Update as a beginner...

[AKQJ10]

[ QUOTE ]
I define a winning player as someone who can win regardless of the level they play because they possess the necessary knowledge and skill to beat the game.

[/ QUOTE ]

But surely there must be some limits to that "regardless" in practice, no? Obviously I will likely never be able to sit down with +EV in the Big Game at the Bellagio, but that doesn't mean I can't be a winning player overall in the games I'm likely to frequent. On the other hand, I grant you that beating the Foxwoods $2/4 or Pacific 5c/10c isn't exactly a career-long goal.

There must be some middle ground in there where you say, "I'm a winning player in this game, but not yet at this game." Right?

[SheridanCat]

[ QUOTE ]

There must be some middle ground in there where you say, "I'm a winning player in this game, but not yet at this game." Right?

[/ QUOTE ]

This is certainly true. Moving from the Bellagio 4/8 to the 8/16 to the 15/30 is a big change and each game requires some adjustments. Same for most locations, online or live, when making a significant move.

I think the process of evaluating your winningness is ongoing - even after you've found a comfortable and profitable level.

Regards,

T

[pzhon]

[ QUOTE ]
what is an adequate sample size to determine(maybe not determine, but get a feel) if your are a winning player? We all know that 5k is not enough...how about 15k? 30k?

[/ QUOTE ]
Actually, it is possible to conclude that you are a winning player after 5k hands. It's just unlikely that you have enough evidence at that point.

These discussions frequently degenerate into anecdotes of big swings and increasingly wild estimates of how many hands are needed. It's better to look at some math.

In limit, the standard deviation is typically about 15 BB/100. (In limit 6-max, 17 BB/100.) The standard deviation of your win rate after n*100 hands is about 15/sqrt(n) BB/100. A rough 95% confidence interval is 2 standard deviations in each direction about your observed rate. If you are 3 standard deviations above breaking even, you have strong evidence that you are a winning player.

If you win 10 BB/100 over 5k hands, you are about 4.7 standard deviations above 0. Although a player winning 4 BB/100 would have to run very well to win 500 BB over 5k hands, a break-even player would have to be much, much luckier. You can't trust the 10 BB/100 observed rate, but you can reasonably conclude both that you have been on a very lucky streak and that you are a winning player.

When I started playing NL 100, I won 14 BB/100 over my first 6500 hands. My SD was about 45 BB/100, so the SD of my win rate after 6500 hands was about 45/sqrt(65)~5.6, and I was ahead by 2.5 standard deviations. At that point, I stopped worrying about whether I was winning at NL 100, though I did not have confidence in the 14 BB/100 figure.

[AASooted]

[ QUOTE ]
Here's a question though...what is an adequate sample size to determine(maybe not determine, but get a feel) if your are a winning player? We all know that 5k is not enough...how about 15k? 30k? Where do we begin getting "an idea" of how we are doing?

[/ QUOTE ]

See this thread for information on how to determine confidence intervals for your win rate. If you have Excel, you can enter the formulas and plug your numbers in.

[4_2_it]

Absolutely there are limits. I guess I tend to get a little anal sometimes [img]/images/graemlins/smile.gif[/img]

[ QUOTE ]
So before online poker, a pro had to play over a year to know if they were a winning player*?

[/ QUOTE ]

yeah...even a year of playing live full time for 50 weeks would only give you a maximum 60k hands...(30/hr*40*50)