Optimal sizing on wet flops: texture based betting - Page 5

On January 29 2015 07:26 dogmeat wrote: Show nested quote + On January 28 2015 22:17 Smuft wrote:   On January 27 2015 18:43 dogmeat wrote: anyway optimal strategy maximases your ev and your argument is completely invalid, you can have balanced PSB range in your example, but its not gto b/c gto play maximases your ev, which is achieved by betting equal fractions of the pot over 3 streets (which allows highest bluff:value ratio), c/r is not an issue at all The bolded is also not true. The bigger the bet size, the more bluffs you can have so the bet size which allows the highest bluff:value ratio is a bet size of allin. it is true for limited stack sizes like 3b pots w/ 100bbs pf,

Let's see if we can agree on this:

1. Sizing that allows highest bluff:value ratio for any 1 street is all in
2. Sizing that allows highest bluff:value ratio for 2 street+ is geometric (equal fractions of pot)

also funny you say that when you previously stated betsizes doesnt matter lol,

I did not say this. I said "The EV difference of any reasonable bet size on the flop in an SRP is negligible". I also said that bet sizing effects EV more in 3b pots and probably more in spots with narrower ranges.

To clarify a bit further, I think that bet sizing matters a lot. It's just useless to try and find some GTO bet size for SRPs on the flop and probably the turn and instead spend your time finding a bet size you feel exploits the current populations tendencies

as for your previous post: this concept is demonstrated for perfectly polarized range, but also applies to polarized range w/ bluffs having x% equity and valuehands having y%, bluff:value ratio decreases but concept is still valid

The bluff:value ratio INCREASES under these conditions. Think about it:

We are usually betting more bluffs than value hands right?

With perfectly polarized ranges (value equity 100%, bluff equity 0%), the type of hand we're betting most (bluffs) have 0% equity.

With a just polarized range (generally, value equity 70-80%, bluff equity 20-30%), our bluffs have gained the same amount of equity our value hands have lost but we bet more bluffs so the overall equity of our strategy is higher and we can bet a higher bluff:value ratio.

Ironically the spreadsheet you just linked can prove this. Go to the "value" and "bluff" portion and adjust the equities, start at 100/0, then change to 80/20, then 70/30, etc. You will see that the more equity our bluffs have the higher value:bluff ratio you can bet.

*** B:V ratio will decrease when you are betting so small that your range has few bluffs but this is rare.

it was typo on my part, i actually ahve it in my older excel file

anyway i agree with you on 1. and even 2. (is betting equal fractions called geometrical?)

i will toy w/ gtorb tomorrow and try to add equities for rivers of each turn nod. i was checking out some draw heavy paired boards today and got 0.5pb > .66 > .75

btw i think you cant cbet flop 65% of timr if that results in only cbetting turn 35% or something % like in ur example

wow what's happening, liquidpoker actually has a good poker theory thread

dogmeat and smuft 2015 mod status imo

On January 29 2015 18:17 dogmeat wrote: i know you are trolling but to be a little constructive i made this excel file for you, mr 'i m too cool to bother with math and i want other ppl to do my homeworks' https://drive.google.com/file/d/0BxjknvTZAZQyekhpdDI3UldVNW8/view?usp=sharing

i am not trolling. im here for enjoyable discussion only, and think it would be worthwhile for its own sake if i can understand theory a little better, and/or help others to understand better by questioning their ideas. "doing my homework" has no value to me as i don't play poker.

thanks for the constructive contribution, but im not sure how you can use a simple toygame that doesn't allow c/r or any sort of branching in the gametree to make your argument? ofcourse ev of c/r isn't a relevant factor in a toygame that doesn't allow c/r (and nor would it ever get played by a K in AKQ game which is effectively what your toygame is).

on a sidenote this toygame becomes particularly worthless for lower flop spr situations as in my example because geometric sizing calls for such small bets that the difference between perfect polarisation and typical wa/wb in real poker becomes significant given the odds your bets offer

On February 01 2015 19:31 Romm3l wrote: on a sidenote this toygame becomes particularly worthless for lower flop spr situations as in my example because geometric sizing calls for such small bets that the difference between perfect polarisation and typical wa/wb in real poker becomes significant given the odds your bets offer

it fits well for 100bb 3bet pots, you rarely play 3streets of betting in 4b pots and i couldnt care less about shortstacking faggots

On February 01 2015 19:31 Romm3l wrote: [QUOTE]On January 29 2015 18:17 dogmeat wrote: thanks for the constructive contribution, but im not sure how you can use a simple toygame that doesn't allow c/r or any sort of branching in the gametree to make your argument? ofcourse ev of c/r isn't a relevant factor in a toygame that doesn't allow c/r (and nor would it ever get played by a K in AKQ game which is effectively what your toygame is).

whats average flop c/r % as pf 3bettor?

On February 01 2015 20:55 dogmeat wrote: Show nested quote + On February 01 2015 19:31 Romm3l wrote: on a sidenote this toygame becomes particularly worthless for lower flop spr situations as in my example because geometric sizing calls for such small bets that the difference between perfect polarisation and typical wa/wb in real poker becomes significant given the odds your bets offer it fits well for 100bb 3bet pots, you rarely play 3streets of betting in 4b pots and i couldnt care less about shortstacking faggots

good for you but not relevant vs the argument i was making

Show nested quote + On February 01 2015 19:31 Romm3l wrote: [QUOTE]On January 29 2015 18:17 dogmeat wrote: thanks for the constructive contribution, but im not sure how you can use a simple toygame that doesn't allow c/r or any sort of branching in the gametree to make your argument? ofcourse ev of c/r isn't a relevant factor in a toygame that doesn't allow c/r (and nor would it ever get played by a K in AKQ game which is effectively what your toygame is). whats average flop c/r % as pf 3bettor?

i dont know, you tell me. check turn% as well, while you're at it. and check flop% vs shortstacks too, unless you couldnt care less about how to play against presumably a nontrivial % of your opponents.

On January 29 2015 21:24 dogmeat wrote: hehe this conversation is going in circles: r0mm31 said 'ppl bet small to diminish flop c/rs" and now hes just trolling

well no, i used it as one example of a possible consideration in some situations that is different from achieving call vs fold indifference, in reply to another poster who thought sizing was all about call vs fold indifference only. not close to the same as saying its the only possible reason to bet small.

you clearly know a lot of good stuff that can contribute, would be good if youcan interact in a more constructive way

so help me ouy here, whatbetsizing do you suggest and why?

btw there are almost no shortstackers at ps nl200+

fml typed out some long shit and deleted it on accident

A GTO poker strategy is one in which the maximally exploitative counterstrategy can gain no edge and is breakeven.

And I was gonna say it seems to me like you guys are doing a lot of arguing without necessarily disagreeing

On February 03 2015 02:59 traxamillion wrote: And I was gonna say it seems to me like you guys are doing a lot of arguing without necessarily disagreeing

there is a need to establish who has the bigger potato
or who's dad is taller

It has something to do with stacksize, position, implied odds, polarity, and range strength. The rest I havent figured out yet.

The trouble with trying to figure out GTO sizing is that it doesn't account for the future. In real life, you account for stacksize / implied odds, but GTO freezes situations in a vacuum.

So the logic becomes broken, between the two.

On February 03 2015 10:25 Minsk wrote: The trouble with trying to figure out GTO sizing is that it doesn't account for the future. In real life, you account for stacksize / implied odds, but GTO freezes situations in a vacuum. So the logic becomes broken, between the two.

If you are saying that GTO doesn't change the way it plays future hands to adjust to opponents' play, that is true.
If however you are making the claim that a GTO strategy doesn't account for future streets in the same hand, that is incorrect. It in fact perfectly accounts for all past events in the hand and all possible future events in the hand.

Ah, you're right.
Actually helped me out a lot.

Then I guess the question becomes,
How do we alter sizing and frequency if there are many bad (next street) situations for our range?
How do we alter sizing and frequency if there are many good (next street) situations for our range?

Yeah, things get complex really fast. Three things I think we have to think about (would like to discuss the 3rd with people):

How do future cards affect our range compared to our opponents range?

How do future cards affect our individual hand compared to our opponents range? (For example if our range in a given spot is extremely ace heavy, but in a specific hand we don't have an ace but instead have a gutshot and a backdoor flush draw, future aces are good for our range but bad for our hand, so future aces are perfect cards to bluff).

How much information do we have about where our specific hand is at? I think even when you have a hand that has quite a bit of equity against your opponent's range, you can be in trouble because it's very hard know where you are at in the hand, and it's very easy to put money in behind or get blown off of the best hand. This is why a straight flush draw is usually more valuable against a good opponent than top pair bad kicker. You know almost exactly where you are at, and can play the hand very well. This information has inherent value, and must be very valuable to someone trying to play a GTO style. We often define a 'mistake' as putting more chips than warranted given the equity one has in a specific spot, or folding when one has the equity to make a call or raise a better choice. That said, there will be many times where a GTO strategy will be forced to do these things as well, so I believe that thinking about them as 'mistakes' is the wrong way to approach it. It is just one of the weaknesses of having a type of hand where it's hard to tell where you are at, and should be taken into account when you are constructing your decision tree.

to all the noobs, just read Smuft's replies, he's 100% correct on everything as far as I can tell.

On February 03 2015 11:04 HungarianGOD wrote: Show nested quote + On February 03 2015 10:25 Minsk wrote: The trouble with trying to figure out GTO sizing is that it doesn't account for the future. In real life, you account for stacksize / implied odds, but GTO freezes situations in a vacuum. So the logic becomes broken, between the two. If you are saying that GTO doesn't change the way it plays future hands to adjust to opponents' play, that is true. If however you are making the claim that a GTO strategy doesn't account for future streets in the same hand, that is incorrect. It in fact perfectly accounts for all past events in the hand and all possible future events in the hand.

this

I dont understand how GTORangeBuilder claims to have a GTO solution to turn and river play if you can change bet sizes within it.

Wouldn't a true GTO solution only have one betsize that cant be altered?