So the definintion of your worst bluff is when your hand you are betting has zero chance of winning at showdown but breakeven or +ev from folds.

Let's say your trying to define your very first value hand. What would be a more accurate definition:

1. A hand capable of being called by at least 1 combo of worse hand plus getting value from other hands which fold.

or

2. A hand which will win at showdown above 50% of the time when called (plus some extra ev when villain folds).

Slight nitpick I doubt that there are many times where your hand has 0 equity and generates enough folds. I guess unless you're specifically talking about river play but then I'm not sure how your valuebet question follows.

I guess I would encourage you to move beyond the terms bluff and value bet and think of hands on a spectrum based on equity and blocking effects which should also help you divide them into bets, checks, and folds.

My definition of a value bet would be a hand that makes more money on average from betting and being called than from betting and generating folds.

+1 to jg. I can't immediately think of anything I'd add.

Excellent definition.

I don't like it. Think about how strong a hand actually has to be for this to happen.

By your definition, if we bet half pot on the river expecting to win 70% of the time when called, it's still not a value bet.

Only very strong hands do better than just winning the pot outright when they get called.

I get the point you're trying to make (i.e. not seeing showdown is usually preferable) but remember you're not really stealing the whole pot if your hand has any equity on the river because part of that pot is yours if you check down.

If my hand has equity then in theory if there is a showdown I win:

Eq*P

Where P is the pot size and Eq is my equity.

So when I bluff I'm only gaining

(1-Eq)*P

When I bet and get called I need to win more than this amount. We'll use 1/2 pot bet size:

Eq*(P+2*.5*P) - .5*P >= (1-Eq)*P
2*Eq*P - .5*P >= P-Eq*P
2*Eq*P + Eq*P >= 1.5*P
3*Eq*P >= 1.5*P
Eq >= .5

So if I have more than 50% equity in the pot I do better by betting 1/2 pot and getting called than I do by betting and stealing the equity that isn't mine. I should also note you're obviously not stealing exactly 50% of the pot due to villain's folding frequency being the true determination of how much you gain when a bluff succceeds.

Feel free to comment critique my thinking that's why I post my ramblings :-)

This is also ignoring things like "betting to prevent a bluff" which makes a case for betting things that actually expect to lose more often when called.

This question comes up from time to time. I think that on betting rounds before the river, the segregation of hands into nice neat buckets doesn't describe games that involve changing hand values. For example, slow playing top set is sometimes good in no limit holdem. However top set will realize a huge pot share if you bet that hand.

On the river, the hands do fall into nice neat buckets. I like just grindings definition here.

Also, I think something can be said about how the pot grows linearly in limit games and how that contrasts no limit and pot limit games where the pot grows geometrically. We need a stronger hand to bloat the pot in big bet games. Though with our strongest hands we're winning pot^x where x = > 1. In other words you own part of your opponents stack so long as he is a rational agent.

Is it worth considering what we mean by equity in the above definition?:

A value hand is one with >50% equity when called

My thinking is a strong open draw with 51% equity when called on the flop does not really have 51% equity when villain is able to overbet the turn when you miss your draw

This is a good point. For example, whilst Q9o might technically have more equity than T9dd on a 8733ddsc board, it's much harder for the Q9o to realise their equity because they can't call any turn bet. So I'd consider this a bluff if the Q9o bets the turn, rather than a value bet.

Though Bob has already said some of thos, every hand except the nuts or a hand that has no chance of winning is some mixture of showdown equity, fold equity, and other ex-showdown equity (such as the amount of additional money earned from the opponent's stack).

Sent from my SM-G900R4 using Tapatalk

replace equity with EV (or capture factor as an EV estimate) ... so there are always two parts for a bet

1) capture factor if called
2) capture factor your fold out

so like just_grindin sad you call a bet for "value" if 1) > 2) and bluff 2) > 1) ... to expand on this, you can call a bet a "pure bluff" if 1) =~ 0 and "pure value" if 2) =~ 0

This is true OOP and on the river. We can value bet a hand that wins < 50% when called and have it be the highest ev line when villain would otherwise check back OTR.

This is true IP and on the river. (Equity-50%) * Bet > 0

Nobody has mentioned position at all, and the definition of 'value hand' has everything to do with position. Kombaiyashii touched on river value hands oop and ip. With respect to flop/turn

Everything (incl. position) is included in EV/Capture Factor.

If I bet really tiny on a river and expect to only get called by better, but raised by worse what would you call that?

spewy?

How do you define capture factor ?

block bet


fraction of pot captured.

So basically, the fraction of the pot I expect to win in the long-run (including further streets of play)

If you enter your hand, villain’s range and board cards into an equity calculator you get the showdown equity, which is the portion of the pot you expect to win. How does that differ from capture factor? By accounting for possible folds by you or villain? If so, then it appears your capture factor somehow incorporates equity realization into the showdown equity calculation. Do you know of a quantification method/formula for capture factor?

When calculating GTO stuff. People consider #1 to be the correct definition. A bluff that will be continued on the next street is also grouped into the "value" hands category even though it is technically a bluff.

I think I actually asked this question before in a thread a long time ago. I'll see if I can find it.

I imagine that most people know that we have previously had a few threads here on Capture Factor / Equity Realization. I think the concept(s) was introduced in a fairly recent math-oriented poker book by a well-known 2+2 member.

IIRC there is no analytical formula for these. They are empirical quantities based upon typical play.

This might be a bad example but I thought the idea was that some hands are able to realize their equity more easily than other hands. 22 is roughly 50% preflop vs AKo. But 22 has a very hard time in realizing its equity after not hitting a set on the flop (even when the flop has no aces or kings), facing betting on future streets, etc.

That doesn't really help define what a value hand is.

I think this is correct.

We could simplify capture factor to something like this:

Pot^x = cf

Or:

X*pot = cf

Or:

Pot/x = cf

-----

I like the first one because I have an affinity for exponents but any of the above could be used to describe cf.

I prefer to think of "value-bets" and "bluffs" as having nothing to do with actual equity, its merely the intentions of the bettor. A bot randomly selected to make bets doesn't value-bet when it happens to be betting with the best hand. A fish throwing out a donk-bet with marginal hand isn't "value-betting" even if his equity is technically best. If I'm trying to get bottom pair to call my bet with middle pair, my bet is still a value-bet even if he's folding better. The reasoning behind my bets, which decide the bet type, doesn't transform due to my opponent's actions.

Yeah I agree. The terms are practically useless for actually determining the correct way to play (i.e., picking the highest EV decision), which very often won't resemble anything we can neatly define as a value bet or bluff or semibluff, so trying to pin down an exact numerical definition completely misses the point.

They are useful as a loose way of organizing your thoughts, and that's it.

It is about a bet that doesnt lose money on the line taken, similar to a value call. How much one includes into that isnt defined. It could be a bet that contains the money in the pot plus the bet made and the opponent calling with his whole range and then there is a showdown, like calling with a value range vs the opponents whole range betting. So, the caller might fold also.

Or it could contain further action, the ev of that line, or even the ev of all the lines. It might not be a value bet if the opponent calls or raises too often, but one adjusts, if one doesnt adjust, it would be a bluff or a semi bluff. Semi bluff might be like a block bet and both might lose money when considering just that line or and the opponent.