How many hands is a "good sample size"

thepokerkid123

thepokerkid123

Visionary
Silver Level
Joined
Jul 31, 2009
Total posts
917
Chips
0
Depends what you're trying to prove.

If you just want to know if you're winning or not, 20-50k will give you a very good estimate.

Significantly positive or negative results are enough to make an educated guess if you're overall winning or losing on a much smaller sample. Not sure on what the numbers would be, but if you burn through 15 buy ins in 10k hands it's a pretty good hint.

If you want to calculate your winrate, >100k.



This really all depends on your standard deviation and how precise you want your conclusions to be.
 
sammyfive

sammyfive

Rock Star
Silver Level
Joined
Oct 4, 2009
Total posts
128
Chips
0
Alright so that is quite a few hands.

so I'm currently 7 tabling 5NL at 30 BB/100 over about 4000 hands but I have done nothing spectacular and feel like it is realistic to keep winning at near at this pace. How long before I can say that i beat 5NL?
 
thepokerkid123

thepokerkid123

Visionary
Silver Level
Joined
Jul 31, 2009
Total posts
917
Chips
0
You probably can already. It just becomes a more reliable statement the bigger your sample size gets.
 
D

DietDave

Rising Star
Silver Level
Joined
Mar 16, 2009
Total posts
20
Chips
0
I'd say about 10k is a good start, but 20k-30k will give a better idea. 100k is great but that takes a long time.
 
shrtstakatak

shrtstakatak

Rock Star
Silver Level
Joined
Jan 17, 2008
Total posts
203
Chips
0
If you want to calculate your winrate, >100k.


This really all depends on your standard deviation and how precise you want your conclusions to be.


When you say winrate, do you mean BB/100 for example? What kind of standard deviation would you consider acceptable? +5%, +10%?
 
No Brainer

No Brainer

Losing keeps me sane
Silver Level
Joined
Aug 27, 2009
Total posts
1,853
Chips
0
30BB/100 is running super hot. 4000 hands is not a large sample size by any means and your 30BB/100 winrate will definately drop substantionally by the time you hit 50k hands.

Basically to answer your question, more is always better. 50k will give you a decent sample to see how well you are doing at micro stakes.
 
joeaugie

joeaugie

Enthusiast
Silver Level
Joined
Jan 16, 2007
Total posts
66
Chips
0
Considering top pros go through 100K-150K break even stretches but boast a 2.5-3ptBB/100 winrate, you have to get in about 1 million hands to be certain you are within .5 of your winrate.

It's a tough one to answer completely, but for purposes of improving, I think playing 20K hands, then reviewing, then 20K, etc is a good way to go.

I know I've had 9K break even stretches at 5NL followed by 11K stretch with a near 10ptBB/100 winrate.

The game just has far more variance than most people realize, which is why bankroll management and emotional control are essential to success.
 
Starting Hands - Poker Hand Nicknames Rankings - Poker Hands
Top