Hello, at the moment I work on this article:

If we bet 2/3 pot on river then we need 1:2.33 bluff-to value.
But does this affect also the Flop and Turn bluff-to-value ratios?

I see often that these ratios are recommended:
Flop 2:1 bluff to value
Turn 1:1
River 1:2

So if I bet 2/3pot on river then I need a bluff to value ratio of: 1:2.33
But does that mean that I need on Flop now 2.33:1 instead of 2:1?

Before the advent of solvers Matthew Janda published a book called applications of no limit hold'em. In the book he suggested using these bluff:value ratios and at the time it was a very unique and interesting concept for how to be "balanced" and make it difficult for your opponents to play against you. However, since solvers/AIs/neural nets Janda himself has revised his theories in his latest book called no limit hold'em for advanced players. I would recommend reading that book (both books are a good read but if you have to pick one, read the latest one)

The definitions of "bluff" or "value" don't really make sense before the river. You bet to get your opponent to fold out his equity (not exclusive to bluffs) and to increase the size of the pot in case you win (not exclusive to value bets). IIRC one example from the book is BTN opens 88, BB call, flop K72, BB check, BTN bets 1/2 pot. BTN is ahead of BB's calling range if BB calls, yet BTN will win more money on average if BB folds; if BB calls the BTN will win less than the current size of the pot on average whereas if the BB folds the BTN wins the whole pot. So was that a value bet or a bluff? You win more if he folds so it's a bluff. You are ahead of his calling range if he calls so it's a value bet. Doesn't really make sense does it.

Don't worry about value:bluff ratios. The only reason you should be performing an action with a certain hand is if that action is the most +EV action. There's a lot of things that will not make sense to you if you think of your action on pre-river streets as value or bluff; there's a lot of bets you should be making that are very mergey and won't fit into the category of a bluff or valuebet (even on the river it's not the most important thing to be thinking about imo).

This is from my frequencies note file:

Value to Bluff Parameters:

Value Made: 2:3 to 1:2 (1.5 - 2 bluff combos)

Value Draw: 2:1 (.5 bluff combos)

We’ll estimate that Villain needs 10-15% less fold equity than the risk and reward calculation first indicates.

Subsequent Street Betting Frequency
(1-x) where x = villain's pot odds % on the previous street.

Ex
If you bet pot on the flop, villain is getting 2:1 odds, and his RE is 33%. This means Hero needs to be betting the turn 1 - .33 = 67% of the time (assuming we're betting with a perfectly polarized range) to make villain indifferent to calling our flop bet.


Pre Flop

Value to Bluff when 3-Betting: 1:2

Usually go a little tighter (1:1.5)
Even tighter OOP (1:1)

Now, if your opponent 4-bets you can fold 2/3 of the time and still put in the money with a premium range of hands.

Bluff to Value when 4-Betting = 1:.826

The frequencies at which 3-bet, 4-bet, and 5-bet bluffs must succeed will usually fall within the following percentages:
3-bet bluffs usually need folds around 67 to 70 (MDF: 30-33%) percent of the time to yield an immediate profit.
4-bet bluffs usually need folds around 54 to 60 (MDF: 40-46%) percent of the time to yield an immediate profit.
5-bet bluffs usually need folds around 40 to 50 (MDF: 50-60%) percent of the time to yield an immediate profit.

If our opponent sometimes defends by calling rather than re-raising, as he almost certainly will, then our 3-bets and 4-bets do not need to work as often to be profitable. That’s because even our weakest bluffs will sometimes get lucky on the flop. Likewise, we can also state the following:

Preflop opens need to defend by 4-betting around 25 to 30 percent of the time if they never defend by calling. This number is lower than it would otherwise be because there are usually players left to act which can help defend.

3-betting ranges which never flat when facing a 4-bet should 5-bet jam around 40 to 46 percent of the time.

4-betting ranges should usually call between 50 to 60 percent of the time when facing a 5-bet jam.

These frequencies are important for determining the ratio of value raises to bluff raises in 3-bet, 4-bet, and 5-bet situations. While they do not tell us exactly what hands belong in each range, they do allow us to check and see if a range is too heavily skewed towards value bets or bluffs.

When a player responds to a 3-bet by only 4-betting or folding, we have established that they must 4-bet between 25 to 30 percent of their preflop opening range. When facing a 5-bet, they must call between 50 to 60 percent of their 4-betting range. So by multiplying the averages of both, we see that approximately 15.2 percent of the preflop raising range should call a 5-bet.

0.152 = (0.275)(0.55)

Stated again, if the preflop raiser 4-bets 27.5 percent of the time and calls a 5-bet 55 percent of the time, he will end up felting 15.2 percent of his opening range.

Adapt to the board texture
The key to understanding our appropriate responses once we’ve figured out who has the range advantage is to recognize the significance of board texture. On more static boards, not only is the range advantage more likely to go to the preflop raiser, but there are likely to be fewer opportunities for the preflop caller to defend their range by raising, since there are few good bluffing spots on boards with no draws.

This means that unless the preflop caller is at a very significant range disadvantage, it benefits the caller to defend their range on the flop mostly by calling and planning to make a move on later streets, rather than raising in a spot where they don’t represent many value hands and their bluffs have little equity.

Conversely, on a more dynamic board with more draws, it makes sense for the preflop caller to make use of the fact that the range equities are likely to be much closer together by applying pressure to the preflop raiser’s range by raising a lot, while reducing their calling frequency to cater to the fact that there are so many turn cards that change the board dramatically. This approach is contrary to how many weaker players tend to play these boards – they do a lot of calling with their draws and tend to raise with an unbalanced range of value hands.

Value to Bluff on the Flop = 1:2

Usually 1:1.5
1:1 at the lowest (with strong range)
Move Premium Made Hands to the Marginal Made Hands (checking range) group if lacking bluffs.
In 3-bet pots, because there is less money behind, your value:bluff should be 1:1.5, or maybe even less than that. (closer to 1:1)

Marginal Made Hands to Junk = 1:1

Value to Bluff on the Turn = 1:1
Just under 1 bluff raise for every 1 value hand raise
.5 bluffs for every 1 strong draw value raised

Value to Bluff on the River = 2:1
Depends on bet size and odds offered
V:B = Villain's Pot Odds
.40psb = 3.5:1
.50psb = 3:1
.66psb = 2.5:1
.75psb = 2.3:1
1psb = 2:1

Post-Flop Raise Math
.5psb RFE: 50% MDF: 50%
.70psb RFE: 60% MDF: 40%
1psb RFE: 66% MDF: 34%

Post-flop raises need to work 50-66% ott (MDF = 34-50%).

~42% of your post flop betting range needs to defend vs a raise (42:58 = 1:1.38)

Defend often enough vs a continuation bet that your opponent cannot easily profit with ATC. If you defend too much more than that you will take away his incentive to c-bet the flop lightly. Defending too much less than that and you will incentivize Villain to c-bet the flop more lightly.

A .75psb needs to work 43% ott to break even (MDF: 57%) so Hero will need to defend ~60% ott to prevent him from being able to make a profitable bluff with any two cards.

Value to Bluff by Sizing (Descending)
Size
100/100/100 (SPR:13)
Flop: 2.38
Turn: 1.25
River: .50

100/80/75 (SPR: 9.25)
Flop: 2.1
Turn: 1.06
River: .43

100/66/50 (SPR: 6.5)
1.8
.86
.33

66/50/50 (SPR: 4.2)
1.48
.78
.33

50/50/50 (SPR: 3.5)
1.37
.78
.33

50/50/33 (SPR: 2.8)
1.22
.66
.25

Thanks Manko!

Statements similar to the above are becoming fairly common in 2P2. However, also fairly common are statements similar to the following:

“It’s important to play a balanced strategy to achieve a GTO-like strategy.”

It seems to me there is a conflict. If ‘bluff and value’ or ‘weak and strong’ are not to be considered what are we balancing?

The word balance in the context of GTO play is often misused and grossly misapplied.

No where does GTO play define that we have to play a balanced strategy consisting of "bluff hands" and "value hands". That is not a requirement of GTO play. It just so happens that often GTO play (or more precisely, the models of equilibrium play created by solvers) has components that are similar to what you define as "balance".

To answer your question, we aren't "balancing" anything. We are taking the most +EV action with each individual hand in the context of our range vs opponents range.

In my previous example of 88 betting on K72. We bet because betting is more +EV than checking. Not because we are balancing our bluffs/value. Like I said before, which one is 88 then? Bluff or value? You tell me.

Seems to me like "balance" should be replaced with a combination of fold to raise frequency and equity given to villains range or potentially hands. Those are the basic characteristics villain considers - how much equity do I have and how often will he fold if I raise?

For example, rather than say 3:2 bluff:value for 3b preflop, use F4 ~ 60%, with a likely calling range having 42% against us.