A4s vs 3bet deep bluff

Hey GingerKid,

2:1 bluff:value ratio is partially constructed based on what our river bet size would be if we fire. Usually when playing deep we can get away with more bluffs in our range than just 2:1 seeing as we may be betting river large on average and hence ok to be weighted towards bluffs.

As for your flop range it looks ok, although depending on preflop defending range I might look for other hands to bluff-raise as opposed to stuff like A4-A6s with the bdfd, seems like these may be slightly better in our flatting range since can be ahead of villain's flop range some percentage.

Assuming opponent just calls flop/turn and folds river way too much we definitely want to exploit that as much as possible and pay less attention to how the board runs out.

Originally posted by w34z3l
2:1 bluff:value ratio is partially constructed based on what our river bet size would be if we fire. Usually when playing deep we can get away with more bluffs in our range than just 2:1 seeing as we may be betting river large on average and hence ok to be weighted towards bluffs.

It is actually purely based on pot odds we give to villian with our raise on flop, with bet turn and river. Here, I should have even less bluffs than 2:1, because turn size is smaller than usual, and river bet size is standard. So it doesnt have to do much with playing deep, but with pot odds we give to him (usually if we play deep and play for stacks we give worse pot odds to villian so we can include more bluffs, but this is 3bet pot, which is why the pot odds are not worse than usual).

pot_odds_flop: 0.25
pot_odds_turn: 0.21
pot_odds_river: 0.31

Now we calculate the number of bluffs on river, value_combos = 13:

bluffs_river = 13 / (1-pot_odds_river) - 13 = 6
bluffs_turn = 19 / (1-pot_odds_turn) - 19 = 5
bluffs_flop = 24 / (1-pot_odds_flop) - 24 = 8

So in total we can have 6+5+8 = 19 combos of bluff on flop, which is as I said less than 2:1 bluff value ratio.

If we want to calculate directly number of bluffs on flop, we can put all together:

bluffs_flop = [13 / (1-pot_odds_river) / (1-pot_odds_turn) / ((1-pot_odds_flop)] - 13

bluffs_flop = 19

I did come to this equitons myself (it is straight forward), didnt read anywhere how other people calculate it. Is it the same method which you use?

Not 100% following your maths atm, I will come back a little later and look again.

I calculate bluff:value in a theory vid here http://www.pokerstrategy.com/video/32265/ if it interests you.

You might be using the same method, just not completely sure yet

I looked your video. Very nice explanation, the equitions from above are the same.
Just to give you feedback, I think you explain by talking in a very good way, but your slides should consist moslty images, and if it is hard to find suitable images, then at least less text per slides and larger fonts, using different font colors to mark important things etc which would all make easier to follow your presentation.

In your video you were saying that vs UTG and MP we should 3bet SC rather than e.g. KQ, AQ, AJ. I usually 3bet only offsuited broadways that are too weak to call.
The reason is:

1) if utg open raises 13%, he defends either by 4bet or call lets say 6.5% range (88 combos). It is something like JJ+, AJ+, KQ.
If you 3bet e.g. KQ, you block many combos from his range that defend vs 3bet:
QQ, KK, AK, AQ, KQ which is 21 combos. This is 25% of his defending range, which means that he will defend in reality less than he thinks, and this makes your blocker hands already preflop +EV even though he has balanced defense range.

2) You bluff SC like 9hTh, you block from his range 8h9h, 9hTh, ThJh, A*Th, KhTh, QhTh
99, TT, which is 15 combos that he folds usually vs 3bet, meaning that you make your preflop EV slightly less, but I guess not a big issue.

3) Having SC postflop is better for the reasons you said, but also having blockers postflop also makes the chance that villian hits top pair less frequent, if he flats AQs, KQs and similar. The chance of hitting the same pair is low, there are just two outs, so 1 in 8 time. When we hit it, we anyway dont need to get married to it. So for those reasons, when we have blocker hand it should increase our chance to take the flop down with cbet because villian will hit less often.

In your video you showed that we can construct ranges where our bluffs have 0% equity, so you can also barrel 3 streets with e.g. AQ even if it doest have equity but you will on each street block AA, QQ which also increases EV of bluff whereas bluffing 3 streets with 78s doesnt have blocker effect.

So what you achive with 3betting SC is that postflop they have more equity for barreling, meaning you can use a few more bluffs on flop, and maybe 2 more on turn and one on river. Another advantage is that they turn into nuts more often than e.g. KQo, which increases again overall EV. So I am which of advantages / disadvantages is stronger in the end. I am just sure that we can 3bet bluff +EV KQo and play postflop only when we hit nuts, whereas I am not sure if we can do it with 9Ts.

Originally posted by GingerKid
So what you achive with 3betting SC is that postflop they have more equity for barreling, meaning you can use a few more bluffs on flop, and maybe 2 more on turn and one on river. Another advantage is that they turn into nuts more often than e.g. KQo, which increases again overall EV. So I am which of advantages / disadvantages is stronger in the end. I am just sure that we can 3bet bluff +EV KQo and play postflop only when we hit nuts, whereas I am not sure if we can do it with 9Ts.

Yup, this is a pretty good summary. Although this refers to being OOP only. I would stick with the higher equity hands like KQo or Axs as IP bluffs.

Part of the reason is that we don't let opponent put us in a situation where he can turn our hand into a bluff-catcher as frequently.

Like imagine we bluff AJ OOP SB vs MP opener. Flop is J82ss. We want to bet for value/protection but we realise we don't have 3 streets. So lets say we bet flop, we are very often going to be put in a situation where we have to check turn and allow our opponent to potentially fire 2 streets us putting ourself in a situation where our hand becomes a bluffcatcher. (If our opponent fires close to a balanced range here we make no money - or even if he fires an unknown range that may or may not be balanced and we have no clear read).

IP we can just check back the turn and our opponent has one less street to make our life difficult. In other words because checking IP is way less harmful we prefer having the higher equity hands because it's easier to realise our equity.

OOP in a 3b we want to avoid checking as much as possible so I like to use the hands that have the best barreling potential and allow us to stay aggressive with the highest frequency. Having said that I will still bluff hands like KQo and AJo also. I think these hands are clearly better vs passive opponents where we are less likely to be put in a tough spot when we flop a dominated top-pair.

Vs good opponents there is also the potential issue of board coverage. If we 3bet for value high cards, and 3bet for bluff high cards then we have a hard time connecting with low textures. Whereas if we 3bet for value high card and 3bet for value mid suited connectors we connect with way more boards. I.e with SC's we can rep the boards that we miss, and also connect with the boards that miss our perceived range.

Like imagine we bluff AJ OOP SB vs MP opener. Flop is J82ss. We want to bet for value/protection but we realise we don't have 3 streets. So lets say we bet flop, we are very often going to be put in a situation where we have to check turn and allow our opponent to potentially fire 2 streets us putting ourself in a situation where our hand becomes a bluffcatcher. (If our opponent fires close to a balanced range here we make no money - or even if he fires an unknown range that may or may not be balanced and we have no clear read).

Having AJ on Jxx board and bluff catching 3 street vs perfectly balanced player, it has to be +EV and a snap call down. We block AA, JJ which makes our equity vs his range better than our pot odds. We might in this case block bluff hands like AQ, AK which would be bad, but perfectly balanced guy shouldnt bluff with showdown value.
If his range would be really strong on flop on that board, lets say JJ+, AK, than we can fold AJ on flop without problem, its like hitting air.

You can make similar comparison with a hand like 9Ts. You hit 9xx, and do you turn it into bluff? I dont, I put in the c/c range, I beat bluffs like AQ, AK, KQs, and I would call those low PP one street and give up, and would have some other premiums which can call more streets. So I dont see a problem with having a bluff catcher.

IP we can just check back the turn and our opponent has one less street to make our life difficult. In other words because checking IP is way less harmful we prefer having the higher equity hands because it's easier to realise our equity.

Yes, it is easier IP because with 2 streets assuming he is balanced, he can have less bluffs in his range, so will fire less often, meaning we get more often to the showdown.
Still he could improve it with overbetting and shoving, thus putting more bluffs in his range, and making your life equally bad.

OOP in a 3b we want to avoid checking as much as possible so I like to use the hands that have the best barreling potential and allow us to stay aggressive with the highest frequency. Having said that I will still bluff hands like KQo and AJo also. I think these hands are clearly better vs passive opponents where we are less likely to be put in a tough spot when we flop a dominated top-pair.

Why is it bad to check in 3bet pot? I guess, when you check hands you can use less bluffs. What I do as more or less standard vs regs, I cbet flop and c/r or c/c turn, so I dont have much of cbet turn range. The reason is that when you bet 1/2 pot size flop, they have to defend around 75% range, meaning that on turn they have many weak bluff catchers and overcards and similar, then when you check to them, most of regs will bet 100% bluffs. Some will bet weak pairs, some will pot control, so those who pot control make their range even more bluff heavy. Also nowadays most regs bet small like 1/3, and when on turn this mean they can use really few bluffs in range, also because most of their hands are bluff catchers means they will check river after called, again decreasing their "allowed" bluffs in range to be balanced. So I think it is a big leak from most regs, and thats why I check big part of range on turn on most boards.

Originally posted by GingerKid
So I dont see a problem with having a bluff catcher.

Assuming opponent fires a balanced ranges our EV doesn't change whether we call or fold our bluffcatchers. Our EV will be 0. So assuming our bluffcatchers always beat villains bluffs and always lose to his value hands we only can make money if villain fires an unbalanced range and we make an exploitative adjustment. (It's actually pretty nice to hold a bluffcatcher in a situation where opponent is bluffing way too often).

Anyway, generally speaking it's more favourable to be the aggressor in poker.

Originally posted by GingerKid

Now we calculate the number of bluffs on river, value_combos = 13:

bluffs_river = 13 / (1-pot_odds_river) - 13 = 6

How did you come up with 6 as a result of those pot odds?

Originally posted by w34z3l

Originally posted by GingerKid
So I dont see a problem with having a bluff catcher.

Assuming opponent fires a balanced ranges our EV doesn't change whether we call or fold our bluffcatchers. Our EV will be 0. So assuming our bluffcatchers always beat villains bluffs and always lose to his value hands we only can make money if villain fires an unbalanced range and we make an exploitative adjustment. (It's actually pretty nice to hold a bluffcatcher in a situation where opponent is bluffing way too often).

Anyway, generally speaking it's more favourable to be the aggressor in poker.

Why do you think bluff catcher will have EV = 0? I think most people miss important thing about blocker effect. In the video where you explained value:bluff ratio on the river, with which you want to make EV = 0 of any bluff catcher from villian, you forgot about the fact that most of villian combos that defend river are blocking your value combos, meaning that you have in reality less value hands in your range and unbalanced range.
Also in articles from ps.com this important fact is completely ignored when they talk about river betting frequencies.

Originally posted by GingerKid
Why do you think bluff catcher will have EV = 0? I think most people miss important thing about blocker effect. In the video where you explained value:bluff ratio on the river, with which you want to make EV = 0 of any bluff catcher from villian, you forgot about the fact that most of villian combos that defend river are blocking your value combos, meaning that you have in reality less value hands in your range and unbalanced range.
Also in articles from ps.com this important fact is completely ignored when they talk about river betting frequencies.

Yup, pretty good question.

The idea is essentially that they are making the middle of our range 0 EV. So the hands with better blockers will of course be +EV, while there are other hands in our range that might be slightly -EV as bluffcatchers.

So a reasonable question from here might be, if our EV is unchanged regardless of how many bluffcatchers we defend, can't we just bluffcatch with only the hands that have the best best blockers and expect our EV to be positive.

Potentially yes but then we might find ourself underdefending and opening ourself up to explotiation - hence putting ourself in a situation where a decent opponent knows he can overbluff since we are not calling with enough bluffcatchers.

It's pretty complicated - and keep in mind that even what we are discussing right now is a hugely oversimplified model. It won't be a perfect model until poker is solved.

Yes, agree.

There are common boards like Axxxx.

If villian plays bet bet bet, and we play call call, and have decision on river, we are very much cappt to having most of hands Ax, sets, 2 pairs. All of those hands have blocker effect. Ax blocks so many value combos, and there is hardly a hand in our range that doesnt block villians value range. And this effect happens on most high card boards which occure most frequently.

The specific spot where we have less blockers in range is when flush completes or 4 card straights, where we have just some of the hands that block good because villian is polarized on such boards.

So I think, if we are the agressor on river and know the range from villian, then we should calculate our bluff:value ratio vs each villians bluff catcher hand taking into account his blocker effect, and then averaging the bluff:value ratio over all combos.
I think this would be much more precise, and you would see that you can include in that case less bluffs in your bet range.
What do you think? Off course this is all so abstract, and nearly not applicable for humans during play, unless they use computers for help.

Yeah, don't get me wrong, I completely agree.

In fact we should just calculate different on every single board texture.

The problem is that it becomes too complex too quickly - this is why poker isn't solved. So if you are seeing ways to improve the model which I'm discussing, that's totally to be expected. Our understanding of GTO is sufficient to a basic overview of how everything should work but not to calculate everything precisely.

Being overly specific here may not actually be that useful when there are gaping oversimplifications in other areas of the model. It's like calculating a number precisely to 5 decimal points and then rounding it up the nearest whole unit =p

You are 100% mathematican or computer scientiest, right ?

Originally posted by GingerKid
You are 100% mathematican or computer scientiest, right ?

Haha no, originally a musician