Basic question about STT's and stats

What, in peoples opinion, is the minimum number of STT's that you need to play before you can start to have some faith in your stats. The two basic numbers that seem to be an indicator of a winning player (or not) seem to be ITM and ROI. How quickly will these converge to a proper indication of your ability?

Thanks

[unreal_nh]

1000+

I'm not so sure about that. That PP player tracker site had me listed as a expert/pro after I won 7/10, and we know how many people go there for info so obviously you only need to have stats for 10. [img]/images/graemlins/smile.gif[/img]

I would agree atleast 1k

[pzhon]

[ QUOTE ]
What, in peoples opinion, is the minimum number of STT's that you need to play before you can start to have some faith in your stats.

[/ QUOTE ]
That depends on your goal.

If you want a close estimate on your ROI, you need a lot of data. If you just want to be confident that you are a winning player, and you are actually winning at a good rate, it doesn't take nearly as many.

The standard deviation of your ROI after n tournaments is about 170%/squareroot(n).

If you want to know your ROI to the percent, you want to play about 340^2 ~ 100k tournaments.

If you are ahead by 1 standard deviation, that is mild evidence that you are a winning player. If you are ahead by 2 standard deviations, that is strong evidence that you are a winning player.

If your ROI is 20%, you only need to play 300 tournaments before it becomes a 2 SD event to break even or lose.

This assumes you are playing 9 or 10 player SNGs. The standard deviation is slightly lower in 5 player SNGs, and slightly higher in 2+ table SNGs.

I don't recommend paying attention to your ITM percentage. Focus on the amount of money you win.

Thanks for the maths, makes more sense to me now.