Monday, November 4, 2013

Exploitative Poker

I listed 4 fundamental strategies. You may be able to play a lot more hands a lot more aggressively in reality. Or you might play a lot more hands more passively. That is, provided you can avoid traps, find profitable bluffing situations, use implied odds and pot odds, shove all in at moments where you have a good sense of equilibrium, and good understanding of opponents, chip stack ratios to pot size, and many other variables.

The thing about playing MORE hands, is the margin of error is much less. You have to be able to make a lot of marginal and borderline decisions if you use more hands, and in most cases, it simply isn't worth it. At least, not unless you are playing your OPPONENT, rather than your cards.

In most cases, I find it easiest to do both. Play opponent's by looking at his range of hands, come in with a tighter range of cards, and proceed much more aggressively.

I will not cover adjusting your initial raising hand percentage because so much of that is contingent upon your postflop strategy and opponent.

But I will look at ONE possible way to handle it. Since you likely won't know if opponent is playing  particular range or not based upon just a few hands, or if perhaps he has planned to switch strategies, it is best to proceed with the assumption that opponent is tighter than advertised. Keep track of his raise percentage and raise 3x his raise with HALF of the hand range that he plays. Technically your average handrange that you play will be on average such that you have a 50% hance of having a better hand then him and could theoretically raise with the entire range with the lower half being a "bluff" and the upper half being a "value bet" over his average range. But to be safe, raise with HALF the range. Then just to potentially handle a 4 bet, the "baseline" assumption is that he will only raise if his range is better than yours. In this case, we are going to assume he ALSO is conservative and raises with the best 50% of hand ranges and calls when he has pot odds over the range, with some strong hands occasionally calling and weak hands occasionally raising to mix up his range so you don't have much of an informational edge. What this means is that when  he 4bets, you should 5 bet enough to break even. If you have his 4bet range, just 5bet with HALF of the range as he does, but if not, instead just raise enough to break even if he were to bluff. Technically, this means you must 5bet with 1/3 of your hands since you also need to be able to do enough to keep up with the blinds, but I go for 1/4 in many cases. In other words, if opponent raises 20% of hands, you should 3bet with 10%, and fold 3 times and on the 4th, 5bet with the assumption that either:
1)You gain everything you lost from folding back if opponent folds, or
2)Get all your money in as a substantial favorite to win a large amount.

Lets take for example if you have 100 big blinds.
he raises to 3, you reraise to 9, he 4bets to 27 and you push/5bet for 100.
Since you will be losing 9 big blinds 3 times for a loss of 27, you must either take down the 27 big blinds 100% of the time OR be called all in with 27BB of equity (or a combination of both that produces an average of 27BB equity.) In other words, in a 200BB pot, you need to on average be 63.5% to win against an opponent who always calls to generate 127/200=.635 (100 back that you put in plus 27). It gets tricky when evaluating normal opponents who may call a particular percentage of the time but can be done. Regardless, the simple rule of thumb assumption is close enough for now.

So add in additional raise with HALF the hand range opponent bets under all conditions. It may be more profitable to make standard all in bet with a wider range (about 60% of opponent's range) while calling should likely only be done with about 40% of opponent's range.

The advantage will come when opponent calls.
If opponent 4bets, 5bet with HALF of his 4bet range.
If opponent calls your 3bet, assess his hand accordingly:
raise % * (100%-fold to 3bet)=hand range.
i.e. If opponent raises 40% of the time and folds to a 3bet 60% of the time, you can assume he continues when facing a 3bet with 40% of his initial 40% range. That is .40 of 40% or top 16% range.


On the flop:
The information you want on the flop is percentage of flops that opponent leads out with a bet on flop, percentage of the time they fold to a continuation bet on flop, and percentage of the time they fold to a turn bet on flop. Also percentage of the time they check raise flop, and check raise turn. Some may like to use river stats but you likely will have too small of a sample size and it won't get to the river very often.

If they fold 50% of the time on the flop or more, you can bet up to a full pot sized bet every time and be profitable. But it may be MORE profitable if you wait until the turn. There is no good way to have stats to know how well opponents adjust to bet sizes on a limited sample size. Ideally you should be able to bet an extremely small amount and still get opponent to fold just as often for maximal exploitation. Ideally you should be able to bet on every street since your hand range is simply stronger, and ideally you should win a lot.

If opponent has a glaring postflop weakness such as folding 67% of the time on any street, or he seems to also fold to 1/3rd pot sized bets, these are the types of opponents you might flat call their bet with any two, or raise preflop to build up a larger pot to steal.

IF opponent calls, be aware of HOW much better you can evaluate his hand strength to have gotten too. Imagine the opponent that raises with 40% of hands, calls with 40% of his range (16% of all hands) and then folds 60% of all flops (calls with 40%) to have the top 16*.4=6.4% of hands, or the top 40% ON that flop, given a top 16% calling range. As such, against the super tight opponents on the flop, you may have to be much more cautious as you approach other streets unless you have information that opponent is even more tight on the next street. But it shouldn't bother you. Afterall, as long as you are making break even preflop, profitable steals on the flop, and being very selective when profitable steal opportunities are no longer available, you should have a GREAT strategy.

Take Notes:
It may be more useful to know specifically what KINDS of hands opponents call with. Do they overvalue or undervalue draws? High cards? Low Pairs? Some opponents will try to hit sets and big draws, others will try to hit pair and only call draws if thy have the pot odds. Some will call with air and attempt to bluff. You want to know these kind of specifics so you can get more technical and less theoretical about opponent's hand range.

The way I approach exploitative poker is that my style preflop is pretty straightforward and designed to get me to start with what is more likely than not to be the best hand and come in aggressively with it. This way I build up a pot with what is likely the best hand. Now, knowing the hand ranges and assessing the situation, I can then look to exploit the mathematics of the flop and err on the side of aggression in most spots whether that be a continuation bet or delayed continuation bet.

When I am OUT of position, I want to know what percentage opponent will bet the flop and what percentage they will bet the turn, and what percentage they fold to a check raise on flop and turn. If it is too high, I may look to call the flop and check raise the turn, or check raise the flop. Remember, I start with the best hand.


Stat Tracking:
Keep Track of the following stats of opponents. Depending on what you try to exploit and how many stats you can handle without clutter you may want to reduce this

Preflop:
1)Preflop raise %
2)3bet
3)fold to 3bet%
Optional:
4)4bet%
5)fold to 4bet%
6)Voluntarily put chips in pot VPIP
7)fold in small blind.
8)fold in big blind.

FLOP:
1) Continuation bet%
2) fold to continuation bet %
3) check raise %
4) fold to check raise %
optional:
5) 3bet %
6) fold to 3bet%
7) out of position lead bet (when not as continuation bet)
8) In position bet (when not as continuation bet).
TURN and RIVER same as flop if you like.

Alternative:An alternative set of options may be tracking raise from X positions such that you know opponents EARLY position stats, MID position stats and LATE position stats all separately. This way you can see how opponents adapt to and value late position. That can be incredibly important.
Additionally if you use a "short stacking" strategy you will not need to know as many stats because you will be moving all in on many flops and can get rid of much of the flop and turn stats and the 3bet will be all in so you don't need to know opponent's 3bet and 4bet stats really.

If I want to be really exploitative against an opponent, I will flat call with more hands than opponent plays and exploit his inability to fold after the flop, OR his inability to call very often. This will potentially provide me with a really loose image against good players which I will not be exploiting, and get me action with  much better hand.

If you have this data and opponent is not playing you or adapting to you, it can be extremely simple to exploit, while also knowing the baseline strategy to shift to every now and then to prevent counter exploitation. I tend not to try to make a killing and instead just grind away my opponents with slight exploitations. Playing half opponent's ranges allows me to play a lot more hands when facing a raise from looser opponents without being too crazy and still having a good edge. The ability to adapt to opponent and play sound against them is enough to enter in enough pots and play a relatively break even style without even needing the cards a lot of times. Add in the occasional showdown where you have the chance to hit draws, and the occasional equilibrium shoves with monster draws and sets on top of that while being able to more accurately assess your hand according to opponent and flop and you can crush the game at low stakes.

No comments:

Post a Comment