Question for David Sklansky or other probability/poker theory experts

[filsteal]

[ QUOTE ]
[ QUOTE ]
Eh, not quite, but close I'd say. Typical standard deviation for 100 hands is around 50 BB/100. So even for 100k hands, the standard error of your winrate will be in the ballpark of 1.5 BB/100. A "true" breakeven player will run at -3 BB/100 or worse over 100k hands about 1/40 of the time. A "true" +1.5 BB/100 player will run at -3 BB/100 or worse over 100k hands less than 1/500 of the time. So while it's not impossible that you've been playing SLIGHTLY winning poker, it's pretty close to impossible that you "should've" been anything resembling a big winner over this time.

[/ QUOTE ]

Thanks, that was really helpful.

Can you describe your calculations?

[/ QUOTE ]

Sure.

The estimate of a standard deviation of your winrate of ~50 BB/100 hands is just a general approximation based on a lot of people's results.

Standard errors of averages (like average winrates) decrease in proportion to the square root of the sample size. So if you play 100k = 1k x 100 hands, then the standard error of the winrate for 100k hands will be the the standard deviation for 100 hands divided by sqrt(1,000), which is around 30. So 50/30 is around 1.5.

As for the probabilities of being above or below certain numbers, it's just based on normal distributions. You get results >2 standard deviations below the mean about 2.5% of the time, and >3 standard deviations below the mean about .015% of the time.

Hope this helps.

[Phone Booth]

[ QUOTE ]
Ok, this is stupid. The correct play against my playstyle would be to rr or limp/rr hands like ATo and 99 even if I made a 14bb raise PF. The villain was limp/calling this raise with hands as bad as T2o and 93o.

[/ QUOTE ]

No, one would rather give you a chance to spew chips on later steets than reraise and be called by dominating hands.

[ QUOTE ]
[ QUOTE ]

The correct way to play against your style from his perspective is to c/c light flop and turn and make a huge bet/raise on the river sometimes for value, sometimes as a bluff. What you've done is to make his style more correct.


[/ QUOTE ]

Only his calls on the flop were like, flop was KT4, 20bb pot, I bet 27bb, he calls with 82. Where is the sense in that? I'm not making that bet without at least a T, and he's calling with virtually every hand. He's not going to bluff later on either. He never made any bets bigger than 10 or 15bb, regardless of the size of the pot.


[/ QUOTE ]

The sense is in that this makes you bet a lone T compulsively, not realizing that very often he has you beat. No one's saying he's playing well, just that you're making his play better relative to how you're playing. And you're cherry-picking worst examples - I'm sure he's made a lot of correct calls against your questionable bets (bad players don't always have bad hands).

[ QUOTE ]
[ QUOTE ]

Btw, the basic way to play again the "paystation" as described is to keep the pot small until the river if you have a good hand that can be outdrawn and make a large bet there if you're quite a bit ahead of his calling range. Unless your hand is unlikely to be outdrawn, in which case you can start growing the pot earlier. If your opponent is consistently paying you off on the river and cares more about the absolute bet size than pot odds, growing the pot earlier in the hand only allows him to make correct decisions on later streets by accident.

[/ QUOTE ]

Why not take a bet where you're winning? He won't bet large in later streets no matter what. My play may be a little more swingy, but it has higher EV.


[/ QUOTE ]

Because you don't know that you're winning and you did mention that he folded some hands, which means it's possible that you've made a lot of bets while dominated (and others he called with defensible implied odds). A vulnerable made hand (middle/bottom pair) is never a huge favorite even if he's calling with a wide range. Sometimes he has you crushed; other times you're slightly ahead. And you'd never know if he improved. And you don't need him to bet large, if he's calling huge river bets with little.

[ QUOTE ]
[ QUOTE ]

Btw, I occasionally play like that "paystation" when I suspect that the table may adjust for the worse.

[/ QUOTE ]

There is no way you can play winning poker by playing like he did. Did I mention he lost 800bb in 350 hands? (and that does count that he took roughly 150bb from me). He was having good beats too.

[/ QUOTE ]

So you losing thousands of BB's over 100K is bad luck but him losing a few hundred BB's is skill? And of course if you're doing this for table-tilt, you won't keep playing that way. Also, note that other players did not adjust the way you did. If you think that sample is significant, do note that other tight players were able to take his money.

[soon2bepro]

[ QUOTE ]
No, one would rather give you a chance to spew chips on later steets than reraise and be called by dominating hands.

[/ QUOTE ]

Oh, you meant what was the perfect way for THAT player to adapt to me. That's different. I was playing a different game against him than against the others. He should've started folding way more hands, both PF and after the flop, and betting much more often in later streets. But he didn't do anything resembling that.

[ QUOTE ]
And you're cherry-picking worst examples - I'm sure he's made a lot of correct calls against your questionable bets (bad players don't always have bad hands).

[/ QUOTE ]

I assure you I'm not cherry picking. He called 5-14b raises PF with 90% of hands and then called pot sized bets on the flop with about 80% of those. He only started folding when the pot got large and on later streets. And even then he didn't fold nearly as much as he should've. I understand your idea of getting to the river to know where you're at, while keeping the pot small which is where he makes most of his mistakes. But he also plays much looser on earlier streets, and also if you know he's going to call so much that your bottom pair will probably be good by the flop, you might as well bet it while there's a relevant chance you're ahead. After that you can just check it down, as even if he makes a hand he usually won't bet it (not for any relevant amount anyway).



[ QUOTE ]
So you losing thousands of BB's over 100K is bad luck but him losing a few hundred BB's is skill?

[/ QUOTE ]

Again, he was having good beats and still managed to lose all that. And of course the other players took his money, anyone would've, but not by adapting, they did so just by playing their regular game. I should've probably taken over 1600 BB just by myself, but I kept getting sucked out on a lot of hands.

[soon2bepro]

Thanks a bunch for that filsteal, when I have some time I'll try to use that formula for my particular case.

[wizexel22]

soon2bepro,

I feel like I can relate. My friend is also a successful winning player and he decided to stake me and I had just started playing. What I had going for me was that I thought analytically and objectively, and I never went on tilt. So just on that he staked me and I won right off the bat for quite a while. Then like you, I just went on this run of just losing which was weird cuz I rarely lost at all before.

One day my friend said to me. "You're on tilt." I replied by saying I never go on tilt. And he said "no, not on tilt cuz of a bad beat, but your life is kinda on tilt...so your game is always on tilt." (This was true as when I was losing was also the worst and hardest period of my life....coicidence?) "Your image is all wack and you LOOK like you're gonna lose. That's why you keep getting called down and taking those beats. I can make the same bets and they might fold to me but call you." He said, it wasn't any one flaw in my game, but that my entire poker mindset was "on tilt".
The whole time I attributed my losing to bad luck, but when I really looked at my game, I realized there were things that I needed to change about my game that had little or nothing to do with poker theory. It was things like clearing my head and not trying to play TOO HARD, or win so much money (cuz the money was very important to me at that time as I wasn't financially stable), but I tried to just remember its a game and have fun by outplaying people. Things like adjusting my confidence and controlling the table image I project (and I'm not talking tightness or looseness, but how "scary" of a player are you? How do you make your image manipulate them into folding or calling. I thought most people should respect my tight/aggressive play, but I realized your average player calls or folds on gut instinct and I noticed my friend was a master of scaring people out of pots, not only using bet sizing, but also using mannerisms and psychological tactics.) I decided to revamp my whole perspective on the game and I even changed the way I dressed and the way I sat in my seat and even to being more friendly at the tables. (This has the effect of getting people looser since they are having a better time.) Also, when you are a presence at the table and aggressive you are "the man" at the table, whereas before I was the "quiet tight guy" being aggressive. This image has different effects on people at the table as well. And though you may not admit it, your confidence is very low right now. You will defend your play with poker theory, but the fact that said you feel like you will lose, and also that you're even asking this question shows low confidence. I can tell you are a smart guy and I think your confidence stems from that fact more than your poker game (at least, thats how it was for me.). So I think my best advice is to get out of that "losing" mentality. It affects your game more than you think. People in general aren't that smart. They aren't analyzing your play or calculating that you lay bad odds or keep track of your % of hands played, etc. They see you taking bad beats, so is their logic "wow, he is consistently going into the river as a heavy favorite?" .... HECK NO! These guys are laughing inside thinking "man! that guy is UNLUCKY!" and you know what? they keep playing at you with their trash cuz they think you are unlucky, and you know what? You end up getting even more bad beats cuz they just won't lay down their hands to you when they should and when they would've laid down their hands to someone else.
A quick example, I was playing shorthanded and I kept getting bad beat by a guy over and over and over and over again. It was really crazy. He would never lay down a hand and would hit gutshots and runner/runners all day. But he would always fold to another guy who was a terrible loose player who was also up big. One time, he told me "I laid down AT" which was top 2 pair. Exasperated I said "what the HECK do you think he had?!" , and he said "I think he had a set.". I almost had a stroke. He gave the loose terrible player credit for a set since he had was "winning" while I had no respect for my game simply cuz I was taking bad beats.

I know this sounds like a bunch of voodoo crap to you. I thought so too...till I changed my whole perspective and I haven't lost a session yet since. (Its been about 20 sessions).

Also, my other advice would be ...it seems you are playing the game from mostly a mathematical style...while accounting for adjustments here and there...it all seems a bit robotical. For example the paystation guy...it seems like you were always trying to lay him bad odds or bet for value. Always trying to but him in situations where he is constantly making mistakes. However, making many mathematical mistakes isn't the same thing (or nearly as bad) as making him make a few huge mistakes. I don't completely agree with Phonebooth, but one thing I do agree with is that you should've controlled the pot size to make it more manageable and this makes it easier to get out when you think he caught up, or make him make big mistakes on the river, when your read is better and he has no more cards to catch up. Or maybe try things like make a very small value bet one time. Then just push all in the next time, feigning a bluff. Sometimes you make small mathematical mistakes on purpose in order to induce people to make huge mistakes in the future.

Honestly, you've played enough hands and those numbers can't just be attributed to just bad luck, and SOMETHING is wrong with your game. But as you seem like a very intelligent guy and a solid player, my guess would be to look at all the factors involved in your poker game that doesn't necessarily have to do with strictly the numbers. Good luck.

[soon2bepro]

wizexel22: You've been very nice and I really appreciate that right now and in this issue. However, and though I did like to read your story, I don't find your suggestions of much use.

First of all I only play online, so table image has more to do with how you play your hands and at most how much time you take for each play. I guess it could be the nickname, since I started losing right after moving to that account, but I doubt this makes that much of a difference.

Moreover on this issue, (and It's funny because I keep hearing the same all the time from a very good friend of mine who's a consistent winner in high stakes even though he doesn't apply much theory to his game), it's not all about bluffing. If they don't respect your bets, then you bet for value. Of course if they already fold too much, you want them to fear you so you can make their mistakes even worse, but that's not always the case. It may be the case that in many situations you feel like the best way to manipulate your opponents is to get them to fold, but I don't think that's all there is to playing winning poker.

Also, I don't think I understand what you say about it being bad to make your opponents make horrible calls because then they'll outdraw you. If you have the winning hand, you want them to try and draw with bad odds.

I understand what you're saying about making the guy make fewer larger mistakes rather than more smaller ones. I just don't think I could be doing that. I think the strategy I was using maximized his mistakes, both in quantity and in quality. And all of this can also be mathematically calculated. You exchange small mistake for large ones. That's exactly what I think I was doing. I was playing like a maniac PF versus everyone else, clearly losing money to them, but in return I got to reap all the rewards of the paystations' mistakes, while maximizing them with my plays.

I understand that it is extremely likely from these numbers that I've been playing losing poker. It's just that aside from those numbers, I just don't see how that is possible. Not against the opposition I've been facing.

I am however worried enough to be trying to find major flaws in my game, but I just don't see any. Whenever I find a flaw it's always a very small leak. For example, the other day I was looking at my starting hand stats and VP$IP% (voluntarily put $ in pot %), when I realized I wasn't playing J8s nearly enough. It seems I had some sort of trauma with this hand back from when I started to play poker (I played only limit back then), and couldn't see the hand for what it really is, a decent two-gap suited connector. But this sort of flaw is very rare in me, and also very irrelevant.

[JavaNut]

Perhaps we should consider a different angle on this.

Someone earlier mentioned a std. dev. of about 50 BB/100 should be common.

A TAG player pushing good edges and avoiding too many coin flips will have a low std. dev.

A player pushing very single edge > 50%, even if in theory that should be +EV, will have an extremely high std. dev. or in other words will have a very high probability of losing for long periods of time. Losing 10 hands in a row when you are a 50.00001 favorite is 1-1000.

Another thing about your play against the calling station you have mentioned is that if you lose a hand to him, he will lose some of that money he won from you to the other players at the table before the next time you can get him HU. That means that you will not be able to win back what you have lost to him, but only a fraction of that amount. Firstly you will have to get your own stack up. In principle that could mean that you could be playing only +EV plays but as his stack is reduced between each play the overall play is -EV.

[JavaNut]

Consider this game setup.

You and villain each starts with $5. You always go all-in and flip a coin for the pot. If you have more cash than your opponent you get the excess returned from the pot before the coin flip. If either of you go broke the one broke will fill up to $5. After each flip villain will pocket $1 if he has cash.

Do you think that you can wins this game even if the coin flip was 50.1% in favor of you?

Is there anybody outthere willing to do the game theory slash maths thingy stuff calculations on this one?

[plexiq]

Hm, i would think you definitely win in this game.

Him pocketing 1$ should be a neutral action imo, all it does is changing the size of the next bet. You have a 50.1:49.9 edge on every single bet, so your EV should be 0.02% per bet.

I dont see how you could be losing here.

But due to the setup, the overall game will be extremely swingy for you. I think the concept is quite similar to the doubling strategy in roulette.

PS:
Simulating 100k runs of 100 coinflips per run yields an average profit of 0.021$/flip.

[El_Hombre_Grande]

The stats set forth in your post cannot be explained by a single session with a loose cannon that repeatedly drew out on you.

The probable explanation is that 70K is not that many hands, your game had/ has developed leaks, the games are somewhat tougher than the time period you took the 70K sample.

Additionally, if you play predictably, on one site, in NL especially, astute regulars with PT and PAHUD may be exploiting those leaks more than you think.