The Sunk-Cost Effect: Post-Flop Play Part IV

Nicolas Levi
Small blind + small stack = big mistakes.

Continuing our discussion of post-flop play, here are two "don't" situations - ones where the message is don't ask for trouble.

1) Avoid problem situations and problem hands.

Position and previous action are keys. If you want to make your life a living hell, limp into pots early with hands like KJ and A9. And don't tell me they were suited!

These are problematical hands. They "look good" and have the potential for making big hands.

The problem lies in the disjunction between the probability of the big hand and that of catching a minor piece of the flop. The former doesn't occur often and when it does, it typically won't win enough to cover the losses when you catch second best.

Here's an example we've all seen (or done!): Mid-position limps with A 9, gets raised 2.5x by the BB and (reluctantly) calls.

The flop is 9 7 2. BB bets half the pot. MP calls. Turn is 4. Another half-pot bet, call. River's a brick. Bet, crying call. BB shows JJ. Reload.

Let's dig into this situation a bit and see why it is creates such havoc.

The initial limp isn't awful. Sometimes we get away with it and see a cheap flop. The problem comes when we get raised pre-flop and call or when we hit a piece of the flop and end up calling several bets.

Why do so many players make these calls? Well, one reason is that this situation invites what behavioral economists call the "sunk-cost effect" - that is, you get pulled into continuing with a line of action because you've already "sunk" costs into it.

It's analogous to the notion of being "pot committed." However, here we really aren't pot committed - certainly not pre-flop and usually not on the flop.

But the tug to go with a hand that has outs after we've already "sunk" valuables into it has a strong emotional pull, partly because we tend to overestimate the potential positive outcomes.

That is, once the slide into the sunk-cost dilemma starts, people caught up in it overestimate the potential gains.

Jackson and Dixon at USIC have looked specifically at how this effect plays itself out in Hold 'Em. And, for the curious, take a look at Dixon's research exploring the psychological links between choice, self-control and gambling.

These problems are ubiquitous in economic settings and finance and haven't been satisfactorily solved so don't be too surprised when you see poker players falling into them.

There are, alas, other difficulties with these hands. Action junkies get pulled in because when they hit, they produce large "reinforcements" (flop two pair with A-9 and you can do a lot of damage to A-K).

As we've discussed numerous times, large rewards have a significant impact on shaping our emotions and our approach to the game. But elementary game theory tells you that the play has negative EV.

Worse, these hands suffer from information poverty. You typically do not know where you are in them. Your opponent's range of hands is large, as it often is when you've limped into a pot.

In addition, you're acting first in these situations and this is never good.

Save yourself a lot of heartache and cash and stay away from these hands. Even the very best players have trouble with them. If you don't see a flop, you don't have to worry about post-flop play.

2) The half-bet from the SB, call or fold rags.

This one has been hotly debated over the years. The standard argument for calling is that you're getting attractive odds, particularly if there are several limpers and a relatively passive player in the BB.

This isn't crazy but it needs to be filtered through some subtle screens.

First, appreciate that you don't really know your implied odds. If there are three callers and you're looking at T 2, it's bloody unlikely you're getting the 9-1 you need (against random hands) to justify the call (assuming the BB doesn't pop it).

Second, you'll be out of position all the way to the river. I don't know about you, but this rarely makes me comfortable.

Third, once you've made the call you're going to be caught up in the "sunk-cost" problem. And if you catch a piece of the flop it'll get even tougher to bail out of the hand.

BTW, I chose the T 2 example here for a reason. It is, of course, known as "Doyle's hand" or "the Brunson" since he won the WSOP Main Event with it twice and found himself psychologically committed to it (as noted, "reinforcement works").

If you saw a telecast of High Stakes Poker last year, it had a magic moment. Brunson picked up T-2, looked at it and dumped it in the muck while making the classic spitting sound people make to ward off evil.

Later he said that he can't begin to count the money he's lost playing that hand.

Calling the half-bet with junk is a long-term risky play. Here's a simple rule: Don't call with any hand that you wouldn't play for a full bet in early position.

Related articles:

Please fill the required fields correctly!

Error saving comment!

You need to wait 3 minutes before posting another comment.

David Peterson 2017-05-12 08:39:20

T = 10

Paul Malmquist 2011-07-21 20:07:36

What does T mean when speaking about a specific card? As in the flop was T 8 2...

Arty Smokes 2011-06-12 21:10:39

I'm probably too risk-averse, but I hate being in the small blind on full ring games so much that I usually fold everything but the top 20 starting hands, except where I'm only playing against the big blind. Being the lead-out person in every round after the flop puts you at such a disadvantage (in reading other players and calculating pot odds, for example) that it's almost never profitable to play anything but monsters. They call it "dead money" for a reason. 80% of the time, being first to act makes a you a dead man walking. You're practically giving away your money if you play T2o, or even KJo, in the small blind. The times you hit top two pair are so rare that I doubt a profit can be made in the long run unless you are a highly skilled post-flop player.

Jed 2009-12-01 20:11:46

Maybe I just play the SB better than most, but if there are limpers in the pot, I almost always call the half-bet with any marginal hand. I'll throw away junk like J-2o etc. but any suited cards or connectors I'll usually stick around. If I hit anything, since I act first, I'll throw out a small bet that sometimes takes down the pot. If I get called or raised, it's easy to get away from and I've only lost a couple dollars. I've won a lot of pots betting out with 8-6o on a Q-6-2 board and having everyone fold.

Arthur Reber 2009-09-16 02:14:00

OnoMakana makes a good point. But the key to understanding the sunk-cost effect is that it is psychological. It is the emotional element that causes the problem. If you can step away from this and view the situation in a "rational" manner then this analysis works.

Alas, as we've discovered in "real world" economic models, the "rational man" assumption is often missing. It's stunningly difficult for human beings to behave rationally.


OnoMakana 2009-09-15 03:38:00

Interesting and informative article. The use of "sunk costs" as an explanation was on the right road but I think it pulled up short. You noted the fault of many players, but did not suggest an alternative.

The lesson should be that "Sunk costs", or costs that have already been spent, invested or lost, should not have any affect on later decision making. The fact that they often do is a problem. You should be playing to win the money that is in the pot, not protecting the money you have invested in it. That money is not yours anymore, it is part of the pot, and if you think making an ill-advised call on the river will likely lose more money there is no reason to invest further, despite your previous bets/calls.

In the end, you should always be looking to win the hands that you are involved in. The amount of money you have put into the pot should not influence your decisions on later streets. A bad call is a bad call, even if half your stack is in the middle.

One note is that this advice is most applicable to deep stack cash games, where the idea of being "pot committed", IMO, does not exist. While still essential, in short stack tournament play, the idea of "pot committed" does in fact exist, but hopefully you're avoiding those situations and tough decisions by just shoving when you think you've got the best of it.

Hope that helps! Good article.

MelMeiko 2009-09-12 05:26:00

Highly informative. There's nothing worse than thinking you've got the nuts, just to be owned by your own impatience.

Tim 2009-09-09 17:47:00

I think that this may be true but what has to be given thought here is the effective stack sizes of the players. If you were around 200-500 bb deep, I think it would be extremely profitable to complete from the sb with many limpers with wide range in an attempt to hit 2prs and such, especially if playing against weak players. I agree however as the stack sizes change, the strategy changes as well. If you were 50 bb deep, you would rarely be doing this.

However what hands you play in ep also depends upon the players to your left. If there are good players to the left of you, I agree completely that folding is standard but if there are weaker players, I might 3 bet or flat depending on whether I wanted to balance my range, or maybe because I believed I could take by a c bet. However I understand the other argument being that what do you expect the op's 3bet calling range to be especially when you're 3 betting marginal hands. And I feel that this factor must be balanced by just game flow and your "feel" which is hard to describe other than how you feel the other player's feel about your actions (i.e. frustrated).

Rray 2009-09-09 07:30:00

Good article.

I'm a beginner and have learned to fold these hands in middle position, though note the temptation to play them 'just in case'.

What are their value in late position- say MP3, C/O and OTB? Would the added equity of getting your opponents to fold by playing position (raising preflop) affect you assessment of these hands' worth?

Best Poker Sites - Editor`s Pick


Sorry, this room is not available in your country.

Please try the best alternative which is available for your location:

Close and visit page