I'm becoming a bit of a stats monkey so I figured I'd write a little post on them for the newer players. The recent discussion in the chat thread and other places lead me to believe some people could benefit from this.
Here's my hud layout. This particular player is a bad regular who I have a ton of hands on.
Voluntarily put money into pot/Preflop raise/Total postflop aggression factor/3bet/fold to 3bet
Attempt to steal/Fold to flop bet/C-bet/Went to showdown/Fold BB to steal
Limp-call %/Squeeze/Aggression frequency
And of course the popup:
Last line in the hud is basically a mess of stats I'm experimenting with. LC and squeeze are staple stats for me now, AFq being something I have yet to really implement into my game. The rest of the stats are things I usually check at least once during every hand.
I'll gear this towards people who understand what the stats mean but don't necessarily know how to apply them. Please don't ask me what a stat means, or else I'll kill a kitten. You can ask SavagePenguin about that.
I guess the best way to do this is to run through some generic situations and explain what stats I look at and what thresholds change my decisions.
First thing's first. Sample size is everything. Some stats converge (meaning they become accurate to that player's actual style) much quicker than others. Some take forever, like check-raise river %. For pf decisions I think a few hundred hands can be enough for most situations. For intricate postflop decisions, usually 10k is enough. Getting that many hands on a player is something you'll have to figure out...datamining where it's legal is essential. Otherwise...figure it out. 1) Stealing: folded to me on the button:
This is obviously where we open up the most out of any other position at the table. Just how loose I go depends on how the blinds play. How tight they are, how passive they are, how sticky they are with their hands postflop, etc.
First stat I look at is 'fold BB to steal'. A good rule is if both the BB and SB add up to 150 or more, we can usually steal any two cards. VPIP/PFR is helpful, but really doesn't matter if both their fBBtS stats are like %85. In that case, we steal every single time it's folded to us on the button.
I think stealing every button until the stat gets down to about %65 or %70 is good. At that point, I still maintain a high steal %, and only adjust when they start to call/3bet me more.
Stats like 3bet%, fold to flop/turn/riv cbet%s and wtsd play a big part of how wide I'm stealing too. Someone could have a highish %74 fold BB to steal, but have a really high %12 3bet and low fold to cbet at like %55, and because of that I'll steal considerably less than against someone who calls a steal less
, but gives up postflop a lot more. In that case you'd actually be making more money on the play by way of your opponent putting in more money before giving up.
Remember to watch out for shortstacks and drop your open size to 3x or 2.5x if it's not already to make sure you don't commit yourself with marginal hands. They generally don't adjust too well to the size difference. 2) Getting 3bet when we steal:
Your staple play here should be to fold. There are a ton of strategy posts about how to play really tricky in 3bet pots, but don't stress enough the fact that folding should be your priority. Against most players, their 3bet ranges will be tight enough that they don't really contain many hands that will fold to action. That in mind, there are definitely ways of exploiting people who 3bet too much.
First thing we look at is actually his username. I'm kidding, ffs. First thing we look at is obviously his 3bet%. To be more specific, we look in the popup at the "vs steal > 3bet". 3bet% changes dramatically for some players vs a stealer in comparison to their overall %.
%5ish and below is fairly tight and consists mostly of hands that are raising for value. %5-8 is where most of the good regulars are sitting at (and where you should be too). There are a considerable number of light 3bets in the %7-8 range as well as the value hands obviously. %9 and over is generally where people start getting out of line. Often times people take that first step and 3bet extremely light, but don't deal with 4bets well, or don't play their hands well postflop.
So if their 3bet% is somewhere like %5 and under, we almost always play our hand strictly for it's face value. Usually people who 3bet that low aren't giving their hands up to aggression, so there's not much for us to do other than get our money in when we're good, and fold when we're not.
When they creep into that %6-8 range, they often also become exploitable in some way since there are now hands in their range that can
fold to some kind of play. Anything wider than %10 or so and we usually find some really exploitable players. The basic ways to combat people like this are 4bet bluffing, flatting and raising the flop cbet, flatting and betting the flop when checked to, and flatting pf and
flatting postflop as a float.
4betting: You're either 4betting to 2.5x the 3bet amount (so as not to commit yourself but still to get fold equity), or you're shoving. It's generally best to 2.5x it. Basically the only stat we're looking at here is 'fold to 4bet %' in the popup.
In a 2/4 game, if we open for $12, SB folds and the BB re-pops to $42, our 4bet bluff will be to about $100. $100/($100+$60) = 0.625 meaning that our bluff has to work about %63 of the time to be immediately
profitable. Most regulars don't have fold to 4bet stats that high (usually %50 +/-%10), but metagame plays a huge factor with 4betting. When you 4bet lots, people pin you as a spewy maniac, and they stack off light in the future.
I like to 4bet a little more often earlier in a session. When there's no history, people give you more respect. Flip it around; if you've been active and have maybe 4bet before on that same table, people will probably give you less respect.
On to flatting and raising the flop cbet. The main stat you should look at is under the popup: Flop As PFR > Fold CB to raise > 3-Bet Pot. %60+ and I happily raise the flop cbet as a bluff, anything lower and we should generally play fit or fold. Good flops to raise are basically those which are unlikely to have hit our opponent's range. Stuff like 567 monotone, flops that 'supposedly' hit your opponents range but don't actually hit that much of it like Qxx, Kxx, etc., rag flops.
I think generally when you have hands that are drawing dead, you should make small 2x-3x raises on the flop that don't commit you should villain shove. When you have something like a flush draw, OESD, overcards + gutshot, that's when you should shove. Of course all this needs to be balanced with real hands too, assuming your opponent is observant. If you make these plays with any consistency, you should often flat AA to a 3bet and shove a rag flop as well. Against those who don't pay attention, there's no need to balance.
Flatting pf and floating the flop (with the intention of betting the turn, or checking back the turn and betting the river when checked to as an even more complicated line) is a lot trickier. Similar flops to the ones I mentioned above should be floated: stuff that misses your opponents range and hits yours. I generally make this play if my opponent's fold to flop raise isn't too high, if his turn cbet is somewhat low, as is his turn check-raise %. Those last two stats in the popup don't apply that well in 3bet pots since they refer to overall stats rather than 3-bet pot-specific ones, but they still help.
The last option is betting the flop when checked to. Unless you're up against a really tricky player, when your opponent checks the flop as the PF 3bettor, he's almost always giving up. Bet it almost always. Just to make sure, check that his ch-r flop % isn't too high. %15+ and we should probably just shut down. Anything below %10 and we can fire away!
This is taking a hell of a lot longer than I anticipated. 3bet pots are extremely complicated and I promise I'll touch on simpler stuff next time. I'll wrap it up now and will come back, prob tomorrow, to post more examples.
One last point though. A ton of the nosebleed players play without a hud. The reason that I've heard from most of them is that 'stats are situational' (plus they're just monsters who don't need huds), meaning that just because someone cbets %75 of the time, doesn't mean he'll do so on a 567sss board, or a Kxx board. He'll cbet those kinds of boards with completely different frequencies. This also applies to other stats as well. Of course this is true, and the key thing about stats is that you have to understand how to apply them properly. Hopefully I'm explaining my points well and that you guys are getting something from this. Like understand that someone's 3bet% is going to be significantly tighter vs an utg raiser than it is against a button raiser.
Anyways, I'll be back tomorrow with more.