2010-07-25, 05:58 AM
Stereo Wrote:http://changingminds.org/disciplines/arg...allacy.htm
It is still 50/50 no matter what happened earlier.
In fact if you have a coin come up heads that often I'd say it's slightly more likely that it's a biased coin, and will keep coming up heads. But if it's a fair coin then it's always 50-50.
HHHHHHHHHHT is no more a statistical anomaly than HHHHHHHHHHH, it just looks tidier as 11 heads in a row. Any specific sequence of 11 results is very unlikely to occur (1 in 2048, or 0.05%, but in general if you flip a coin 11 times, you are going to hit one of the sequences.
Well, its not a biased coin b/c its a theoretical coin that has a 50/50 chance. I like the 2nd part, maybe I wasn't looking at the "anomaly" characteristic of every sequence.
Omg, actually thats all I needed, you could have said "all sequences are 'anomalies'" and that would have been enough. I was having the same question discussion on facebook and just wasn't convinced. I'm kind of disappointed there wasn't more discussion of all the dispute theories here, so...
Say i flip a coin infinite times, wouldn't it work itself into a 50/50 split. If its a 50% chance each flip, then after infinite flips it should reflect that probability otherwise i'd say the probability isn't really a correct probability after all. So if i got 10 heads wouldnt it eventually work to a 50/50 split and wouldnt that make tails more likely than heads. It seems unlikely that something unlikely would become even more unlikely more often than becoming slightly less unlikely. (i used unlikely a lot there, but the point is there, the verb'age isnt b/c its getting late for me =P)