I sold a total of 20 flips. 7 of them were heads and 13 were tails. Since you knew the probability came from a uniform distribution, the prior is a beta distribution with parameters (1,1). So effectively you had 8 heads and 14 tails. 

The optimal probability is therefore 8/22 ~ 36%. The market closed at 39% so this market resolves to NO. 

I have sent out a certain number of total flips. If one person had access to all of the flips, they could do some Bayesian updating (with a careful prior choice) to get the best possible estimate of the true probability p. I'll call this estimate p̂. 

This market resolves YES if the final probability on the other market is exactly p̂ when I do the final flip, and NO otherwise. 

#	Name	Total profit
1		Ṁ94
2		Ṁ28
3		Ṁ7
4		Ṁ7
5		Ṁ6

🏅 Top traders

Related questions