winrates at 2/4 party, pacing question

[Jdanz]

not to start a penis measuring contest but after not playing limit for quite some time in liu of other games i came back to limit this summer and am wondering what my goal should be for winrate at party 2/4, i feel like i'm playing well but i don't know if the results are statistically significant yet or about where i should be.

so just checking up and i'd like to see how other's on the board are doing.

party 2/4 - poker tracker - 6.5k hands 2.27/100

-JDanz

[JTG51]

You're doing well. Some are doing a little better. A few are doing significantly better. Almost everyone is doing worse.

[Jdanz]

so what do you see as a goal for a very good but not great powekr player?

-JDanz

[BigBaitsim (milo)]

2.5/100 is pretty good.

What do you think would be the pinnacle of 2/4 win rates that would be sustainable? Assuming good game selection, expert play etc - so we have something to shoot for before moving on to bigger things

[bisonbison]

Obsessing about win-rates is results-oriented thinking.

[SpaceAce]

[ QUOTE ]
so what do you see as a goal for a very good but not great powekr player?

-JDanz

[/ QUOTE ]

2.5 or so is a good win rate. I can tell you that the PartyPoker $2/$4 tables are beatable for much higher than even 2.5, though. I have many, many times more $2/$4 hands than you do in my PokerTracker database and my per 100 win rate is astronomical. My win rate at PartyPoker $2/$4 Texas Hold'Em is far and away the best of any game at any limit for me.

SpaceAce

[SpaceAce]

[ QUOTE ]
Obsessing about win-rates is results-oriented thinking.

[/ QUOTE ]

That's ridiculous. I don't see how his two posts in this thread count as "obsessing" and anyway, most of the people around here play to win. These forae and 2+2 Publishing would not exist if people were not concerned with results because no one would be trying to improve their game if they didn't want to win.

SpaceAce

[MarkD]

Your winrate converges so damn slowly that it isn't something you should "shoot for".

[MarkL444]

last 18k hands (since i got p tracker)
4.14/100
July almost 4.9k hands
5.02/100

im going to go out on a limb and say fold ATo, KJo, QJo, QTo more.

[bisonbison]

1) I didn't say that the poster was obsessing about win rates, I said that obsessing about win rates is results-oriented thinking. It was a warning, not a condemnation.

2) Clearly, poker is popular because it's possible to win money, and these forums are popular because people feel it will help them play better, more profitable poker.


That having been said, the conscious control you have over your long term win-rate is not exercised by worrying about your win-rate. That control is exercised through focus on better play.

We've all had questions about what a good rate is, what a bad rate is, what a rate that you should move up because of and so on and so forth, but it's still putting the cart before the horse.

If you're worried about your win-rate over 6.5k hands, you need to read up on variance and modify your expectations of accuracy.

yeah that was kinda the wrong term, but it does converge a bit quicker when your playing 2k hands a day /images/graemlins/smirk.gif

[sfer]

I would guess a board average is around 2.5-3.5 BB/100. Some are substantially higher and lower.

I'd worry more about playing well (and posting hands) and ignore your winrate until you hit at least 10K hands. Even better at like 20K hands.

[Bob T.]

What do you think would be the pinnacle of 2/4 win rates that would be sustainable? Assuming good game selection, expert play etc - so we have something to shoot for before moving on to bigger things

You should probably move on to bigger things before you reach the pinnacle of 2-4 party poker holdem.

[bisonbison]

I would guess a board average is around 2.5-3.5 BB/100.

I think this is high. I think 2/3rds of the winning players on this board probably fall between 1-3 BB/100 in the long run. Maybe 5-10% of people here could expect to be above that. Course, I 4-table habitually, so my perceptions may skew towards the lower end.

[tech]

If you are above 2BB/100, you are doing quite well. Trying to compare your win-rate to others is quite difficult because there are so many factors that affect it (multitabling, small samples, running good/bad, chest-puffing /images/graemlins/smile.gif, etc.).

[lefty rosen]

4 big bets an hour is about the best you can do in an online game. Stupid Party 2/4 tables and stupid Pacific tables these win rates can be maintained forever in theory...

[Jdanz]

it would be results oriented thinking if i hadn't specifically asked whether 6.5k hands can be considered significant. Worrying about the results rather then the -EV of a particular play is poor thinking, worrying about statistically signifcant long term results is just good policy.

-JDanz

[Blarg]

I have over 20k hands and still see substantial fluctuations in my win rate. You really do need a pretty hefty sample size before you get too settled on where you really stand. I'm thinking that at 20k I still have quite a ways to go before one weekend or one week's playing won't have that much affect on my win rate. Perhaps in terms of skill, too, but mainly just in terms of pure numbers of recorded hands.

[Tosh]

[ QUOTE ]


party 2/4 - poker tracker - 6.5k hands 2.27/100


[/ QUOTE ]

Statistically insignificant.

[jslag]

[ QUOTE ]
Statistically insignificant.

[/ QUOTE ]

For those of you guys that seem to have a grip on their statistics, what is the consensus on a good sample size (# of hands) to properly evaluate your average win rate (BB/hour or BB/100 hands)?

I have been killing Party 2/4, but consider myself to be running hot. I'd like to know that I'm a consistent winner at that level and then move up rather than 4-tabling 2/4. My interest is getting to big bet poker (10/20 on up). I do jump up to 3/6 now and then and usually have winning sessions, though not as profitable as putting time into 2/4 tables.

Obviously, I need more hours at 2/4, but I'm interested in how I can determine on a statistic level that I'm ready for a level change. I have the required bankroll for 3/6 already, but it would be nice to have my statistics line up as well.

Blah... did any of that make sense? Just looking for the generally accepted numbers (e.g. - "20k hands, 3 BB/hr recommended to move up")... seems to me there are a lot of math wizards here.

Thanks.

[Alobar]

I'm stoopid on the math end so my answer isn;t scientific. But if you've got a good winrate (say at least 2bb/100) after 20K hands at 2/4 and the bankroll for 3/6 I think it would be pretty safe to say you arnt going to go bust at 3/6.

My winrate at 2/4 was over 4BB/100 after 40K hands, 10K hands later it was 3.25/100. So even after a pretty desent sample size you can expect to see the numbers change

[Octopus]

I'll take a shot at this.

To calculate the standard deviation of your win rate per hand, you can divide the standard deviation per hand by the square root of the number of hands. (We normally talk about win rate/100 hands and standard deviation/100 hands, so divide the standard deviation by the square root of the number of hands you played over 100.)

For example: In the orginal post, Jdanz says he has 6500 hands at a win rate of 2.27/100. Using my own 2/4 standard deviation of 16.55, the standard deviation on this estimate is 2.05, which means that there is an 86% chance he is playing at zero BB/100 or higher and a 73% chance he is playing at 1BB/100 or higher. Statistitians would say something inscrutible like, "we can not reject the hypothesis that he is a losing player." He is probably a winning player, but we just do not have enough data to be confident.

How much data do we need? Well... Assuming your win rate standard deviation per 100 hands is 16BB, you would need 25,600 hands in order for the standard devation on your average win rate to be 1BB. That is, if your win rate after 25,600 hands was 2BB per 100 hands, you could be 97.5% confident that you were a winning player (and similarly, 97.5% confident that your win rate was below 4BB/100). If your win rate after that many hands was 3BB/100, you would be nearly 99.9% sure you were a winning player. If it were only 1BB/100, then you would be about 84% confident you are a winning player.

To get the standard deviation on your win rate down to one small bet, you would need slightly over 100k hands. You can be 95% confident that your true win rate is within 2 standard deviations of your estimated win rate. (This is a typical significance level used in statistics.)

As for when to move up, others can answer that better than I. I would only say that you can move up a lot earlier if you are willing to move back down for a while should you do poorly. Holdem is sufficiently volatile that you have to play a fairly large number of hands just to tell if you are a positive EV player. (At a B&M, 25,000 hands is like 40 hours a week over 3 months.) The best advice I have heard is that you are ready to move up when you can readily identify the mistakes your potential opponents are making and you have the bankroll to handle the fluctuations.

[sthief09]

[ QUOTE ]
not to start a penis measuring contest but

[/ QUOTE ]


[ QUOTE ]
last 18k hands (since i got p tracker)
4.14/100
July almost 4.9k hands
5.02/100

im going to go out on a limb and say fold ATo, KJo, QJo, QTo more.

[/ QUOTE ]



I guess you're right and everyone else is wrong