|
|
| Author |
Message |
Holgininho
Joined: 01 Jan 2007
Posts: 468
Location: Essen
|
 Posted: Fri Mar 23, 2007 7:10 am Post subject: |
 |
|
I have a question that relates to the topic, I think: Given that stats aren't really reliable because of variance - how can I tell the difference between being in a downswing or being a losing player (or between an upswing/being a winning player)?
I have 28k hands in my database, mostly short-handed, limits ranging from 0.5/1 to 3/6. Overall I've been winning 1.7 BB/100, but over the last 10k hands I've been losing nearly 1 BB/100.
I've been feeling unlucky over the past weeks, but then maybe I have only been really lucky before. It isn't really hard for me to cope with losing, I feel confident when I play, and, thanks to bonuses and rakeback, my bankroll has been growing regardless of my losses.
But it still is a nagging question, especially because I have moved up rather fast: Is it a downswing, or am I a losing player? Surprisingly I was most successful at 2/4 and 3/6 in the last weeks, while losing big time at 1/2. But then this doesn't say much at all, as I'm only talking about 3 or 4k hands on each limit.
So, what can I do against my confusion - apart from trying to improve my game? |
|
| Back to top |
|
 |
Piemaster
Author of THE POKER MINDSET
Joined: 15 Jan 2004
Posts: 6931
Location: London
|
| Posted: Fri Mar 23, 2007 8:55 am Post subject: |
|
|
Holgininho - I hope you don't mind, I split this into a new topic because it isn't really the same topic as the other thread, but should provide an interesting discussion on its own.
It can be difficult to separate the effects of luck and the effects of bad play. What you can say is that the more stats you have, the more likely it is that those stats are an accurate reflection of your ability. For example, if you have been running at -1BB/100, then the number of hands that is over will determine how you can interpret that run.
5,000 - Virtually meaningless
10,000 - A concern, but could easily just be variance
20,000 - Likely you are a losing player, but could be a very bad downswing
30,000 - Almost certain that you are a losing player
Of course, the exact numbers make a big difference too. If you were running at -2 BB/100, then you be confident that you were a losing player a lot faster. -0.1 it would be very difficult to be sure even after a lot of hands.
In your case I think there is nothing to worry about. In the good old days, 10,000 hands was considered a reasonable indicator of your ability. But these days games are tougher, players are more aggressive and more players play short-handed (you didn't say in your post whether your gameswere FR or SH). 10,000 hands is now not really considered enough hands to know either way. The fact that you are winning at a good rate over nearly 40,000 hands makes it likely that this is just variance. Providing you are not on tilt (and from your post it doesn't sound like you are), you should be fine. |
|
| Back to top |
|
|
the_hawk
Chelsea FTW!
Joined: 13 Jul 2005
Posts: 4340
|
| Posted: Fri Mar 23, 2007 9:49 am Post subject: |
|
|
Just to put some numbers to this from normal distributions.
Firstly, consider your entire career so far. At 1.7BB/100 net over 28k hands you can be, by my reckoning, 85%+ confident that you are a winning player overall. This depends on your "true" SD, which you can look up in PT but I would guess is something like 20-25BB/100 for a SH player. At an SD of 20 you should be 92% sure you're truly breakeven or better, and at SD of 25 you are 87% confident of being a true winner. If your SD goes down your confidence goes up, and vice versa.
Now, what are the corresponding figures for a 10k stretch where your "apparent" earn is -1BB/100? It emerges that you can be about 33% confident that you are truly better than breakeven over such a period with these numbers. That doesn't sound very good, and if it was all the evidence you had then you'd be a likely true loser. Happily this isn't the case here.
If we vary the calculation a bit again and assume you're a true 1.7BB/100 winner, we can calculate the chances of you doing that badly over any given 10k stretch, which is probably the closest we can get to answering your specific question. It turns out to be about 10-14% depending on the SD you use, so whilst unlikely, by no means impossible - and hardly inconceivable that you'll get one such run like that within a span of 30k hands overall.
Hopefully the numbers don't distract too much - but the cliff notes are that you're probably OK. |
|
| Back to top |
|
|
Holgininho
Joined: 01 Jan 2007
Posts: 468
Location: Essen
|
| Posted: Sat Mar 24, 2007 5:25 am Post subject: |
|
|
| Piemaster wrote: | | Holgininho - I hope you don't mind, I split this into a new topic because it isn't really the same topic as the other thread, but should provide an interesting discussion on its own. |
Of course I don't mind. 
Thank you both for the helpful answers. It really is a bit complicated, but throughout my short poker career I've had the suspicion that my winnings were due to an upswing and my losses due to bad play. And just when I was fairly confident I had a decent sample size to show I might really be winning a downswing started to hit. |
|
| Back to top |
|
|
Bugsbunny
Wascally
Joined: 07 Apr 2004
Posts: 7626
Location: Drinking Carrot juice
|
| Posted: Sat Mar 24, 2007 8:31 am Post subject: |
|
|
| Holgininho wrote: | | Piemaster wrote: | | Holgininho - I hope you don't mind, I split this into a new topic because it isn't really the same topic as the other thread, but should provide an interesting discussion on its own. |
Of course I don't mind.
Thank you both for the helpful answers. It really is a bit complicated, but throughout my short poker career I've had the suspicion that my winnings were due to an upswing and my losses due to bad play. And just when I was fairly confident I had a decent sample size to show I might really be winning a downswing started to hit. |
No - you have it backwards. Winnings are ALWAYS due to your superior god-like poker skillz. Losses are ALWAYS caused by absurd bad luck. because we have angered the poker gods somehow (as in not offering them a proper sacrifice or something).
Seriously Pie has it just about correct, although the exact numbers depend on your Standard Deviation\, and even at 30,000 hands if you're losing at a rate of -1 BB/100 there's still about a 4% chance that your true earn is greater than 1 BB/100 (if we assume a SD of 20, which would not be atypical for a 6 max player). Once you get to 50K hands though, if you're running at -1BB those chances drop to 1% - but there's still about a 4.5% chance that you're true earn is > .5BB/100, and a 13% chance that you're actually break-even or better.
And just to show how much SD can effect things (still using 20) - even at 200K hands, if you're losing at -1BB/100 there is STILL a better than 1% chance that you're actually a break-even player or better. Now 1% isn't a lot, and break-even isn't what we shoot for, but the point is that even after that many hands you still can't be 100% certain that you're not actually a winner (although you can be near certain). |
|
| Back to top |
|
|
Misunderstud
1K Club
Joined: 10 Jan 2006
Posts: 1661
Location: Here, stupid
|
| Posted: Thu Mar 29, 2007 3:58 am Post subject: |
|
|
| Holgininho wrote: | | Given that stats aren't really reliable because of variance - how can I tell the difference between being in a downswing or being a losing player (or between an upswing/being a winning player)? |
Given that stats aren't really reliable, why has everyone used them to answer your question?
If you find yourself frequently thinking 'That was a bad beat, I had the odds to make that play and he didn't', 'I know I played that hand correctly because that's what it says in ITH', 'Those players are making obvious mistakes' etc. etc. you are probably in a downswing.
If you find yourself thinking 'I'll take a chance on this', 'Damn, I didn't see that coming', 'I shouldn't really have been in that hand', 'That was lucky - about time!', 'Oops!' etc., you are possibly a losing player.
I divide my poker career into two sections: (a) The first six months, during which time I started knowing nothing, progressed to thinking I knew everything, nearly went bust and decided I really did know nothing after all. (b) The last nine months, when I went back to basics and tried to do things properly.
Some more unreliable figures (we can't resist them, can we?  ): During period (a) my stats at .10/.20 were 2.27BB/100 over 18K hands and at .25/.50 0.36 over 12k; during period (b) I won 0.61 over 30k at .10/.20 and 2.2 over 23.5k at .25/.50.
I think I'm playing well enough at the moment to beat .50/1 quite easily, but I'm winning only 0.75 over 11k because it took me 9k hands to break even after a bad start.
So I'd suggest you forget the stats altogether, concentrate on playing each hand the best way you know how and then ask yourself if there is a better way.
Moving up fast is not necessarily a good thing because a good run can pitch you into a higher limit before you're ready to be there. If you feel comfortable at the table and confident that you play as well as most of your opps and better than many, then you are probably in the right place.
Really, though, if you're intelligent enough to ask the question in the first place, then you're probably intelligent enough to know the answer already. |
|
| Back to top |
|
|
Holgininho
Joined: 01 Jan 2007
Posts: 468
Location: Essen
|
| Posted: Thu Mar 29, 2007 3:06 pm Post subject: |
|
|
| Misunderstud wrote: | | Holgininho wrote: | | Given that stats aren't really reliable because of variance - how can I tell the difference between being in a downswing or being a losing player (or between an upswing/being a winning player)? |
Given that stats aren't really reliable, why has everyone used them to answer your question? |
Good news. My overall win rate is up to 1.88 again. 
You make some very good points, the only problem is that over a period of 10k hands (for me this is about one month of playing) I trust my thoughts and feelings even less than those stats.
When I lose 100 BB in one session, I don't feel good, but my thoughts are more or less exactly like you wrote most of the time (of course there can be some "Oops!" involved as well). I can deal with that.
But if you are running at -1BB/100 for a whole month things become difficult. You can be quite sure you are not the worst player at the tables, you might even be winning if it wasn't for the rake (as I wrote: My bankroll had kept growing despite my losses at the tables). But how can you be sure you belong to the small minority of players who are able to beat their opponents AND the rake?
Even as a good player you aren't supposed to win more than 1 or 2 BB per 100 hands. Therefore I figure it doesn't take a big leak to turn from a marginal winner into a marginal loser. And if your leaks aren't as obvious as chasing weak draws with insufficient pot odds and your sample of hands is not yet big enough to be reliable, it becomes hard to tell to which group you belong. |
|
| Back to top |
|
|
Misunderstud
1K Club
Joined: 10 Jan 2006
Posts: 1661
Location: Here, stupid
|
| Posted: Fri Mar 30, 2007 6:42 am Post subject: |
|
|
| Holgininho wrote: | | Good news. My overall win rate is up to 1.88 again. |
Mine's down to 0.5 again 
OK, I see your point. It's impossible for me, and probably you or anyone else, to determine with certainty whether you're a winner or a loser on that size of db, so let's fall back on stats again.
Stat 1: My db contains over 21K players, which is probably large enough to be representative. Of those, about 60% are losers and 40% winners, whichever criteria I use, so the % of winners doesn't appear to be THAT small.
What % of players are serious enough about their game to participate in forums such as ITH? I don't know, but I'll wager it's less than 40. So it seems reasonable to assume that everyone who contributes to these discussions is either a winning player or at least on their way to being one. You obviously understand how the game works, even if, like most of us, you haven't mastered it. That fact alone probably makes you a winning player.
Provided that your BR is adequate for your limit, and not so large that you are blase about the monetary amounts involved (you wish!), its size is largely irrelevant. (Of course it is possible that you, like any of us, could be among the minuscule minority who go broke through no fault of their own, but you're probably more likely to be run over by a bus, so I wouldn't worry about that either.) The rake is also irrelevant unless it's ridiculously punitive - your earn rate takes that into account. So
Stat 2: You are -1BB over a month, or 10K hands, so you have suffered a 100BB downswing. That is not unusual. Most people's Poker Patterns charts exhibit a mountain-range quality rather than a steady incline - win 110, lose 100. 30K hands is really very little to go on. I managed a 100BB downswing even at .5/.10 during period (b), where I was beating the game by 5+ BB/100 (though there was no rake at the time). Theoretically, I wouldn't be unduly concerned to lose 150 or 200BB at FR (though I would be perturbed, and probably start panicking a bit in reality). Short handed (which I don't play), I believe the swings can be much bigger.
Some other things you could try:
If you're playing 10k hands a month you must either be multitabling or playing really long sessions. Your earn rate will always fall slightly if doing either, so try single tabling/shorter sessions for a while.
You always sit down with confidence (so do I), but in a downswing that confidence is much more susceptible to evaporation (oh, oh, here we go again). When all your starting hands begin to look the same, it's time to bale.
I don't know how you use PT, but I usually display hands from the limit I'm at. That can get depressing. If so I reset to display from today and treat that as day 1. The past is the past and I leave it there. This is just psycholigical nonsense, but I find it helps.
Finally, let me answer your original question with some more:
(A) If the answer is 'You are a losing player', what will you do next to improve your game?
(B) If the answer is 'You are a winning player', what will you do next to improve your game?
Are the answers to (A) and (B) the same? |
|
| Back to top |
|
|
|
|
