Two Plus Two Publishing LLC Two Plus Two Publishing LLC
 

Go Back   Two Plus Two Poker Forums > >

Poker Theory General poker theory

Reply
 
Thread Tools Display Modes
Old 06-09-2021, 07:10 AM   #51
Yeodan
veteran
 
Yeodan's Avatar
 
Join Date: Mar 2008
Location: Belgium, Antwerp
Posts: 2,945
Re: How does GTO win?

Quote:
Originally Posted by dude45 View Post
Will quantum computers be able to solve poker ?
It depends.

It's been a while since I've looking into quantum computers.

But as far as I understand you can't really compare them to normal computers.
You can't just say they're much faster better normal computers, that's not how it works.

Their computation power is insanely much higher than even our best normal computers, but it's hard to get the actual answers back out.

If they manage to get a solver to run on a quantum computer, then yes, it can probably solve poker, or very close to it.
It would probably solve for 10 sizing's on each street and every possible flop within a few seconds.

The real question is, will they ever be able to get this kind of software to run on a quantum computer.
I believe they will, but as far as I know we're not capable of that just yet.


That being said, aren't there real time solvers being used already?
With the right setup you could probably run a range vs range on a specific board pretty quickly not?


I also don't think poker getting solved will be an issue.
Look at chess, computers have been way beyond human players for a good while now.
But chess is still very popular, maybe even more popular than ever?

Actually completely learning solved poker would take years and years of study, probably a lifetime.
I doubt it's possible for a human to play perfect NE poker, even if they had instant-access to all the answers while studying.
Yeodan is online now   Reply With Quote
Old 06-09-2021, 08:21 AM   #52
aner0
old hand
 
aner0's Avatar
 
Join Date: Oct 2013
Location: https://discord.gg/TVbwaBHmhx
Posts: 1,559
Re: How does GTO win?

Quote:
Originally Posted by Yeodan View Post
That being said, aren't there real time solvers being used already?
With the right setup you could probably run a range vs range on a specific board pretty quickly not?
No. This takes at least multiple minutes per flop
aner0 is offline   Reply With Quote
Old 06-09-2021, 01:31 PM   #53
JustLuck
veteran
 
Join Date: Jul 2007
Location: germanys first female supernova
Posts: 2,107
Re: How does GTO win?

Quote:
Originally Posted by aner0 View Post
Defending "10-sizing style" vs a 1 sizing betting range is underperforming compared to the maximally exploitative strategy, yes. But not by a long stretch, by a very short one, we're talking in the 0.5%s.

This is wrong though. Depending on which spot we are talking about. EVs can and will change massively when you are incorrectly solving for 25% as part of multi sizing strategy when compared to 25% as a single size.

Not only will EV's change but also the resulting strategy on every street will look vastly different. If you were to nodelock your incorrectly calculated strategy vs your opponents actual strategy you'd be losing heaps of EV on the flop already.

Unfortunately there's no way to measure this as it would require far too much nodelocking but i threw a quick sim together to make my point:

player 1 is ip
EVs's taken after flop bet:
top sim. small bet only / optimal bb response vs the villain's actually strategy
bottom sim: small bet only/ optimal response nodelocked from a multi sizing flop sim.

https://imgur.com/a/GVDHNom

edit: i suppose whether that's too small or big of an EV loss is relative but the point is that your strategic response isn't optimal. There is no "close" or "good enough" in GTO. Strictly speaking the response from your 10 sizing is incorrect and loses EV.

Last edited by JustLuck; 06-09-2021 at 01:44 PM.
JustLuck is offline   Reply With Quote
Old 06-09-2021, 02:03 PM   #54
aner0
old hand
 
aner0's Avatar
 
Join Date: Oct 2013
Location: https://discord.gg/TVbwaBHmhx
Posts: 1,559
Re: How does GTO win?

Quote:
Originally Posted by JustLuck View Post
This is wrong though. Depending on which spot we are talking about. EVs can and will change massively when you are incorrectly solving for 25% as part of multi sizing strategy when compared to 25% as a single size.

Not only will EV's change but also the resulting strategy on every street will look vastly different. If you were to nodelock your incorrectly calculated strategy vs your opponents actual strategy you'd be losing heaps of EV on the flop already.

Unfortunately there's no way to measure this as it would require far too much nodelocking but i threw a quick sim together to make my point:

player 1 is ip
EVs's taken after flop bet:
top sim. small bet only / optimal bb response vs the villain's actually strategy
bottom sim: small bet only/ optimal response nodelocked from a multi sizing flop sim.

https://imgur.com/a/GVDHNom

edit: i suppose whether that's too small or big of an EV loss is relative but the point is that your strategic response isn't optimal. There is no "close" or "good enough" in GTO. Strictly speaking the response from your 10 sizing is incorrect and loses EV.
Idk what your sims are. All I know is that your first claim was that "So for practical purposes less sizings are usually better.", which is false by definition, because people in game are not playing your abstraction, they are playing poker and they can choose whatever sizings they want, and if you defend 1-sizing style, a studied enough human will beat you even if you play exactly like your sim.

You have moved the goalpost down to the trivial statement that GTO doesn't make as much money as maximally exploitative, which is something no one with a brain would dispute.

"So for practical purposes less sizings are usually better." is as silly (probably more) as saying in RPS playing Paper is definitely the best. If everyone listened to this advice Scissors would actually be best, not paper. By the same token, if everyone studies 1 sizing sims, you will exploit them very easily by studying multi-sizing sims, even if they were to apply their outputs to a much better precision, and that is shown to be evident in the terms of the prop bet i offered you (Multisizing strategy always makes more money against 1 sizing strategies than against multisizing opponent) that you smartly rejected.

I know (from my sims) of preflop spots in which I can make >100bb/100 (exploitative, so over Nash) with certain combos by exploiting a PERFECT 1-sizing sim very very easily, by playing strategies outside the abstraction.

Last edited by aner0; 06-09-2021 at 02:11 PM.
aner0 is offline   Reply With Quote
Old 06-09-2021, 02:28 PM   #55
JustLuck
veteran
 
Join Date: Jul 2007
Location: germanys first female supernova
Posts: 2,107
Re: How does GTO win?

Quote:
Originally Posted by aner0 View Post
"So for practical purposes less sizings are usually better.", which is false by definition, because people in game are not playing your abstraction, they are playing poker and they can choose whatever sizings they want, and if you defend 1-sizing style, a studied enough human will beat you even if you play exactly like your sim.
smartly rejected.
As for what's practical: Teaching implementable strategies. Implementable means limited bet sizes . If you think you can accurately pull of a 4 sizing flop strategy I have a real propbet offer for you.

I'd take a human studying sims that are limited in sizings over a human studying sims with 10 flop sizings all day everyday.

As for your propbet: I offered to show you that using the response you get from your multi sizing sim will actually lose significant EV compared to the actual solution. The resulting strategy you want to use loses significant EV on the flop alone. Sadly there's no way to measure how much EV it would lose across the entire game tree.

Poker isn't as simple as mashing 4 sizings into monker and using the output as your default strategy. You need to solve for single size if that's what your opponent plays. Or for 2 sizes if that's their strategy.

Solvers only solve for the strategy you input. Nothing else and nothing less.
"covering the size" means exactly nothing. That's why a 10 sizing sim will still be garbage. No matter how much time and resources you put into it.
It's too hard to learn from as it doesn't accurately represent the game tree.

Ultimately the way to use poker solvers is to try and roughly represent the game tree. Not mash in unlimited sizings so you have a response to everything.

I showed you the response is wrong and loses EV.


Edit: I feel like i've been over this very recently when I called zenith poker snakeoil. You wouldn't happen to be affiliated with them?
JustLuck is offline   Reply With Quote
Old 06-09-2021, 02:34 PM   #56
aner0
old hand
 
aner0's Avatar
 
Join Date: Oct 2013
Location: https://discord.gg/TVbwaBHmhx
Posts: 1,559
Re: How does GTO win?

Quote:
Originally Posted by JustLuck View Post
As for what's practical: Teaching implementable strategies. Implementable means limited bet sizes . If you think you can accurately pull of a 4 sizing flop strategy I have a real propbet offer for you.

I'd take a human studying sims that are limited in sizings over a human studying sims with 10 flop sizings all day everyday.

As for your propbet: I offered to show you that using the response you get from your multi sizing sim will actually lose significant EV compared to the actual solution. The resulting strategy you want to use loses significant EV on the flop alone. Sadly there's no way to measure how much EV it would lose across the entire game tree.

Poker isn't as simple as mashing 4 sizings into monker and using the output as your default strategy. You need to solve for single size if that's what your opponent plays. Or for 2 sizes if that's their strategy.

Solvers only solve for the strategy you input. Nothing else and nothing less.
"covering the size" means exactly nothing. That's why a 10 sizing sim will still be garbage. No matter how much time and resources you put into it.
It's too hard to learn from as it doesn't accurately represent the game tree.

Ultimately the way to use poker solvers is to try and roughly represent the game tree. Not mash in unlimited sizings so you have a response to everything.

I showed you the response is wrong and loses EV.


Edit: I feel like i've been over this very recently when I called zenith poker snakeoil. You wouldn't happen to be affiliated with them?
I'm not affiliated with anybody, I know them though, and on this specific issue they happen to be correct.

I don't need to apply the multisizing strat more accurately than the 1-sizing person to beat them, I only need to exploit them in very basic ways by stepping outside of their abstraction, with little regard for balance.

A solver output is not something you implement, it's something you learn from, ingame no one is implementing nothing remotely close to any 1 or 200 sizing sim, and everyone has a **** ton of leaks, so you can keep trying to "implement" your strategy where I very easily pick up on the most fundamental leak in a 1-sizing abstraction and hammer it for free $.
aner0 is offline   Reply With Quote
Old 06-09-2021, 02:44 PM   #57
JustLuck
veteran
 
Join Date: Jul 2007
Location: germanys first female supernova
Posts: 2,107
Re: How does GTO win?

Quote:
Originally Posted by aner0 View Post
I'm not affiliated with anybody, I know them though, and on this specific issue they happen to be correct.

I don't need to apply the multisizing strat more accurately than the 1-sizing person to beat them, I only need to exploit them in very basic ways by stepping outside of their abstraction, with little regard for balance.

A solver output is not something you implement, it's something you learn from, ingame no one is implementing nothing remotely close to any 1 or 200 sizing sim, and everyone has a **** ton of leaks, so you can keep trying to "implement" your strategy where I very easily pick up on the most fundamental leak in a 1-sizing abstraction and hammer it for free $.
My god if you are a zenith customer that explains everything. Let me put it blundly. They are wrong about it. Their product is basically a scam.
And you fell for it.

Don't believe me? https://forumserver.twoplustwo.com/2...ghlight=zenith


but anyway keep trying to "cover" every size if that's what you want. It won't give you the right answer (and you know that) but it will give you an answer that's "close enough" (whatever that means).

So enjoy your "close enough" answers that are calculated based on ranges that aren't even real. Maybe watch some finding equilibrium while you are at it.
JustLuck is offline   Reply With Quote
Old 06-09-2021, 02:50 PM   #58
aner0
old hand
 
aner0's Avatar
 
Join Date: Oct 2013
Location: https://discord.gg/TVbwaBHmhx
Posts: 1,559
Re: How does GTO win?

Quote:
Originally Posted by JustLuck View Post
My god if you are a zenith customer that explains everything. Let me put it blundly. They are wrong about it. Their product is basically a scam.
And you fell for it.

Don't believe me? https://forumserver.twoplustwo.com/2...ghlight=zenith


but anyway keep trying to "cover" every size if that's what you want. It won't give you the right answer (and you know that) but it will give you an answer that's "close enough" (whatever that means).

So enjoy your "close enough" answers that are calculated based on ranges that aren't even real. Maybe watch some finding equilibrium while you are at it.
Could be that, o could be that they are just your fair competitor and you like to slander your competitors, like you've already done to Bencb.

Just take the L already and move on
aner0 is offline   Reply With Quote
Old 06-09-2021, 02:57 PM   #59
JustLuck
veteran
 
Join Date: Jul 2007
Location: germanys first female supernova
Posts: 2,107
Re: How does GTO win?

Quote:
Originally Posted by aner0 View Post
Could be that, o could be that they are just your fair competitor and you like to slander your competitors, like you've already done to Bencb.

Just take the L already and move on
Well for one Zenith isn't a competitor and RYE is a joke so...
Zenith and RYE are both scams that no professional player uses. DTO has 100s of professionals as clients.

Maybe I know a thing or two about running sims as learning material.

As for taking the L. i have shown that your 10 sizing solution loses EV to the 1 sizing sim. That's all i was claiming. And no. The EV loss isn't small. Your response isn't "close enough". And you will get murdered if you apply a 10 sizing sim against a 1 sizing villain.
Whether he'd get murdered if the spot was reversed doesn't matter. Your strategy will still be incorrect.

Precision matters in sims and you can't simply wing it by adding more sizings to you sims. That's madness. The resulting ranges will never reflect what reality looks like
JustLuck is offline   Reply With Quote
Old 06-09-2021, 03:03 PM   #60
aner0
old hand
 
aner0's Avatar
 
Join Date: Oct 2013
Location: https://discord.gg/TVbwaBHmhx
Posts: 1,559
Re: How does GTO win?

Quote:
Originally Posted by JustLuck View Post
As for taking the L. i have shown that your 10 sizing solution loses EV to the 1 sizing sim.
sure buddy
aner0 is offline   Reply With Quote
Old 06-09-2021, 03:04 PM   #61
JustLuck
veteran
 
Join Date: Jul 2007
Location: germanys first female supernova
Posts: 2,107
Re: How does GTO win?

Quote:
Originally Posted by aner0 View Post
sure buddy
Sorry I ruined zenith poker and your library of 10 flop sizing sims for you
If you ever wanna learn how to actually study solver strategy I can recommend a few tools for you.
JustLuck is offline   Reply With Quote
Old 06-09-2021, 03:08 PM   #62
aner0
old hand
 
aner0's Avatar
 
Join Date: Oct 2013
Location: https://discord.gg/TVbwaBHmhx
Posts: 1,559
Re: How does GTO win?

U could've saved us the silly argument and just try to sell ur stuff
aner0 is offline   Reply With Quote
Old 06-09-2021, 03:42 PM   #63
JustLuck
veteran
 
Join Date: Jul 2007
Location: germanys first female supernova
Posts: 2,107
Re: How does GTO win?

Quote:
Originally Posted by aner0 View Post
U could've saved us the silly argument and just try to sell ur stuff
i don't care about selling stuff. I'm just here to help people understand that there is a huge difference between playing against 1 size and playing against one of the sizes as part of a multi sizing strategy. Even if they are exactly the same bet size.


I put a lot of time and effort into building sims. Unlike zenith who just run some of the worst sims i've ever seen and it quite frankly bothers me that they get away selling their garbage to the community.

All their stuff uses poor abstractions. They run 3 way sims to 3 iterations. They use 10 buckets for turn/river in monker etc etc. It's frankly some of the worst stuff i've seen.
This is painfully obvious to anyone who's run their fair share of preflop sims
JustLuck is offline   Reply With Quote
Old 06-09-2021, 03:49 PM   #64
aner0
old hand
 
aner0's Avatar
 
Join Date: Oct 2013
Location: https://discord.gg/TVbwaBHmhx
Posts: 1,559
Re: How does GTO win?

Quote:
Originally Posted by JustLuck View Post
there is a huge difference between playing against 1 size and playing against one of the sizes as part of a multi sizing strategy. Even if they are exactly the same bet size.
my point exactly
aner0 is offline   Reply With Quote
Old 06-09-2021, 03:58 PM   #65
JustLuck
veteran
 
Join Date: Jul 2007
Location: germanys first female supernova
Posts: 2,107
Re: How does GTO win?

Quote:
Originally Posted by aner0 View Post
my point exactly
Nice. So we do agree that "covering a betsize" doesn't really mean anything?

Glad to see we got there in the end!
JustLuck is offline   Reply With Quote
Old 06-09-2021, 04:06 PM   #66
aner0
old hand
 
aner0's Avatar
 
Join Date: Oct 2013
Location: https://discord.gg/TVbwaBHmhx
Posts: 1,559
Re: How does GTO win?

Quote:
Originally Posted by JustLuck View Post
Nice. So we do agree that "covering a betsize" doesn't really mean anything?

Glad to see we got there in the end!
it means taht we can beat the worse abstraction defense strategies way harder than they can beat ours, pretty simple
the mistake a 1 sizing sim makes by playing the wrong adaptation to the multi betting range is way more severe, because "there is a huge difference between playing against 1 size and playing against one of the sizes as part of a multi sizing strategy. Even if they are exactly the same bet size."

say for example, if you defend BB v SB RFI with a range generated by a no limps sim, the guy who limps will absolutely demolish you. The other way around it's not as severe
aner0 is offline   Reply With Quote
Old 06-09-2021, 04:17 PM   #67
JustLuck
veteran
 
Join Date: Jul 2007
Location: germanys first female supernova
Posts: 2,107
Re: How does GTO win?

Quote:
Originally Posted by aner0 View Post
it means taht we can beat the worse abstraction defense strategies way harder than they can beat ours, pretty simple
the mistake a 1 sizing sim makes by playing the wrong adaptation to the multi betting range is way more severe, because "there is a huge difference between playing against 1 size and playing against one of the sizes as part of a multi sizing strategy. Even if they are exactly the same bet size."

say for example, if you defend BB v SB RFI with a range generated by a no limps sim, the guy who limps will absolutely demolish you. The other way around it's not as severe
No one is arguing the opposite. This isn't a competition where your choices are 10 sizings or just one. Both are really bad choices. The 10 sizing solution will give very poor results just like the 1 sizing solution.


if you want to argue that your results are "good enough" well okay. That's your decision. Just know that sometimes your outputs will be pretty far from reality.

Your example is super easy to solve. You do 2 solves. 1 vs RFI only 1 vs RFI/limp mix.

What you can't do is use a single solve against both strategies.

Last edited by JustLuck; 06-09-2021 at 04:23 PM.
JustLuck is offline   Reply With Quote
Old 06-09-2021, 04:28 PM   #68
aner0
old hand
 
aner0's Avatar
 
Join Date: Oct 2013
Location: https://discord.gg/TVbwaBHmhx
Posts: 1,559
Re: How does GTO win?

Quote:
Originally Posted by JustLuck View Post
No one is arguing the opposite. This isn't a competition where your choices are 10 sizings or just one. Both are really bad choices. The 10 sizing solution will give very poor results just like the 1 sizing solution.


if you want to argue that your results are "good enough" well okay. That's your decision. Just know that sometimes your outputs will be pretty far from reality.

Your example is super easy to solve. You do 2 solves. 1 vs RFI only 1 vs RFI/limp mix.

What you can't do is use a single solve against both strategies.
yes, maximally exploiting is best.

which one of the 2 is most exploitable (by far)
aner0 is offline   Reply With Quote
Old 06-09-2021, 04:42 PM   #69
JustLuck
veteran
 
Join Date: Jul 2007
Location: germanys first female supernova
Posts: 2,107
Re: How does GTO win?

Quote:
Originally Posted by aner0 View Post
yes, maximally exploiting is best.

which one of the 2 is most exploitable (by far)
They are both perfectly valid strategies right? Not limping preflop doesn't sacrifice much EV unless it's a very low or no rake structure?
Still if you encounter someone who doesn't limp you better fire up a preflop sim yourself and work out how to counter no limping. And vice versa.
"including a bet size" in a sim means nothing. You need to actually input their betting strategy if you want to get reliable results.


You can't just treat the raise range (of a raise/limp mix) the same as RFI only. Well you can but that would cost you a ton of EV. Or maybe that's close enough for you too?
JustLuck is offline   Reply With Quote
Old 06-09-2021, 04:45 PM   #70
aner0
old hand
 
aner0's Avatar
 
Join Date: Oct 2013
Location: https://discord.gg/TVbwaBHmhx
Posts: 1,559
Re: How does GTO win?

Quote:
Originally Posted by JustLuck View Post
They are both perfectly valid strategies right? Not limping preflop doesn't sacrifice much EV unless it's a very low or no rake structure?
Still if you encounter someone who doesn't limp you better fire up a preflop sim yourself and work out how to counter no limping. And vice versa.
"including a bet size" in a sim means nothing. You need to actually input their betting strategy if you want to get reliable results.


You can't just treat it as the same. Well you can but that would cost you a ton of EV
I meant the BB defense strategies when I asked which one is more exploitable
aner0 is offline   Reply With Quote
Old 06-09-2021, 04:59 PM   #71
JustLuck
veteran
 
Join Date: Jul 2007
Location: germanys first female supernova
Posts: 2,107
Re: How does GTO win?

Quote:
Originally Posted by aner0 View Post
I meant the BB defense strategies when I asked which one is more exploitable
You don't get it. This isn't a competition about which one is worse/better. Both strategies are "reasonable". Simply covering a sizing will give you a wrong output.

If all you want to do is argue that the wrong output is still sort of okay. I have no interest in discussing this further. You are knowingly accepting a wrong output as correct.

If you want a good answer to your BVB question you need to run multiple preflop sims. If you want a good answer against your opponents (very bad!!!) strategy of b25% only. You will still need to run a sim for this specific scenario. No shortcuts. No looking at a sim that has 25% as an option.
These two things are entirely different. And you agree.
"oh it's close enough" It's not.. it's really not.

That's how losing players run sims. Run your sims properly and stop with that preflop multi sizing nonsense. Why do you think no one at zenith plays HS?

Why does no one at high stakes mix between 3 sizings pre? I can assure you I have more than enough resources to run preflop sims with as many sizins as i want. I choose 1 per sim. But i run many different sizings pre. ONE AT A TIME
JustLuck is offline   Reply With Quote
Old 06-09-2021, 05:04 PM   #72
robert_utk
Pooh-Bah
 
robert_utk's Avatar
 
Join Date: Jan 2005
Location: ValueTown
Posts: 4,401
Re: How does GTO win?

If a strategy, A, which limps at some frequency is strictly +ev versus a strategy, B, that only open raises, then the BB that defends correctly against strategy A will will also be even more +ev against strategy B without ever needing to know if the BTN is playing strategy A or B.

If betting strategy A is a Nash solution, then the matching calling strategy is also a Nash solution. Either player playing B for whatever reason would be losing EV.

I think this is correct, unless one of the very experienced posters from earlier in the thread corrects me.

What is confusing here is that every finite set of bet sizes has a finite set of ranges, and these ranges might be different even when both players agree that the optimal bet size in a given spot so happens to be one of the sizes that the player with fewer sizes in strategy has solved for previously.

The final strategy should have as many sizes as possible given the chips in play, and there should be optimal ranges for all those sizes, in my opinion. As Aner0 has pointed out multiple times, the strategy with more solved sizes will more closely match the final theoretical solution.
robert_utk is offline   Reply With Quote
Old 06-09-2021, 05:20 PM   #73
JustLuck
veteran
 
Join Date: Jul 2007
Location: germanys first female supernova
Posts: 2,107
Re: How does GTO win?

Quote:
Originally Posted by robert_utk View Post
If a strategy, A, which limps at some frequency is strictly +ev versus a strategy, B, that only open raises, then the BB that defends correctly against strategy A will will also be even more +ev against strategy B without ever needing to know if the BTN is playing strategy A or B.

If betting strategy A is a Nash solution, then the matching calling strategy is also a Nash solution. Either player playing B for whatever reason would be losing EV.

I think this is correct, unless one of the very experienced posters from earlier in the thread corrects me.

What is confusing here is that every finite set of bet sizes has a finite set of ranges, and these ranges might be different even when both players agree that the optimal bet size in a given spot so happens to be one of the sizes that the player with fewer sizes in strategy has solved for previously.

The final strategy should have as many sizes as possible given the chips in play, and there should be optimal ranges for all those sizes, in my opinion. As Aner0 has pointed out multiple times, the strategy with more solved sizes will more closely match the final theoretical solution.
This doesn't have to be correct. EVs run insanely close between RFI only and limp/raise mix. In the BVB example it's very possible that the BB is making -EV plays because he incorrectly assumes villain plays RFI only and vice versa.

This can easily be verified by node locking a preflop sim and watching the EV of the BB drop off a cliff if he's forced to respond incorrectly. I posted an example of a postflop sim above The same concept applies here. We take a response that was calculated for a different problem and apply it to an entirely different spot. Somehow the response isn't good anymore??? SHOCKING

As for more sizings being better. In a theoretical sense no one is even arguing that. However that doesn't mean that you can simply add more sizings to your sim to get an answer for any sizing strategy your opponent chooses.

I'd recommend 2-4 sizings depending on the street and stacksizes. Anything more is just madness.
JustLuck is offline   Reply With Quote
Old 06-09-2021, 05:23 PM   #74
plexiq
veteran
 
Join Date: Apr 2007
Location: Vienna
Posts: 2,364
Re: How does GTO win?

Quote:
Originally Posted by robert_utk View Post
If a strategy, A, which limps at some frequency is strictly +ev versus a strategy, B, that only open raises, then the BB that defends correctly against strategy A will will also be even more +ev against strategy B without ever needing to know if the BTN is playing strategy A or B.

If betting strategy A is a Nash solution, then the matching calling strategy is also a Nash solution. Either player playing B for whatever reason would be losing EV.

I think this is correct, unless one of the very experienced posters from earlier in the thread corrects me.
The phrasing is a bit imprecise, but you are essentially correct.

A NE strategy for an abstract game that allows a set of possible sizings can't be exploited by a strategy that uses a subset of these sizings. This follows trivially from the definition of a NE, idk how this is even an argument.

Last edited by plexiq; 06-09-2021 at 05:33 PM.
plexiq is offline   Reply With Quote
Old 06-09-2021, 05:28 PM   #75
aner0
old hand
 
aner0's Avatar
 
Join Date: Oct 2013
Location: https://discord.gg/TVbwaBHmhx
Posts: 1,559
Re: How does GTO win?

Quote:
Originally Posted by JustLuck View Post
Both strategies are "reasonable"
I could show you a trivial exploit against the no limps BB defense that has some combos exploiting BB for >100bb/100. You will not find this the other way around.
aner0 is offline   Reply With Quote

Reply
      

Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off


Forum Jump


All times are GMT -4. The time now is 11:19 AM.


Powered by vBulletin®
Copyright ©2000 - 2021, Jelsoft Enterprises Ltd.
Copyright 2008-2020, Two Plus Two Interactive