In the new thread on the Baluga Theorem, there's an offhand comment from DK Barrel saying:

So I wanted to try to do some calculations to get a feel for what would happen if we really did raise preflop and then triple barrel every single hand.

I did some calculations. In my calculations I assume the following:

1. We raise to 5bb preflop. We neglect the blinds (they get raked away).
2. We get 3 callers preflop.
3. We fire a flop bet. If one person calls, we fire a turn bet. If our opponent calls that, we fire a river bet. I assume we are deep enough not to get all in with these bets, and I also assume that we don't get more than one caller on the flop.
4. Each barrel that we fire postflop is pot-sized.
5. I assume we will take the pot down on the flop 1/3 of the time. I assume that when we get a flop call, that caller will fold 70% of the time to a turn barrel. I then assume that if that bet is called, the caller will fold an additional 30% of the time to a river barrel (since calling the turn is usually stronger than normal).
6. I assume we are bluffing, i.e. that we never win at showdown if all 3 barrels get called.

Under these assumptions, the EV of triple barrel bluffing is a little bit worse than -10 big blinds. To be quite honest, I'm a bit surprised it's this close to zero.

What I can never be sure of, of course, is whether my assumptions are reasonable enough to give me a decent idea of anything. But, I wrote some code to carry out these computations. The code can easily be changed to accommodate changing assumptions #1, 2, 4, 5, or 6. So if you want to help me with my thought experiment, tell me what you think is a reasonable way to change the percentages and I can run the numbers. Hopefully we can get a lot of different perspectives on what we think are reasonable calling percentages and I can run the EV under all the different sets of assumptions.

So, I think this is an interesting idea. A few tweaks that we can make here.
I think ~50% of the time that we get 3 caller pre flop, and we pot it on the flop we are going to get 2 callers. 50% we get 1 caller. (Sure whatever, sometimes we get 3 callers, who cares.) This once again blots the pots, and lessen the chance that we win on successive street.

Also, the bet on the turn can be smaller than a pot sized bet, and the same thing will be accomplished. 90% of the pot, even 80% of the pot will certainly accomplish the same folding tendencies. On the turn, if we .80PSB instead of 1PSB, this will save us 12BB the 30% of the time that we get called. On the river is we are making the same .8PSB, (.8*156) so 124BB, we are saving another 24BB over the 180BB bet we would have had to make.


Also, pot pot pot puts almost all players all in.
20BB on the flop (into 20) get a call. 60BB (into 60) on the turn, get a call. River, 180BB into 180BB?... Your average player is not sitting 265BB deep in almost any game. Average player at 1/2 sitting with $530? 2/5, average player might be sitting with $1325, but only if it's an action game. Or uncapped.

I think it's pretty bad to triple barrel all the time, or even a lot at all.

Just set your entire bankroll on fire. It will last longer than this strategy.

From a game theory standpoint, this is pretty bad, since any decent player will adjust by simply calling down lighter.

LLSNL is already filled with calling stations as it is, so the best adjustment to live play is actually to bluff as infrequently as possible, and value town always.


But for your thought experiment, triple barreling every hand we involve ourselves in is obviously going to result in -ev. There are situations where triple barreling can be +ev, but its not something I use very frequently.

It is a good thought experiment, but the unstated assumption is that no one at the table will adjust. The problem is that even at 1/2, there will be players that will adjust and call down lighter. Even the fish can figure out you are FOS, although the most timid will simply wait it out until they have a monster.

These players burn up fairly quickly. Online, they usually go until busted and disappear. Live, they are normally waiting for a bigger table to open up and will go as soon as a seat opens. For me, the more interesting question is how to handle such a player, not how to be one.

I believe the best strategy against such play can be derived from Texas Bonus Poker. Basically you're looking to show down with TP or even SP.

I think you should try it and report back.

I think triple barreling all the time is a bit suicidal because our opponents will have top pair or better when they get to the river that they aren't willing to fold but I think double barreling can be extremely profitable since live players like to peel one on the flop incredibly light.

What I was really hoping would happen was that there would be some kind of discussion of, if you really think my calculations are off, how they are off, and what would be a reasonable way to fix them. (Also, my calculations are already showing that this is -EV. It's not like I'm planning on going out and doing this. I'm just a little surprised that our EV can be that close to zero when we're bluffing 100% of the time.)

At some point, also, we'll need to account for the fact that sometimes we win at showdown! I know how to do this, but I want to get a reasonable estimate from the rest of the forum of how often we think people are going to bluff catch us on each street first.

I know you ran this effectively as an algorithm, but here are some qualitative points:

1. You will win small pots and lose huge pots. The flop barrel is the most effective one (we all know it's impossible to flop a hand) but it's in the smallest pot; the river barrel is the least effective one (villain stuck it out through the flop and turn, and no one folds) but it's in the largest pot.

2. Villains will adjust. It's part of the game. I appreciate the idea of the thought experiment in a vacuum. But poker isn't played in a vacuum. Fool me once, shame on... see, you can't fool me again.

3. Your risk of ruin makes this impossible in the real world. Again, a limitation of the thought experiment. You could blow through your entire bankroll doing this in the real world... in one session. That's not *why* it's unprofitable (but psst, it's unprofitable) - it's to say that it will be nearly impossible to weather the swings to take this from thought experiment to the real world. Risk of ruin with under 50 buy-ins is going to approach 100%. Ignoring EV, as a strategy, it's just not feasible... not possible.

4. Your assumptions are way off. You're saying that when LLSNL villains call the flop, they fold the turn 70% of the time? No way. I mean, **** man. 70% is huge. Do this - take your code and disable the third barrel. Look at your analysis for just 2 barrels. You're going to find your code with these assumptions tells you that barreling flop and the turn is massively profitable, but that's only because your assumptions are wrong, particularly that 70%. In addition, re: the third barrel, the river folding % is too high. No one has ever called the flop and the turn and folded a river. Ever.

Grunch i've actually done this live and at low stakes online. All I have to say is that this strategy gets expensive very quickly. Now for 4 or 5 handed play ithe discussion may be more interesting.

This point is actually what inspired the thread--because without running numbers, I thought this too.

If you go look at DK's post, he said "I ought to triple barrel every hand" after a comment where he said he can't remember the last time he got 3 streets of value with TPTK. The implication is:

-If you can't get 3 streets of value with TPTK, it is either because your opponents are folding hands you beat to 3 barrels, or because you aren't 3-barreling for value enough.
-If the reason is because your opponents are folding hands you beat to 3 barrels, then you should be 3 barreling a ton since your opponents will not often have TPTK beat and will frequently be folding.
-But if our opponents are calling down enough to make 3-barrel bluffing unprofitable, then doesn't that mean we should be looking for 3 streets of value with good top pair hands way more often than people on the forum usually advocate?

There is a contradiction in logic here somewhere and I am trying to find it. So all the people who are qualitatively telling me that 3-barrel bluffing doesn't work should give me folding percentages that they think are reasonable and let's figure it out.

Your turn fold is way too high.
Also if you're triple barrelling every hand then peoples river fold will be 0%

-10bbs per hand is massive.

That's just results-oriented.

I already said the turn 70% is ridiculous. Also, I'm not entirely sure what you're getting at... I think you answered your own question very well.

You established that the EV of triple barrel bluffing is -10BB.

You then said you're surprised it's so close to 0.

But dude! -10BB is ****ing insane.

That means you lose an entire buy-in every 10 hands.

That means you lose an entire bankroll in 12 hours.

I think you very soundly determined this is a majorly losing strategy.

OK, I'm going to stop posting in this thread for tonight after this one, but I see what the problem is--misleading thread title.

My thought experiment is not about triple barreling every single hand we get dealt cards. It is about triple barreling every hand we choose to play.

Meaning, I'm not thinking about what would happen if we just blindly raise and triple barrel every hand. I'm thinking about what happens if we either fold preflop OR raise and triple barrel. Then we have some control over how many times we try this play, so it's harder for people to adjust; and we also can narrow our range so that the percentage of the time we expect to win at showdown is not super low.

That's the main reason I'm asking for percentages--once we can agree on what percentages are reasonable (because I don't know if mine are, and as you're saying, maybe they're not), then I can go back and say, "given these percentages, how often do I have to win at showdown for triple barreling to be a +EV proposition overall?" When I say I'm surprised that the number I got was close to zero, I mean that I expect that if we toss in just SOME showdown wins, it would be positive. But the conventional wisdom is that when we are triple barreling, we need to win at showdown a lot. I want to see if I can back this up quantitatively, or if our assumptions would lead to different conclusions than we all think.

What stakes are you looking for fold percentages for? And I'm assuming heads up pots?

This changes nothing.

And for once I'm with the crowd.

For your question of better percentages: I think flop fold is 25%, turn fold 30%, river fold 5% would be more accurate.

Since you could probably never come up with even reasonable estimates for the kinds of things that you'd need to know to carry out such an analysis (I'm not sure for instance how you came up with pretty much any of your assumptions, and don't see why anyone would believe they're an approximation to reality), the right way to study it is most likely by simulation.

Yeah most people seemed to miss the point Vernon.

TBH, I think your point has some validity.

You have to have a good image to pull this off. I wouldn't try it, but I can see it working if you are only doing it once or twice in four hours.

For the record, I have effectively triple barreled when given the proper circumstances.

I'm not just going into a hand and saying 'I am tripling here.' I'm thinking 'If I get the right opponent(s) and the right board texture I can make this work.'

I remember one hand where I had been playing pretty tight and elected to open KT in MP. Flop was 789. I bet. Guy called. Turn was a 4. I bet. Guy called. River was a 2. I bet. Guy folded 9T. Now, he didn't have a very strong hand and he missed his straight draw; but I knew this player was capable of making a fold on the river to my aggression and perceived image of being tight.

I don't want to make this results-oriented, but it can work under the right circumstances and I wouldn't be afraid to do it ... Just not every hand and not even every time I raise.

Answer of the year lol

Yeah those percentages are too tight. That means they're only getting to showdown with 14% of their range when faced with 3 barrels which seems unrealistic.

Really this depends on your starting ranges...if you only open jj+ then tripple barrel 100% its probably profitable.....but if your opening anything approaching a reasonable range your going to get called down so light.

Is this thought experiment including our ability to read the writing on the wall? Are we ever folding to a raise? Villains are not going to know we are triple-barreling, so they're not (always) going to flat with their made hands. What happens if we get raised, esp. on the turn? We just call and then barrel the next street? Or do we simply keep raising until we see felt?

If the V raises on the turn, it cost about the same as the times that we get flatted and barrel the river anyway.
And we still lose.
And it happens about as much.

So yes, sometimes that will happen.
But for the sake of the discussion, I think we are ignoring that possibility.

You don't think the folding %'s will decrease after doing this 5 times in a row? If you're not playing against a head of lettuce, they'll realize that you're the biggest maniac ever and will start calling you down with anything or re-raising you. Even if your assumptions were true for the first 1 or 2 hands, they'd quickly approach 10% or 5% after they see, "Gee, he raises and 3-barrels every single hand, 10 hands in a row now!"

Edit -- or did you just mean every time that you raised preflop? Not raising preflop every hand? In that case, it's still true that people would adjust (as others have pointed out) and the percentages would be decreasing.

I'm pretty sure the experiment ends if we get raised at any point in the hand...if I understand it correctly.