Exploiting Leaks

I was going to put this on my blog, because... Well, because I write enough stuff over there. But this does a lot more good in the forums, so here you go:

So you practise game selection. And you come across a table with a guy who plays a third of all hands, but clearly isn't a strong loose-aggressive player, but rather just someone who plays too many hands. Score! You found a fish! Sit down, play, profit. Easy money. Right? Except that doesn't quite suffice. It's not enough that you have players at your table who have leaks - because they don't have leaks if you're not exploiting them. With the exception of stuff like calling river bets with a hand that can never win or not betting the river in position with the nuts, there are very few plays that are inherently wrong; that are bad in a vacuum.

What I'm getting at is that you should be thinking about what mistakes your opponents make and how you're going to take advantage of them. Just saying "they make mistakes" is to only get halfway there. In fact, it's probably even less than halfway. Because unless you exploit them, they may not be mistakes at all.

Consider someone who 3-bets on average 20% of hands against you. Is he making a mistake?

Or someone who floats almost every flop you c-bet.

Or someone who flats a really wide range on the button preflop.

It's not enough to look at a player's VPIP and think that you've found a good table. You need to actually take advantage of the mistakes that players is making to make it worth it. I've played extensively with some 35/15 type players that are actually break-even at 200NL. They could probably improve on their preflop hand selection, but whatever they're doing postflop clearly outweighs any disadvantage they start with.

And let's not forget our friends the regs. If your table selection tool is to look at VPIPs, then you might wrongly deduce that a table full of regs is one where you don't have much of an exploitative edge. Think again. Regs don't have the big flashy "leak" of having a high VPIP, but they can (and do) certainly have other habits that you can exploit. Think about your opponent, about where he tends to overdo something that you can exploit, and then try to get into those situations with him with a plan for how to abuse it. Maybe he check-raises the flop an awful lot from the blinds. That can be a pretty big leak if you know how to handle it - but most opponents don't, so the eager flop raisers get away with it.

So, just to drive the point home one last time: Leaks are there for the picking. But if you ain't picking, they ain't leaks. Playing "bad players" is not an immediate recipe for success - you have to either find bad players that happen to have a style that doesn't work against how you play, or you have to adjust how you play to the mistakes they make. I know which of the two is the more common approach, and I also know which one is the more profitable. The two are not the same.

Follow-up question: How DO you exploit a loose player at your table? My guess is that most of you play your default game and let the money roll in on their own. But that's leaving a lot of money on the table. Or rather, in the pockets of the other players.

Raise & isolate them w position.


... then back the truck up!

gd. post (as always). Looking forward to the comments generated by it (not mine obv.)

Probably a good start. But two questions immediately follow:

1. With how wide a range? and
2. What if you don't have position on them - do you not consider it a good table because the fish isn't on your right?

I'll take a shot for the sake of discussion plus, I might learn something. Should also note that I play FR limit but I think this a concept that obv. applies to nl and limit.

1) We want to be opening up wider, but I think we want to be doing it with hands that have a greater chance of winning at SD unimproved.

2) Confused myself here. At first was going to say that I wouldn't consider it a bad table, but a strategy adjustment is needed. Then, I remembered hearing a quote that Ivey said he couldn't beat his grandma OOP. So now I'm really not sure.

How about Regs?
My play would be 3 bet light when in position until they play back at you.Then become tighter, having created the image you are 3 betting light, wait for a strong hand out of position against the reg then 3bet hoping he will read you as still 3 betting light.

another stimulating post by F Paulsson.

Most regs will have HoldemManager or PokerTracker HUDs running. They'll judge you not just by how you play against them but how you play on average, so while metagame situations often occur (if I've 3bet someone four times in a row that they've opened, they're almost certainly going to start getting pissed off) your "image" is mostly comprised of what you've done in the past, not what you've done this session.

But yes, exploiting how regs react to 3bets is important. Having good ranges for value and bluffs versus regs is imperative. Personally, I use something like this in position:

Vs reg who folds often to 3bets, opens in early position, doesn't 4bet much
For value: AA and KK.
Bluffs: All suited cards 9 and lower (e.g. 82s, 96s) that I don't want to flat for value (e.g. 98s), and offsuit connectors and one-gappers below 9 (e.g. 97o, 65o) and above 4.
Vs reg who folds often to 3bets, opens in early position, 4bets frequently
For value: QQ+,AK
Bluffs: Same as above.

Vs reg who calls 3bets often, doesn't 4bet much
For value: AJs+,AQo+,KQ,TT (but will fold to a 4bet), JJ+,AK (will shove versus 4bet)
Bluffs: Same as above, except I skip the offsuit hands.

Vs reg who calls 3bets often and 4bets frequently
For value: Same as immediately above, but will now shove AQs and TT versus a 4bet (still folding AQo and AJs)
Bluffs: Skipping the smaller suited hands (e.g. 52s), but rest of bluffrange stays in.

... etc.

Note that I only use four classifications here and that guarantees that my ranges will be far from optimal versus many opponents. I wish I had the mental capacity to tailor my ranges to each and every possible value for their fold-to-3bet%, their 4bet-range%, etc. but I don't. So I use some vague estimates. I consider a "high" fold-to-3bet% to 60%+. I consider a "high" 4-bet-range% to be 2.5%+. I will, in other words, play very differently versus someone who folds to 3bets 59% of the time compared to someone who folds to 3bets 60% of the time, but that's the price I pay for not using finer classifications.

By the way, don't copy these ranges and incorporate them immediately into your game; there are a LOT of caveats that I didn't write down here. I just wanted to give you an idea of what my ranges look like and how I approach 3betting, not to give a starting hands chart to use.

Does the regs Aggression Factor % influence you much before making the decisions above

No. I don't know if it should or not, but I don't look at it.

I hate having passive fish on my left, I end up barreling too much and head straight to tiltsville when I have to b/f river w a top pair type hand a couple of times. So yeah, I tend to not consider it a good table simply cuz I'm bad at it, and this is probably a leak for me - I spend a lot of time finding tables w nits/TAGs on my left and LAGs/fish on my right (heaven!) - and why the heck should I pass up an opportunity to play against a fish???

Looking forward to thoughts on this - I tend to barrel fish until they show strength, but passive fish frequently won't show any strength until the river, and I end up feeling like I spewed off three streets oop. I'd guess a couple of thoughts would be "don't barrel too much oop and keep your bet sizing on the smaller side"? But then I feel like I don't get sufficient value from top pair hands, which are frequently the nuts against passive fish who call down (much, much) worse.

It really annoys when you spew to a fish and they immediately leave the table with your money.