Open Side Menu Go to the Top
Register
the Vegasmidas blog the Vegasmidas blog

05-22-2015 , 09:55 PM
I once folded every hand except QQ+ and AK for 6 hours to hit a bonus on PS back when the VPP's were in dealt mode. It was less boring than reading that post
the Vegasmidas blog Quote
05-23-2015 , 05:38 AM
Quote:
Originally Posted by theSimonman
...
55% * (SB+BB+2.5BB's) - 12% * 7.5BB's + 33% * (equity * (8.5BB's+SB+BB) + (1-equity) * -7.5BB's) = 0
_______
Aight, solve that formula for the needed equity, and you'll get Part 2.
Equity = 57,4%
the Vegasmidas blog Quote
05-23-2015 , 11:54 AM
Quote:
Originally Posted by Dr.FatCat
I once folded every hand except QQ+ and AK for 6 hours to hit a bonus on PS back when the VPP's were in dealt mode. It was less boring than reading that post
boring its u on this thread... thanks
the Vegasmidas blog Quote
05-23-2015 , 11:54 AM
SHOTS FIRED

BANG BANG BANG
the Vegasmidas blog Quote
05-24-2015 , 12:45 PM
In

Fun thread, gl
the Vegasmidas blog Quote
05-24-2015 , 03:34 PM
Very enjoyable reading. More PLO talk pls!
the Vegasmidas blog Quote
05-25-2015 , 01:50 AM
Quote:
Originally Posted by theSimonman

A comprehensive intro to building 3betting ranges - Part 1

Terminology:
3bet - Re-raising against a preflop raise
EV - expected value. The value (compared to folding) that a decision has.
Equity - The chance of winning a hand with no further action
Fold to 3bet % - How often somebody folds to your 3bets. You should be able to find this stat somewhere in your tracker/hud.
Fold Equity or FEQ - The chance (typically as a percentage, but is equivalent to a BB/hand or share of the pot / hand winrate) of getting a fold
Value range - The hand range that 3bets for value, meaning that your expected EV is bigger than half the pot + Fold Equity for every single hand.
Bluff range - The hand range that 3bets as a bluff, generally simply meaning that your EV is higher than 0.
So, let's say you're in the BB, and you want to create a 3betting range against the player on the button. There's loads of things that we don't know:
1) How does the player on the button play?
2) How would we find our value hands?
3) How could we find our bluff hands?
We'll start by thinking about a very simple idea:
Betting or raising in poker creates a trade-off between equity and fold equity.

We'll think about the validity of this by assuming that there was no trade-off, meaning that your equity would stay the same no matter how much fold equity you generate.
Before, our expected value was equity * pot.

Now, our expected value is fold * pot + call * equity * pot.
Essentially, we add a case where we have 100% equity (we always win the pot) to the case where we have the same equity as before. The result must be a bigger number than before.
We would always bet or raise if we put in any chips.

In reality, things look a bit differently:

equity * pot (as before) compared to FEQ * pot + (equity-x*FEQ) * pot. So now, for every bit of FEQ that we generate, we take away some of our equity (that's what the "x*FEQ" means). This makes sense. Now it's our goal to find the spots where the trade-off is advantageous, and that's what we'll do next.

Our approach is the following:
1) Find out how much equity we need in order to consider bluffing
2) Based on that, find hands that could 3bet bluff
3) Out of those hands, find the hands with the best FEQ-equity trade-off
1) How much equity do we need?
fold * pot - 4bet * 3bet-amount + call * (equity * (pot) + (1-equity) * -3bet-amount) = 0
We either get a fold, a 4bet, or a call. If we get a fold, then we win the pot. If we get 4bet, then we fold our bluff and lose our 3bet-amount. If we get called, then our equity determines how often we win or lose the pot.

In order to solve that formula for the needed equity, we need 2 numbers and one range:
calling range - What does the player on the button call with?
We'll need this range in order to look at the equities of different hands. Before we can do this, we'll find all the other numbers first.

fold - How often does the player on the button fold? You'll find this number by feel, or by looking at your HUD. For this example, we'll simply assume 55%.

4bet - How often does the player on the button 4bet? Again, you'll have to estimate, or look up.
The way we'll find this number right now is pretty simple:
[TT-AA,AK] / 1326 combos = 46/1326 = 3.5%
This is the 4bet value range. You can simply adjust it if you think people 4bet tighter, or looser for value.
Now, generally somebody should 4bet bluff around half the time, which would be another 3.5% range. However, i would assume that people at small stakes don't 4bet bluff enough.
[bluff range] < 3.5% = 2.5%

Entire 4bet range = 6%
So, this Button player folds 55%, and 4bets a 6% range. In order to continue, we'll have to estimate how often he raises the button in the first place. I'll simply assume 50%.

This means that he 4bets 6/50, or 12% of the time.
fold = 55%
4bet = 12%
call = 100% - 55% - 12% = 33%
We can solve for the equity now:
55% * pot - 12% * 3bet-amount + 33% * (equity * (pot) + (1-equity) * -3bet-amount) = 0
OK, not quite. We also need to think about the betting and raising amounts.
Let's assume the following:

The player on the button opens to 2.5 Big Blinds, and we 3bet to 8.5 Big Blinds. This means that we risk 7.5 BB's for our 3bet.
55% * (SB+BB+2.5BB's) - 12% * 7.5BB's + 33% * (equity * (8.5BB's+SB+BB) + (1-equity) * -7.5BB's) = 0
_______
Aight, solve that formula for the needed equity, and you'll get Part 2.
All this for no limit texas holdem?
the Vegasmidas blog Quote
05-25-2015 , 07:46 AM
he gave you logic , you could probably apply it to 'pot-smoking' or whatever you want... just plug in different variables
the Vegasmidas blog Quote
05-27-2015 , 06:47 AM
Have you ever noticed the voice in your head?

My voice is VERY active while i play poker. It usually says:

"That guy is so bad"
"That play is so bad"
"Course a russian raises me here"
"What a bad turn card"
...

And it keeps repeating itself, and it's goal is to supercharge me with emotions (which it sometimes achieves).

btw, the F.UCK do i share such a huge-a.ss post about 3betting and nobody finds the right equity number for me?!
the Vegasmidas blog Quote
05-27-2015 , 07:22 AM
Bad at math, but i took a shot and came up with 20,35%
the Vegasmidas blog Quote
05-27-2015 , 10:57 AM
Quote:
Originally Posted by Klemi
Bad at math, but i took a shot and came up with 20,35%
I take it. Part 2 in progress.
the Vegasmidas blog Quote
05-27-2015 , 02:20 PM
equity = 20.3463 %

the lack of correct format made that really hard to put into wolfram.
the Vegasmidas blog Quote
05-28-2015 , 10:07 AM
The player on the button opens to 2.5 Big Blinds, and we 3bet to 8.5 Big Blinds. This means that we risk 7.5 BB's for our 3bet.
55% * (SB+BB+2.5BB's) - 12% * 7.5BB's + 33% * (equity * (8.5BB's+SB+BB) + (1-equity) * -7.5BB's) = 0
-> equity = 20%
That's quite boring (why?), but it should make you think (why?).
Little mind game:
If the minimum equity that we need (against the calling range) is 20%, then which hands should we 3bet?

TT-22,AQs-A2s,KQs-K8s,QJs-QTs,JTs-J9s,T9s-T8s,98s-97s,87s,76s,AQo-ATo,KQo-KJo

That's the kind of calling range that we could expect (based on our frequencies).
At this point, I'll need you to understand that this approach of finding a 3bet range isn't THE approach. It doesn't even yield great results in every spot. But there's one thing that we want to focus on here, and i think that's very important:

The FEQ/EQ trade-off

I've mentioned this before, and I've even defined it. But what are we doing here exactly?

Take a hand like T9s.




And look at the Button's calling range vs his entire range.
What stands out?
- He folds a LOT of better T's: JTo,QTo,KTo.
- He calls some hands that we dominate: T8s,98s.
- He calls many hands that we are a slight dog against: KQo,AQo,QJo,...
Based on that, let's focus on the FEQ/EQ trade-off. I'll re-define it for you:

The FEQ/EQ trade-off is advantageous if a hand has as much equity as possible against the calling range and as little as possible against the folding range.

Right. We want to fold out better hands, while getting called by hands that don't have us crushed. And if somebody 4bets hands that have us crushed: Excellent!
If somebody 4bets JJ-AA and AK, then the best hands to 3bet/fold must be hands like [insert hand].

OK, it's your turn again:
You have a 4bet range and a calling range (and a folding range). You can actually treat the 4bet-range and folding-range as one big range for this exercise (why?).
Find a group of hands that have the best FEQ/EQ trade-off, meaning that it's equity versus the button's entire range is similar to it's equity against his calling range.
Give me a reasonable group of hands (and a reasoning) that fit those criteria and you'll get part 3.
the Vegasmidas blog Quote
05-28-2015 , 10:57 AM
Quote:
Originally Posted by Wafflecrunch
equity = 20.3463 %

the lack of correct format made that really hard to put into wolfram.
Can you show what you put into wolfram? I am completely new to this.
the Vegasmidas blog Quote
05-28-2015 , 11:00 AM
Lol math

What does that wall of numbers have to do w poker?
the Vegasmidas blog Quote
05-28-2015 , 12:32 PM
Quote:
Originally Posted by Jajajaja
Lol math

What does that wall of numbers have to do w poker?
He can't help everyone, find it yourself man...
the Vegasmidas blog Quote
05-28-2015 , 02:17 PM
Quote:
Originally Posted by Palmersquall
Can you show what you put into wolfram? I am completely new to this.



@Simon You mean we have a 5bet range and a 4bet calling range? Also considering oop play how often do we really get to realize especially draw equity in 3bet/4bet pots, so we should be aware of textures where we expect to get backplayed/stabbed at on flop/turn/river a lot more, meaning we have to be aware of which board textures are favorable and we for 1) have to cover them and 2) need villain to not entirely realize what is good or bad to stab.

In my opinion you can either strengthen your 4bet calling range with the addition of certain monster hands like some AQs, some AKo, AQo, QQ plus some hands that avoid domination scenarios and provide low board coverage with draw potential, while your 5bet Range should contain hands that avoid 20% equity situations towards more equity when called, so be something like KK+,all AKs, some AQs, A5s, some A4s. It ALL depends on how villains 4bet bluffing range is constructed. So constructing a good FEQ/EQ-weighted range has to be correlated towards BU 4betting range and frequencies.
Treating your 4bet call/fold range as one entirely does not really help to significantly generalize the range.
the Vegasmidas blog Quote
05-28-2015 , 03:29 PM
Quote:
Originally Posted by Wafflecrunch
It ALL depends on how villains 4bet bluffing range is constructed. So constructing a good FEQ/EQ-weighted range has to be correlated towards BU 4betting range and frequencies.
Treating your 4bet call/fold range as one entirely does not really help to significantly generalize the range.
I am not gonna make it a habit to reply to posts, but i couldn't resist.

Everything I've been doing / asking you guys to do is geared towards one idea, which is the fact that you generally lose equity by raising.
  • We want to decide which hands lose the least equity by 3betting.
  • In order to do that, you simply need to know your equity against the range that calls and your equity against the range that does not call (because we would fold our bluffs against a 4bet anyway).
  • The worse our equity against any hand that doesn't call, the better, because we never play a pot against those hands.
  • All that said, doing this won't give you a perfect range, and it certainly won't give you any A4 suited kind hands, or AQ.

However, we might be able to find a piece of our bluff range this way, and if you've never done this before then i think this little exercise might end up teaching you quite a lot about poker (theory).
There's much talk about "Naah don't bet, you'll only get called by better hands" and "yea you have to bluff cause you're at the bottom of your range", but all that is useless preflop, because "bluffs" and "value hands" are very hard to define, and often it's quite murky.
Once you introduce the trade-off the way we did, suddenly this whole theory is applicable both to preflop and postflop situations as far as bluffs are concerned. True, we haven't taken 4bets into account. But i did say there was gonna be a part 3 (and maybe part 4).
the Vegasmidas blog Quote
05-28-2015 , 04:51 PM
I feel this is what my blog would be like if my life had not taken a sudden divergence 5ish years ago when I had a kid... and if I could write... and was funny. What I'm saying is from one old guy who has been doing this for 10+ years to another, good luck. Very enjoyable read so far.
the Vegasmidas blog Quote
05-28-2015 , 04:53 PM
I am just so tight up in certain thinking patterns to cross the hump of "should I really turn this hand into a 3bet bluff knowing I will not even be able to realize its equity in a 3bet pot when called, while I can do more with it in a raised pot in a better equity scenario". I kind of see of what this exercise is intended to do, judging hands according to their equity when called, 3betted, or even 4bet called and thus constructing ranges to min/max equity lost by gaining all advantage of preflop foldequity to be had. As I play at stakes where rake is a huge factor this is even more important I feel.

I will make my attempt at constructing those ranges. I just missed the point about this being an exercise in realizing how FEQ can be used to manipulate ranges so that you lose the least EV from aggressive preflop actions.
I kind of remember watching you play on Bryces stream and curious about some 3bet situations you chose to engage in with certain hands that for me seemed to be used completely different from what I suspect their max ev would be at. Seeing where you are coming from it kind of makes sense now to me. So in terms of that I think I already learned something and I will try to incorporate some of it into my studies.
the Vegasmidas blog Quote
05-28-2015 , 05:45 PM
Ok, I completely focussed on the hands that lost the least equity when 3betted compaired to playing against a 50% preflop opening range, while being smashed against 4bets.

Pretty much what became obvious that pairs bigger than 77 and high cards would perform much better against the 50% range, as against the defend range. Especially Ace highs had too much value passively defended and even at least the suited ones performed nearly as good vs the defending range as the 4bet range.

So it quickly became transparent that suited compared to offsuit low cards and low pairs would suffer the least equity loss playing against a defending range, while giving them up against a 4bet while playing worse against 50% opening due to much better low board coverage from BU.

Now it was all about judging gappers vs connectors in terms of their playability and equity against 50% vs 3bet defend. Again it seemed that gappers seem to be less equity loss struck than connectors, only if you went down to 32s you would experience a gapper like behaviour in terms of equity.

So my 3bet bluffing range, giving all the parameters (50% pfr, defend range: TT-22,AQs-A2s,K8s+,QTs+,J9s+,T8s+,97s+,87s,76s,AQo-ATo,KJo+, 4bet range: JJ+,AKs,AKo) would look like this:

77-22,74s+,62s+,52s+,42s+,32s (6.64% of hands total)

When we remember that we need 20% equity at least against the defending range in order to achieve neutral ev, this is how we look like against BU defend range:



This range overall only loses ~2.6% equity compared to facing a 50% PFR range:



While getting smashed against the 4bet range:



If you would have to get a little less bluff heavy (6.6% is a lot), the hands to deduct would be the hands that would get the least crushed against the 4bet range, while being equally good against the other two ranges, which obviously would be the pairs:



But deducting all pairs would mean getting down to 4% overall, which might make our 3betting range overall become too nutted and allowing BTN to defend tighter ranges against our overall frequency. So in order to get the "bluff" segment up we would start adding pairs starting with deuces up to fives and even up to 77 if BU really defends a wide standard range.

Another thing that became quite apparent was that we cover low boards much much better than the buttons defend range. Meaning we either flop overcard equity, a pair or an overpair with our entire range and put BU in a ton of ugly spots.

I hope this was what you were looking for Simon.
the Vegasmidas blog Quote
05-29-2015 , 02:58 AM
Quote:
Originally Posted by theSimonman
The player on the button opens to 2.5 Big Blinds, and we 3bet to 8.5 Big Blinds. This means that we risk 7.5 BB's for our 3bet.
55% * (SB+BB+2.5BB's) - 12% * 7.5BB's + 33% * (equity * (8.5BB's+SB+BB) + (1-equity) * -7.5BB's) = 0
-> equity = 20%
We 3b on BB to 8.5bb total, why is there our 3b+sb+bb ?

-
Thanks for the excellent read so far!
the Vegasmidas blog Quote
05-29-2015 , 05:02 AM
Quote:
Originally Posted by moment
We 3b on BB to 8.5bb total, why is there our 3b+sb+bb ?

-
Thanks for the excellent read so far!
That's not our 3bet, but button's initial raise + call of our 3bet. When we win, we win his investment + SB and BB that are dead money from the start. You could also set up that part of the equation(when we're called - after 33% *) this way:

equity * [whole pot] - investment = equity * (8.5bb * 2 + sb) - 7.5bb

I find this more intuitive; you win your equity share of the whole pot after he calls and deduct your investment(note that blinds are a forced investment that counts as dead money). Using this equation you only need to change the equity part one place. This way it's easier to see how changing your (estimated realized) equity will affect the outcome of the EV calc. Also you can deduct rake easier by adding another factor:

equity * [whole pot] * (100% - rakepercent) - investment = equity * (8.5bb * 2 + sb) * 0.955 - 7.5bb

If you're a baller, you might also need to consider rake cap. If rake is capped after 3bet is called, simply subtract rake cap from [whole pot]. Not sure why I think ballers need to be told this stuff though.

BTW SimonMan, you are not alone. I'm also only 40ish. I live in Norway, too. True story.

Dissing Norway like that. Not cool, guys.
the Vegasmidas blog Quote
05-29-2015 , 05:14 AM
Tacohead, thanks for pointing out!

Good note to mind rake as it changes our ranges
the Vegasmidas blog Quote
05-29-2015 , 06:36 AM
If you start making this about stakes, there are so many more parameters thrown in making the equation stake dependent, talk about the smaller SB on NL25 for example and of course the rake, which micro players should always be aware of, if you play on stakes where you hit the rake cap already in a raised pot when a cbet got called it does a lot less to change the required equity compared to scenarios where you don't cap out just by getting to the flop in a 3bet pot. Speaking about this model of course focussing on FEQ vs EQ.
the Vegasmidas blog Quote

      
m