View Full Version : Are you a winning player?
masonx
03-13-2005, 03:26 PM
There was a formula for calculating % certainty at which you could say you were a winning player.
We will give generic stats.
Winrate: 2bb/100
SD: 10bb/100
#of hands: 10,000
someone calculate this and state the equation so we can all use it /images/graemlins/smile.gif
please!?
I completely forgot the formula
wtfsvi
03-13-2005, 03:40 PM
try this post (http://forumserver.twoplustwo.com/showflat.php?Cat=&Board=inet&Number=1342415&Forum= All_Forums&Words=confidence%20interval&Searchpage= 0&Limit=25&Main=1342415&Search=true&where=bodysub& Name=167&daterange=1&newerval=1&newertype=y&olderv al=&oldertype=&bodyprev=)
BruceZ
03-13-2005, 05:27 PM
[ QUOTE ]
There was a formula for calculating % certainty at which you could say you were a winning player.
We will give generic stats.
Winrate: 2bb/100
SD: 10bb/100
#of hands: 10,000
someone calculate this and state the equation so we can all use it /images/graemlins/smile.gif
[/ QUOTE ]
Standard deviation of win rate (standard error SE) for N*100 hours is:
SE = SD/sqrt(N) = 10/sqrt(100) = 1 bb/100.
Win rate of 2 bb/100 is 2 standard errors above 0, so there is a confidence of 97.7% that we are a winning player. Get this from a table of the standard normal distribution, or Excel function NORMSDIST(2) = 97.7%.
Note: Your units of SD are technically not correct. They should be 10 bb for 100 hours, or
10 bb/sqrt(100 hours).
masonx
03-14-2005, 08:46 AM
bruce i thought as long as you kept winrate in BB/100 then you could also use SD in BB/100
vBulletin® v3.8.11, Copyright ©2000-2024, vBulletin Solutions Inc.