Using Math to Beat Poker

[RedeemerKing]

I was thinking of doing this and would like to get some opinions before I start. Can low limit hold em be beaten by using strict mathematics. Say with 4 people behind you you would raise any hand that is in the top 20% of hand possibilities. What about on the flop. If you take all the hands you beat (or will beat if a draw misses) and all the hands that your opponent could have that beats you. Take the ratio of those 2 numbers and make sure they are at least the pot odds. Now I know this will be off since you know your opponent wont have 72 offsuit and hands like that, but will your profit, if this works, make up for it.
Can you use just math to win. Only playing hands which are statistically more likely to be the best, given the number of opponents. Would you then be able to develop a chart that maps out every situation and what to do using the probability of your opponents hand.

This is hard to explain and I dont know if Im saying what I really want to say. Here is an example.

3 people are left behind you. You raise with Q8 since this is statistically likely to be the best hand. You get 2 calls.
flop comes 8,J,K rainbow
can you figure the chances of either or both of your opponents having a K or J or a straight draw and the odds of them hitting their straight and weigh this with the pot odds.

Would you then be able to develop a formula for every situation. You would know if 2 people are in with you the odds they have you beat at that point or will have you beat

This probably sounds like nonsense, but I think it might be worth it. So I guess what Im really asking is ignoring tells, position, type of opponents, and all the other stuff just using strictly math can you beat low limit hold em?

[chris_a]

[ QUOTE ]
what Im really asking is ignoring tells, position, type of opponents, and all the other stuff just using strictly math can you beat low limit hold em?

[/ QUOTE ]

First of all, it depends on how closely you model each situation with "strictly (sic) math".

As someone who has implemented various AI algorithms for poker, I can say that you'll gain a lot by modelling how your opponents betting affects your winning chances.

My opinion: If you did it the way you suggest, without regard to adjusting your winning chances when your opponents raise, you will probably lose in all but the smallest micro-limits. You'll probably break even at absolute best at medium limits or higher. Ignoring position is very detrimental. Ignoring opponents betting is horrible.

[PygmyHero]

I'm with Chris. I just don't see how the answer can be yes. Poker is a game grounded in mathematics, but that's only the beginning of course.

Maybe I don't understand the motivation here. Are you just trying to show that this could be a valid method of playing? Because even if it were profitable I would expect even a slightly above average player to show better results over any decent period of time.

Even if you did come up with a solid set of guidelines do you realstically feel you would have the discipline to hold to them on every single hand?

And of course your opponents won't hold to guidelines either. I defy anyone in this forum to write back that they've ONLY EVER drawn with the proper pot odds (or more generously, the proper implied pot odds - probably going to get a bunch of wise guy responses on that). So how would your 'system' account for the fact that players will occasionally (maybe even knowingly) take the worst of it?

It kind of seems to me like part of the weakness of this playing method would be that the only hands you would win are the ones were you are the pre flop favorite and then you CONTINUE to be the favorite as every board card is exposed (and the occasional successful draw). And in any event it also seems like this plan would make you showdown every single hand.

Again, I'm in complete agreement with Chris. Because you wouldn't be factoring in your opponents actions, playing tendencies, and positions, you would probably lose. Also, a perceptive player may get some sort of idea of what you're doing - and if that happens (no matter what system you're using), you're really in trouble.

[jpg7n16]

[ QUOTE ]
. . . Say with 4 people behind you you would raise any hand that is in the top 20% of hand possibilities. . .
Can you use just math to win. Only playing hands which are statistically more likely to be the best, given the number of opponents. . .
3 people are left behind you. . .
So I guess what Im really asking is ignoring tells, position, type of opponents, and all the other stuff just using strictly math can you beat low limit hold em?

[/ QUOTE ]
Your examples use position, and you ask if you can avoid it? Hmmm...

Position can help determine odds. Knowing your opponent can help you know whether or not their hand is in the upper limit of the range of possible hands for your formula...

an example: you know the Big Blind only calls/bets with high pocket pairs (he'd be an idiot to do that... but say you know him and he does) now when a flop comes with k 9 3, and he is in the pot, you have a better mathematical advantage in that you know he either has AA, KK, QQ, JJ, TT, so you could avoid considering him being on a straight or flush draw. You have AK. Therefore the math is against him having AA and KK since you have one of each of those. 3/5 times he would have Q's J's 10's, and your K's A kicker would beat that so you should bet (or induce a bluff depending on how aggressive he is). Even more so since you lower his chances of having AA-KK, increasing your chance to win. You get what I'm saying?

The profitability of your bets are even dependent upon the playing styles of your opponents, so the math of profitable plays needs to know that info.

Poker is a game of math and psychology combined. You use psychology to affect the math sometimes (bluffing math don't count).

But position, player profiles, style of play, size of bet, number of opponents, tells, etc. ALL help with the math, and it should be impossible to do math without some of them. Ie. position - you can't determine the odds of your hand being the greatest if you don't know how many people are in the hand.

So theoretically, you could win without some of them (ie. tells) but the rest should be key to whether you win or not.

[gergery]

Yeah, and you figure out which hands are statistical favorites in those situations, you could bunch them together and give guidance for what situations they could be played in. You might call those “groupings”. I’d guess there might be roughly 8 Groups of hands……..

[Little Fishy]

you can buy computer simulations where you play against the computer. this software is all math. I haven't seen any software that is capable of playing multiple opponents at the same time however I'm sure that with enough computer power and programming skill and time it would be possible. what's even more scary to think about is that it is theoretically possible to create programs that can adjust play given the play (ie stats, hands % played, % raised etc) of opponents. if some1 did develope a program like this and put it to use on a paysite they would become rediculously rich. they could have dozens of thesse programs running syphoning $$ off of players who assumed that they were playing other people.

how's that one to make you distrust internet gaming.

-little fishy

[gambool]

[ QUOTE ]
So I guess what Im really asking is ignoring tells, position, type of opponents, and all the other stuff just using strictly math can you beat low limit hold em?

[/ QUOTE ]
No, unless all of the other players were also ignoring tells, position, type of opponents, and all the other stuff..

If you find such a game, let me know! [img]/images/graemlins/grin.gif[/img]

[fluff]

[ QUOTE ]
[ QUOTE ]
So I guess what Im really asking is ignoring tells, position, type of opponents, and all the other stuff just using strictly math can you beat low limit hold em?

[/ QUOTE ]
No, unless all of the other players were also ignoring tells, position, type of opponents, and all the other stuff..

If you find such a game, let me know! [img]/images/graemlins/grin.gif[/img]

[/ QUOTE ]

www.partypoker.com

[einbert]

What you are talking about is assigning your opponents a range of hands and then acting according to your estimated pot equity based on that range.

Example, someone with VPIP of 50 from BTN-3 limps from BTN-3 and it's folded to you on the button. You have J9s which has 55% equity against the range that includes the top 50% of hands, so you raise.

While this strategy is relatively simple to program for most preflop situations, estimating hand ranges for your opponents after the flop and acting correctly based on those ranges is far, far more difficult after the flop. Too much information to interpret and consider and the situations are just too dynamic.

You could make a nice little preflop bot though, that could do some of your work for you ^_^.

[Brainwalter]

Are you making a bot? Sure sounds like it.