Thread: SNG statistics
View Single Post
  #22  
Old 06-15-2006, 12:19 PM
Frinkenstein Frinkenstein is offline
Senior Member
 
Join Date: Mar 2006
Location: Springfield University
Posts: 552
Default Re: SNG statistics

[ QUOTE ]
Maybe I've misunderstood something, but I used a 95% confidence interval to calculate that: http://en.wikipedia.org/wiki/Confidence_interval

<p - z sqrt(p(1-p)/n), p + z sqrt(p(1-p)/n)> = <.19 - 1.96 sqrt(.19(1-.19)/150), .19 + 1.96 sqrt(.19(1-.19)/150)> = <.13, .25>

Don't you think he's got any idea wether he's a winning player when he's played 150 SnGs with a ROI of 19%?

[/ QUOTE ]

Yeah, when I started playing SNG's I had a 95% confidence of having an ROI greater than 20% after 100 games.

100 (i.e. after 200) games later my ROI was -5%. The confidence interval does not capture the variability of SNG play. either that or I beat the odds and went on a huge losing streak. Probably a bit of both.