Bluffing the Turn

Last night I played an hour and a half long session in which I gained just over a buyin playing full ring. As I was checking out of my tables, I ended up heads up with a donkey at one, so I decided to stick around and focus entirely on that table. His only moves were to bet and raise, so while frusterating to play he was easily exploitable. Eventually someone sat down at the table (at which point we were about even), and he complained that our HU match was ruined. I informed him that Stars has $50NL 1-on-1 tables which we could move to if he desired. He said he wanted to, so I was on my way to playing my first heads up cash session ever.

It took me 28 hands at the HU table to take his buyin. I ran pretty well, but also had a huge edge. Here are my lifetime HU stats:

166ptBB/100 for the win!

Ok, on to what I want to talk about today as per the title of the post. At microstakes, it is possible to be a profitable player without ever bluffing the turn or the river. Bluffing the flop (which 90%+ of the time is in the form of continuation bet) is however likely a prerequisite for success. That said, being able to pick good spots to bluff on the turn and the river could easily increase one's winrate substantially, and is a necessary skill to develop as one moves to higher stakes.

So, when do we bluff? If we're thinking about bluffing one of the later streets, we need to double check three things:
1) It is feasible that we actually have the hand our bluff represents based on the action so far.
2) Our opponent is not a calling station.
3) Better hands can realistically fold.

If our opponent is a calling station, we're probably getting called down regardless of how great our bluff was, and even poor players can sometimes figure out when a line doesn't make sense and by default often call to see what is going on. Further, if only worse hands fold to our bluff, we would have been better off checking the hand down or perhaps picking up a small value bet on the river, rather than blowing the weak hands we beat out of the pot.

Lets look at a couple examples:

Villain in hand one is a 10/6/2/1000 nit. As far as I know he's straightforward postflop.
Poker Stars, $0.25/$0.50 NL Hold'em Cash Game, 9 Players
MP1: $97.20 (194.4 bb)
MP2: $41 (82 bb)
Hero (MP3): $67.90 (135.8 bb)
CO: $59.10 (118.2 bb)
BTN: $50 (100 bb)
SB: $47.50 (95 bb)
BB: $68.70 (137.4 bb)
UTG: $50 (100 bb)
UTG+1: $49.60 (99.2 bb)

Pre-Flop: Hero is MP3 with
2 folds, MP1 calls $0.50, MP2 folds, Hero raises to $2.50, 4 folds, MP1 calls $2

Flop: ($5.75) (2 players)
MP1 checks, Hero bets $3.50, MP1 calls $3.50

Turn: ($12.75) (2 players)
MP1 checks, Hero bets $8, MP1 folds

Results: $12.75 pot ($0.60 rake)
Hero mucked (a pair of Fours) and won $12.15 ($6.15 net)

In this first hand, we continuation bet our small pair into a two tone board and get called by our uncreative, nitty villain. At this point, our range is anything we raise with (overcards like AK/AQ, large pairs, small pairs, suited connectors, etc.) Since our nit limped preflop in MP, it is very likely he is playing a small/medium pair, something like 22-TT. It's also possible he's on suited connectors or a hand like AJs, but fairly unlikely.

So, given that information and that our villain called our c-bet, what are the villain's most likely holdings?

1) A middle sized pair, second or third pair based on the board. TT/99 most likely, with 33-66 a possibility as well.
2) A set. This is somewhat unlikely however, as many villains would raise this two tone flop with a set, as they really don't want to see another spade fall.
3) A flush draw. Very unlikely as our villain is fairly nitty preflop, but we can't entirely rule out a hand like QJs or AJs.

Based on the likelyhood of these hands, we are behind almost the entirety of 1), nearly dead to 2), and probably even money against 3). On the turn, we catch the Kh. This doesn't help our hand, but it is a good card for our range - as our villain almost certainly doesn't have a king, and we very possibly c-bet our missed AK or KQ. Of course, it's also possible we had a big pair like AA/KK/QQ all along. Thus, this is a good time to fire a second barrel, because we can now fold out nearly (if not entirely) all the hands that beat us in 1). If our opponent has a set, he will let us know here and we obviously will fold.

This bluff worked because it was consistant with the betting so far (we very possibly had AK/KQ, or a big pair the whole time), and our villain was a standard nitty reg rather than a calling station. It was a +EV move (and almost certainly the highest EV choice based on our villain's range) because there were many better hands that were likely to fold.

Let's take a look at another hand which is a bit more complex.

Poker Stars, $0.25/$0.50 NL Hold'em Cash Game, 9 Players
MP3: $53.55 (107.1 bb)
CO: $61.30 (122.6 bb)
Hero (BTN): $51.85 (103.7 bb)
SB: $8 (16 bb)
BB: $43 (86 bb)
UTG: $52.90 (105.8 bb)
UTG+1: $50.90 (101.8 bb)
MP1: $48.95 (97.9 bb)
MP2: $11.50 (23 bb)

UTG is a loose fish, don't have stats offhand. MP3 is an unknown.
Pre-Flop: Hero is BTN with
UTG calls $0.50, 3 folds, MP3 calls $0.50, CO folds, Hero raises to $2.50, 2 folds, UTG calls $2, MP3 calls $2

Flop: ($8.25) (3 players)
UTG bets $0.50, MP3 calls $0.50, Hero raises to $5, UTG folds, MP3 calls $4.50

Turn: ($18.75) (2 players)
MP3 checks, Hero bets $11, MP3 folds<

Results: $18.75 pot ($0.90 rake)
Hero mucked (high card King) and won $17.85 ($10.35 net)

In this hand, I really just wanted to take down the blinds + 2 limps with my button raise, but got caught. However, the min-bet + min-call on the monotone flop tells me that neither of these players have a made hand of any sort that is worth protecting, because otherwise they'd want to give people poor odds to catch up to the flush should a 4th heart come. A made flush should also be worried about a 4th heart because if said flush isn't the nuts, it is now beaten by bigger flushes (which are likely) and if it is the nuts, the 4th heart will likely stop any further betting.

So I decide to raise the flop, which is a pretty standard play in this situation. Both players have shown weakness, so consider this a "continuation bet" that is actually a raise. Technically I have a gutshot straight draw at this point, but it's not really much of a semi-bluff. The fish folds, and the unknown calls. The turn blanks (actually I pick up a double-gutter draw at this point, but I'm not terribly excited if I get a call on the turn and then make my straight, since I'm against a made nut flush a lot of the time at that point). So, we have a draw we're not excited about, but what is our villain likely to hold?

1) A flush draw. The way the betting has gone, this is by far most likely. Very likely the Ah or Kh. Could very possibly have a hand like Ah8x, AhJx, or KhJx for the draw + one pair. Many of these hands should fold to a significant bet on the turn, unless our villain is a huge station (we don't know at this point). If he has a hand like KhQx, we had TPGK and the 2nd nut flush draw, and probably isn't going away.
2) A made flush. Possible, but misplayed (terribly so if it's not the nut flush) if this is the case.
3) A set, 2 pair, TPTK type hand. Possible, but so brutally misplayed at this point that I don't really want to think about it. If our villain is poor enough to play such a hand this way, they're also probably bad enough to lay down a 2 pair type hand on the turn giving us credit for a made flush.

Based on these holdings, what does our second barrel on the turn accomplish? We fold out many of the hands in 1) which have better equity in the pot that we do. Our only real hope at taking down the pot at this point is by bluffing, though there's an outside chance we spike our straight on the river and it holds up, though most likely we're then against an ace high busted draw, so we get no further value. Overall, this bluff probably is not as a good a play as the example above, because we do not know if our villain is or is not a calling station. We do know that better hands can fold, however. Finally, due to our flop line, our turn bet is completely consistant with our representation of hands in {AA, KK, QQ, JJ, AQ}.

The reason why turn bluffs have the potential to be so effective is due to the concept of leverage, originally articulated by Howard Lederer. It's a very simple concept: people are reluctant to call turn bets because there is always the threat of a big bet on the river that they'll then have to also call to see showdown. That isn't to say we should never bluff the river, and I do so fairly often now (at least compared to myself a few months ago) when it is consistant with the line I've taken in the hand. However, always keep in mind that bluffing the turn is much scarier than the river because of this leverage.