Poker Tracker stuff - how many hands before I can assume error?

[RiverTheNuts]

Well my top 5 profitable hands are:
AJs - 4.07 BB/hand
AKs - 3.1
KK - 2.93
AA - 2.87
AQs - 2.5


My bottom 5 are:
JJ - (.95)
98s - (1.37)
ATs - (1.48)
87s - (1.65)
QJs - (2.18)

Is this pretty normal for someone who limps these into too many non multiway pots and then plays them too aggressively when they hit?

This is after 15,000 hands in PT, which brings me to my main question: about how many hands do I need to normalize the values to where I can start assuming that it is operator error and not just variance? For example Im making $$ with 99 and TT, but not JJ, could be because I overplay JJ alot.

Any and all input would be great

[bugstud]

[ QUOTE ]
Well my top 5 profitable hands are:
AJs - 4.07 BB/hand
AKs - 3.1
KK - 2.93
AA - 2.87
AQs - 2.5


My bottom 5 are:
JJ - (.95)
98s - (1.37)
ATs - (1.48)
87s - (1.65)
QJs - (2.18)

Is this pretty normal for someone who limps these into too many non multiway pots and then plays them too aggressively when they hit?

This is after 15,000 hands in PT, which brings me to my main question: about how many hands do I need to normalize the values to where I can start assuming that it is operator error and not just variance? For example Im making $$ with 99 and TT, but not JJ, could be because I overplay JJ alot.

Any and all input would be great

[/ QUOTE ]

It only takes one hand that drops/gains you a stack to massively impact your stats...posting hands is a better cure than waiting for PT to tell you things. However, you might be overplaying JJ. Post 'em! /images/graemlins/grin.gif

[excession]

As a broad rule I use 20 instances for an indication and 50 for a 'read'.

Certainly if a 'core' hands is losing oevr the last 20 times you have it, it's worth replaying in PT to see what is going wrong - if it's bad beats then fair enough, but it's better to check early than later if you get a flag that something might be wrong.

For marginal hands like suited connectors you can filter them in PT to show profit from EP and LP and whether limped, raised or cold-called. You may need a bigger DB than 15k to get enough statistical significance here though - I'd say maybe 30k hands or so..