you’re making all the right points about the way the probablity of the individual toss works.