I’ve had a few players ask me for tips to improve their game, and thought I’d summarize here some of the things I think might help players ranked around 200 to 800 move up into the top 100.
Work on pre-flop ranges
I’ve been surprised by how many players don’t seem to take pre-flop strategy seriously. Decisions on every street feed into your over all win rate, and it gets harder as you play better players to make bad decisions pre-flop and then recover from that with superior post flop strategy.
- decide on what hands are worth opening from each different seat at the table, and play more hands from the button than from the cutoff, more hands from the cutoff than from the highjack, etc.
- decide what hands make good 3 bets from each position, facing a raise from each position, and in general, 3 bet more (especially when facing raises from players that raise more frequently)
- call less frequently; this goes hand in hand with the point above (3 bet more), but especially as you have aggressive players still to act behind you that might squeeze, and where you are neither the button nor last to act, there just aren’t that many hands that really perform best by just calling
- flat 3 bets even less; if someone raises in front of you, and then someone raises over that, flatting really exposes you to the risk of a 4 bet from the initial raiser, and with a 3 bet already out there, the value of speculative hands diminishes a lot (their implied odds are diminished in 3 bet pots as stacks behind aren’t as large relative to the chips already in the pot)
- consider the frequencies with which different players are making different kinds of plays, and record those statistics
Work on Balancing Ranges
The goal here is usually not to achieve balance, but more to understand what balance looks like so that you can tell where your opponents are not balanced, and how you can react to exploit that.
- understand where you should make bigger bets, and where you should make smaller bets; at a very high level, you want to make big bets where you have a more polarized range (a nut advantage), and where your opponent’s range has a reasonably high density of hands that will want to continue
- understand optimal bluff frequencies, and how bluff frequencies are a function of bet size. A pot sized bet where the bluffs have zero equity (usually only occurs on the river) should have 1 bluff for every 2 value bets, where a 1/3 pot sized bet should only have 1 bluff per every 4 value bets (just compare the size of the bet to the size of the pot plus the bet).
- understand how having equity in your bluffs allows for bluff ratios higher than this (while those bluffs still have equity)
- your checking range needs protection: you need to check some good hands at least some of the time, especially out of position
- if you have a hand that is mostly ahead of your opponents range, but mostly behind their calling range, a bet is not a value bet, it is a bluff. Really weak hands with no show down value have the most to gain from a bluff, but there are also still spots where a hand mostly ahead can benefit from a bluff of this sort (when your opponent’s range has significant equity).
- Check with some middle strength hands that are less vulnerable to cards to come
- Make “protection” bets with some other middle strength hands more vulnerable to being out drawn
- bucketing hands into 1 street of value, 2 streets of value, and 3 streets of value (and no streets of value) hands is useful in a lot of spots, but it is also important to remember that the flop, turn and river can change things. If you make a bet on one street, it can be useful to decide in advance how various cards on the next street modify the type of hand you have.
Apply Pressure, but also Patience
Continue to work on segmenting your range given the cards that are now visible:
- hands that should check/fold
- hands that should check/call
- hands that should check/raise
- hands that should bet/fold
- hands that should bet/call
- hands that should bet/raise
As you face decisions on every street, think of all of the hands you’d arrive at that spot with over millions of iterations, and how different hands should be divided into the categories above. You want to think about what hands belong in each category in the actual situation you find yourself in. You want to have a decent number of hands taking aggressive actions, but I find it helpful to try and focus on making the right play with the right hand, and finding the right hands to fold is important too.
Play Exploitatively
If you happen to be an advanced AI, you can likely just play at the perfect frequencies and do quite well without paying any attention to the mistakes your opponents are making, but I think the biggest value from beginning to understanding optimal frequencies comes from understanding where you opponents are deviating from them, and how you can respond profitably to that. Especially pay attention to the frequencies you see from the opponents you think you’re most likely to be able to get chips from, and pretend everyone else is playing perfectly. I personally like to try and adjust my frequencies enough to derive extra EV, but subtly enough so that it’s harder for opponents to realize where my ranges are now also liable to counter exploitation. As an example, against a fairly extreme call station, I like to continue to include some bluffs in my range, but I’m going to bluff at frequencies significantly below theoretically optimal levels.
