|
|
|
November 20th, 2008
10:47 pm - Poker wonkery and more
I don't comment much on my poker research because who the fuck cares, but I have run some serious simulations using strategic templates.
Most poker players will tell you that a player cannot "beat the rake" at any cash game lower than 4/8. To beat the rake means to play profitably enough that, factoring in the money the casino "rakes" from each pot (typically 10% of the pot up to $3, plus a $1-2 side drop for jackpots), you can conceivably turn a profit from winning play. With the casino siphoning money from each pot, it's harder for a solid player to do so, and as mentioned, some believe it's impossible at the 3/6 level and below.
I tested a starting hand strategy devised by King Yao in his book Weighing the Odds in Hold'Em Poker, with a general solid postflop strategy... over 1,000,000 simulated hands at a typical 3/6 game that features loose, passive players with average ability. To meaningfully beat the rake, you must turn a profit of about $0.10 per hand. The Yao strategy lost about $0.12 per hand... which was still more profitable than the strategies of the other 9 simulated players, but still a losing strategy. With 1,000,000 hands, I had thousands of instances of each possible Hold'Em hand, a large sample from which to determine win rates.
The key is that certain marginal hands, like small pairs and low to middle suited aces, marginal hands that can be profitable but must be played right to turn an overall profit... lost anywhere from $0.15 a hand to $0.60. The worst Hold'Em hands had a baseline loss of $0.30-0.40 per hand, so anything beyond that indicated the hands in question were being overplayed. Pocket twos and pocket threes had losses in the $0.40-0.50 range. Strategies typically have you playing these for one bet in late position and looking to flop a set... but the Yao strategy had me raising with them in late position if 1st in. I can immediately tell this is a losing strategy, since they frequently get outdrawn on the flop. You spew two bets and then, unless you flop a set, you just lost two bets. With the suited aces, the Yao strategy also had me raising in late position when 1st in. Not only do they miss the flop most of the time (they really need to flop a flush draw to make money), but when you flop an Ace, anyone willing to call your raise likely has a stronger ace, beating you.
The Yao strategy is moreso meant for middle limit games at the 20/40 and 30/60 level, where players are smarter and more tactical. When you play a loose game where 4-5 players see a flop and call just about any bet from there, deceptive plays don't work as much. You need to take more of a drawing strategy, a la Omaha Hi/Lo, where you raise only for value and typically keep things cheap until you see you've got a big hand.
But really, what I'm looking for with these trials are leak hands: potentially profitable marginal hands that show an overall loss. Clearly, big hands like pocket aces still make big money overall. Some marginal hands do turn a profit. The key is to figure out HOW marginal your starting cards can be at this level, how aggressively you can play them, and still allow you to turn a profit. Yes, postflop strategy matters too, but most strategies postflop are fairly standard, and I always continue working to mold the programmed strategy to what I would do in a live game.
I think a starting hand strategy that is just right in its selectivity can allow a player to turn at least a big bet an hour (the typical goal of a winning limit Hold'Em player). The motherlode is to devise an overall strategy that allows you to average 3 big bets an hour. If you can do that... you might actually be able to grind a subsistent living out at 3/6. But of course, if you could do that, you'd be an amazing player for that level. And it wouldn't be the high life, and you still have to pay taxes on that money (well, if you're honest), but $18 an hour on average isn't bad. And if you can do it at 3/6, you can definitely do it at 4/8. I don't buy that it can't be done.
It will take a couple days, but I'm programming Lee Jones' beginner strategy from Winning Low Limit Hold'Em, with a solid postflop strategy per the book's parameters, before running a similar trial to see if this can beat the rake at 3/6. I'm not so much interested in fine tuning this strategy as I am in developing a baseline expectation for what a typical, predictably solid player can expect to do in a 3/6 ring game.
From there, I will dig Small Stakes Hold'Em by Sklansky/Malmuth/Miller off the shelf, program the tight games strategy from the book and see what it can do. This will require me to read it through again and learn the strategy back to front so I can program it right (it involves a lot of situational tricky play after the flop). Jones' strategy is pretty simple, and typically is marginally profitable, but SSHE is designed to give experienced low limit players a decisive edge. It's a tad dated and online players have since met adaptive resistance from other players... but this strategy can still work in typical loose passive casino and cardroom games.
From there, in both cases, I will examine the stats for leak hands, and see about fine tuning the SSHE strategy and whether that increases profitability. There will be a lot of statistical analysis and trial/error involved. Hopefully, I will find that the SSHE strategy isn't as dated as online players think it is and that it can still make a killing at a cardroom.
Also, I discovered from my simulations of active 3/6 games that turning a profit in Omaha Hi/Lo at the 3/6 level is fairly easy... just as Jeff Hwang said it could be in Pot Limit Omaha Poker. If I can fine tune my approach, I could well walk in the door at New Grove in Everett in the near future and play consistently profitable.
I want to get a lot of sleep this weekend. And I need to get a lot of writing done. I expect to pour forth a deluge of text during next week's long weekend.
Also, I contacted UTEMPS and told them to reopen my file, stating availability for 12/1. Notice I went to them first and not my old agency. Not a coincidence.
|
|
|
