Actually, one needs to know the exact pre-flip probability before saying that the odds are 50/50. In real life, knowing the exact pre-flip probability is much harder than it sounds in theoretical math land. In theoretical math land, one just says "assume that this coin is fair and has a 50% of landing on a heads or a tail on any individual flip." In that land if I flip a coin and it turns up heads 23 straight times, I throw my hands up and go "WOW, what are the odds of that happening..well the odds are 1 in 2 to the 23rd power!!! That's an amazing statistical anomaly!!" In real life land, by somewhere between the 6th and the 8th straight "heads" outcome, an observer starts to think that the original assumption of a "fair, 50-50 coin" might not actually be true...by the 15th straight heads call they are REALLY sure the original assumption was false. So, the more you toss a coin that doesn't show a 50/50 outcome, the less likely it is that the coin is a fair 50/50 coin. That's hard to think about though...