How many hands to establish winrate?
 

After approximately how many hands would you consider a winrate to be accurate?

After 7600 hands at Party $25 NL, I have a winrate of 21 PTBB/100. I'm aware that this is not sustainable, but I'm wondering how many hands are required to be reasonably confident that a winrate is accurate.

Re: How many hands to establish winrate?
 

[ QUOTE ]
After 7600 hands at Party $25 NL, I have a winrate of 21 PTBB/100.

[/ QUOTE ]

IMO with good table selection this is entirely sustainable at $25NL. If you keep playing your A game there's no reason why it should drop after 8K hands.

25K seems a good number. You can easily have 5K fluctuations (especially if you play a lot in one day) so 25K gives you an idea. Some say you don't know for sure until 100K, but the $25NL is so soft there's not that much variance if you're playing well. Your big hands always get paid off and suckouts happen with the expected frequency.

Re: How many hands to establish winrate?
 

Yep I was going to say 20k

Re: How many hands to establish winrate?
 

probably on a slight rush.. but it's still time to move up

Re: How many hands to establish winrate?
 

[ QUOTE ]
I'm wondering how many hands are required to be reasonably confident that a winrate is accurate.

[/ QUOTE ]
That depends on your standard deviation. If it is 50 PTBB/100, then a rough 95% confidence interval for your win rate in PTBB/100 after n hands is your result +- 1000/squareroot(n).

[ QUOTE ]
After 7600 hands at Party $25 NL, I have a winrate of 21 PTBB/100.

[/ QUOTE ]
In that case, your confidence interval might be 21+-11.5 PTBB/100.

If your standard deviation is lower, your confidence interval might be smaller.

Re: How many hands to establish winrate?
 

I thought concise and informative answers were banned in SSNL. POTD. [img]/images/graemlins/smile.gif[/img]

Re: How many hands to establish winrate?
 

Sorry about bumping this old thread. I made a mistake and reply in the wrong place.