Two Plus Two Publishing LLC
Two Plus Two Publishing LLC
 

Go Back   Two Plus Two Poker Forums > >

Notices

Mid-High Stakes Limit Discussions of mid-high stakes limit Texas Hold'em

Reply
 
Thread Tools Display Modes
Old 12-27-2017, 02:20 PM   #51
Clumsy Surgeon
enthusiast
 
Clumsy Surgeon's Avatar
 
Join Date: Jun 2017
Posts: 54
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by DeathDonkey View Post
I hear you and I donít think you are wrong. On some level everyone who posts on this site is ok with sharing info and making the games tougher. You are somewhat of an exception tbh.

I guess I look at it like new technology is sometimes invented and sometimes discovered. Figuring out Poker is more of a discovery of the right way to task a computer with the problem. And it was going to get solved regardless. At least academics publish their work so people like me can tinker with CFR algorithms on my laptop. Otherwise a few unscrupulous programmers would / will make millions running bots online and nobody would be the wiser.

Also itís not all peaches and cream being an academic these days. Their world is crumbling too
Why do you say this? For all academics? Certain fields?
Clumsy Surgeon is offline   Reply With Quote
Old 12-27-2017, 03:13 PM   #52
UpHillBothWays
adept
 
Join Date: Nov 2013
Posts: 976
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by Clumsy Surgeon View Post
Why do you say this? For all academics? Certain fields?
1. supply/demand of key positions
2. sexual allegations coming out in many areas now (though this isn't anti academics per se, just something that many departments are now dealign with given how crude and sexist and domineering generations of male professors have been over the past few decades)
3. upheaval in journal acceptance/reviews
4. increase in scrutiny of publications due to #3

to name a few
UpHillBothWays is offline   Reply With Quote
Old 12-27-2017, 06:37 PM   #53
up2ng
enthusiast
 
Join Date: May 2014
Posts: 54
Re: hand vs. cepheus. let's try to understand this one

I'll certainly defer to you guys on the specifics for how the algorithm for the bot is actually programmed -- I'd imagine that the reasons behind why it was done the way that it was has a lot to do with speed and efficiency vs current computing power.

I just hope that we can all get behind the fundamentals behind how betting ranges are determined when checked to heads up on the river in a gto solution. We determine what we can bet for value and bet those hands. We then add some bluffs for balance based on pot side. We check back all other hands. Agreed?
up2ng is offline   Reply With Quote
Old 12-27-2017, 07:06 PM   #54
Bob148
Carpal \'Tunnel
 
Bob148's Avatar
 
Join Date: May 2012
Location: takin it to the streets
Posts: 9,676
Re: hand vs. cepheus. let's try to understand this one

Quote:
I just hope that we can all get behind the fundamentals behind how betting ranges are determined when checked to heads up on the river in a gto solution.
I think it depends on the makeup of the ranges and how they interact. In the pure (bluffcatchers) vs (nuts + bluffs) scenario, yes you're right. However, having a range that is only (bluffcatchers) is clearly not part of a gto solution. Thus building ranges with this process is not going to give a gto solution for real poker:

Quote:
We determine what we can bet for value and bet those hands. We then add some bluffs for balance based on pot side. We check back all other hands. Agreed?
Instead, we need to account for the possibility of a check raise, as well as the size of the pot and the size of the bet before we can decide on bluffs.

Here's some river discussion free of charge:

https://forumserver.twoplustwo.com/1...22/?highlight=

Got a bit sidetracked at the beginning, but the meat of the discussion came to temperature here imo:

Quote:
For example:

pot = 1

player A: out of position players range = (0.4, 0.5, 0.6, 0.7)

player B: in position players range = (0, 0.65, 1)

player A checks, player B bets 1 pot with 100% of the listed range, player A?

ev of calling for player A:

(0.4) = 0ev call

(0.5) = 0ev call

(0.6) = 0ev call

(0.7) = 1 pot +ev call

-----

What if we give player B this range?: (0.4, 0.65, 1)

and keep player A's range the same: (0.4, 0.5, 0.6, 0.7)

player A checks, player B bets 1 pot with 100% of the listed range, player A?

here are the evs for the listed hands:

(0.4) = -0.5 pots -ev call

(0.5) = 0ev call

(0.6) = 0ev call

(0.7) = 1 pot + ev call

notice here that if player B checks with (0.4), he chops 25% of the time, which equals 12.5% equity, which equals a profitable checkback of 0.125 pots + ev. So if player A wants to make player B indifferent to checking or betting with (0.4), then player A should allow player B's bluffs to profit exactly 12.5% of the pot. Thus player A should call with 43.75% of his entire range, which allows such a profit for player B. This means that player A should check call at these frequencies:

(0.4) = 0% call

(0.5) = 0% call

(0.6) = 75% call

(0.7) = 100% call

With these frequencies, player A may not be exploited by player B.
Bob148 is offline   Reply With Quote
Old 12-27-2017, 07:07 PM   #55
thesilverbail
old hand
 
thesilverbail's Avatar
 
Join Date: Aug 2009
Location: on Mr Giblet's secret service
Posts: 1,649
Re: hand vs. cepheus. let's try to understand this one

no it's not so simple and one-directional. The "bluffing" and "value" ranges are determined together. Sometimes it is helpful in interpreting a solver solution, to think about it one way. e.g., if it’s pretty clear what the candidates for your value range should be, then you can figure out how much of the bottom of your range (modulo blockers) can be supported as a bluff. Polarized vs condensed range situations are the most common example of this, but this is not that common in limit holdem. But consider another scenario, where ranges overlap significantly. You might have some candidates for a value bet, and you try to figure out an appropriate number of bluffs, but then you realize you don’t have enough bluffs (i.e. some hands you are looking to bluff with actually have >EV by checking), this means villain can fold a wider range than you were thinking which means you might have to pare back your value bets, rinse and repeat.


So really they’re just 2 sides of a coin.
thesilverbail is offline   Reply With Quote
Old 12-28-2017, 01:37 AM   #56
up2ng
enthusiast
 
Join Date: May 2014
Posts: 54
Re: hand vs. cepheus. let's try to understand this one

Quote:
I think it depends on the makeup of the ranges and how they interact.
See, I don't think it depends on this at all. If we somehow happen to arrive at the river with perfectly symmetrical ranges the solution might look one way and if we have a massive range vs range advantage the solution might look another way, but the method for determining which hands we can bet for value remains the same as far as I can tell.

Quote:
In the pure (bluffcatchers) vs (nuts + bluffs) scenario, yes you're right. However, having a range that is only (bluffcatchers) is clearly not part of a gto solution.
I'm not sure what you're getting at here. In such a situation determining our value range becomes trivial and not worth discussing since those hands win at showdown 100% of the time. But I think that it's pretty clear that if we have hands in our range that will win at showdown 90% of the time we should be betting those for value also. Same with hands that will win 80% of the time, and so on. There is a threshold at which point betting the hand in question moves from a value bet into spew. Our task is to determine which hands can be bet for value. Including hands that are spew is incorrect. I hate to keep harping on this concept but I feel that it's extremely important and I am honestly shocked that we can't all agree on this small point.

As for our opponent having a range that is only (bluffcatchers), I never said that that has to be the case. However, when we are in position on the river and our opponent checks to us, that opponent should have MOSTLY folds and bluffcatchers -- some slightly stronger than others. The best bluff catchers might be able to beat some of our worst value bets, but mostly they will lose when we value bet. They will also have some checkraises -- some for value, and some as bluffs. These checkraises and the bluffcatchers together make up our opponent's "continuing range". When determining whether or not our hand should be included in our betting range as a value bet, we compare how it is doing against our opponent's continuing range (which, by definition, takes our opponent's checkraises into account). If it's strong enough we bet for value.

All of the excerpt that you cut-and-pasted regarding river play basically makes the same point as thesilverbail makes here:

Quote:
You might have some candidates for a value bet, and you try to figure out an appropriate number of bluffs, but then you realize you donít have enough bluffs (i.e. some hands you are looking to bluff with actually have >EV by checking), this means villain can fold a wider range than you were thinking which means you might have to pare back your value bets, rinse and repeat.
This is an interesting point and it's something that I have not thought about very much, but I don't see any flaws with this in theory.

I would say that if we are in a situation where we don't have enough bluffs on the river that might mean that we have played a previous street suboptimally. For example, we might semi-bluff certain types of hands on a previous street -- say, flush draws. In that example, if the flush card comes in on the river then we have no river bluffs. This is probably an indication that how we constructed our bluffing range on that previous street was unbalanced.

The other possibility where I could legitimately see this happening is on extremely narrow range situations -- in those spots there are just only so many total combos to go around. So then if the board runs out a certain way (somewhat rare) we just might not have any bluffs on the river.

In these cases, assuming both players actually played well up until the river, it's an interesting concept that we might actually have to value bet less often so that our opponent cannot just auto-fold his whole range. However, notice that we have still not yet come up with any legitimate situation where it is correct to ADD MORE value bets to our betting range. That's because the very next hand that we try to value bet is spew. It just is.

Determine our value bets by comparing how the hand is doing against our opponent's range. THEN add bluffs. Come on guys, work with me here . . .
up2ng is offline   Reply With Quote
Old 12-28-2017, 02:49 AM   #57
DeathDonkey
Carpal \'Tunnel
 
DeathDonkey's Avatar
 
Join Date: Feb 2004
Location: Scottsdale
Posts: 12,375
Re: hand vs. cepheus. let's try to understand this one

Sorta random but relevant:
https://youtu.be/tRiaGahlyy4
DeathDonkey is offline   Reply With Quote
Old 12-28-2017, 03:17 AM   #58
thesilverbail
old hand
 
thesilverbail's Avatar
 
Join Date: Aug 2009
Location: on Mr Giblet's secret service
Posts: 1,649
Re: hand vs. cepheus. let's try to understand this one

The extreme situation with no bluffs rarely happens in limit holdem, but i've seen solver solutions where weird things happen. Like you suggested, usually they show up in extremely assymetric range situations e.g. EP vs BB defend on a weird runout. If you tried to follow the simplistic approach of fixing a value range and then choosing (1-alpha) worth of bluffs, you would go badly wrong. The main reason would be because the (1-alpha) model doesn't apply when your worst bluff have non-zero s.d. value. (the formula for alpha is derived assuming pure air or nuts). A common example, you open as EP, BB defends, and you barrel to the river with a busted flush draw. Your worst FD (say JTs) has non-zero sd-value because BB should definitely only be x-calling a lot of his worse FDs since your range is so strong. I have seen a few completely reasonable scenarios where the defender ends up folding maybe 40% of his range in a 5 bet pot.

>However, notice that we have still not yet come up with any legitimate situation where it >is correct to ADD MORE value bets to our betting range. That's because the very next >hand that we try to value bet is spew. It just is.

In NL there are often situations where it is correct for one player to bet his entire range (typically an overbet) and the other player folds his entire range. In that case, there are no real "value" bets since they're not getting any value, but you need to make them to "protect" your bluffs.
thesilverbail is offline   Reply With Quote
Old 12-28-2017, 03:22 AM   #59
thesilverbail
old hand
 
thesilverbail's Avatar
 
Join Date: Aug 2009
Location: on Mr Giblet's secret service
Posts: 1,649
Re: hand vs. cepheus. let's try to understand this one

In fact honestly, and this falls in the dangerous category of maybe giving away actual useful information, after studying these solutions, I've realized that a far too common mistake made by people who know a little game theory (myself included), is to devote too much mind space in a hand to what their own range is and where they are in it and not enough to what their opponent's is and what it should be.
thesilverbail is offline   Reply With Quote
Old 12-28-2017, 04:11 PM   #60
UpHillBothWays
adept
 
Join Date: Nov 2013
Posts: 976
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by DeathDonkey View Post
Sorta random but relevant:
https://youtu.be/tRiaGahlyy4
Any more?
UpHillBothWays is offline   Reply With Quote
Old 12-28-2017, 06:01 PM   #61
JLot
enthusiast
 
JLot's Avatar
 
Join Date: Feb 2015
Posts: 59
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by DeathDonkey View Post
Sorta random but relevant:
https://youtu.be/tRiaGahlyy4
You are the man. Everyone should watch this - thank you.

PS. We miss you deeply in slippery chip land UHBW.
JLot is offline   Reply With Quote
Old 12-28-2017, 07:26 PM   #62
JLot
enthusiast
 
JLot's Avatar
 
Join Date: Feb 2015
Posts: 59
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by UpHillBothWays View Post
Cepheus bets
uhbw calls (20)
uhbw shows pair of eights with ace, nine, and five
Cepheus shows pair of eights with ace, nine, and seven [Cepheus won w/ A7o]

what do you guys think here? agree it's a clear value bet? agree it's not following FTOP? other thoughts?
Won't you take me to... dun dun dun... Value Town...
JLot is offline   Reply With Quote
Old 12-28-2017, 07:33 PM   #63
DonJuan
old hand
 
DonJuan's Avatar
 
Join Date: Jul 2006
Location: Lonely on Top
Posts: 1,274
Re: hand vs. cepheus. let's try to understand this one

Why you guys are so late everyone in plo is already using solver w 500+ Ram server in some other country sim out solution 24/7. and you guys are here talking about an old technology
DonJuan is offline   Reply With Quote
Old 12-28-2017, 11:20 PM   #64
UpHillBothWays
adept
 
Join Date: Nov 2013
Posts: 976
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by JLot View Post
Won't you take me to... dun dun dun... Value Town...
LOL. Omg if you pulled that on me after a bet like this Iíd die laughing.
UpHillBothWays is offline   Reply With Quote
Old 12-28-2017, 11:22 PM   #65
UpHillBothWays
adept
 
Join Date: Nov 2013
Posts: 976
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by JLot View Post
You are the man. Everyone should watch this - thank you.

PS. We miss you deeply in slippery chip land UHBW.
Agreed re the video.

Thanks Jlot. You are missed as well. Slippery chips, meh, unless weíre talking towers ;-)
UpHillBothWays is offline   Reply With Quote
Old 12-29-2017, 03:15 AM   #66
DeathDonkey
Carpal \'Tunnel
 
DeathDonkey's Avatar
 
Join Date: Feb 2004
Location: Scottsdale
Posts: 12,375
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by DonJuan View Post
Why you guys are so late everyone in plo is already using solver w 500+ Ram server in some other country sim out solution 24/7. and you guys are here talking about an old technology


Meh I mean people pay big $ and use the solvers but not many have any idea how they really work
DeathDonkey is offline   Reply With Quote
Old 01-13-2018, 09:34 AM   #67
maka2184
adept
 
Join Date: Oct 2006
Posts: 748
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by UpHillBothWays View Post
ok so the recent cepheus thread got me interested in playing it again. kinda rambly post as it's my thoughts, but i was interested so i thought others might be.

here's a hand:

Hand #31 of 100 dealt by Cepheus, Cepheus pays SB (5), uhbw pays BB (10)
Cepheus calls and raises (#1 of 3)
uhbw calls (10) with A4o
8♦8♠9♣ are revealed.
uhbw checks (i'm doing black/red k/r vs. call w/ Ahighs on favorable flops so this time was k/c. red-> k/r)
Cepheus bets
uhbw calls (10) (plan is to simply call down)
3♣ is revealed.
uhbw checks
Cepheus bets
uhbw calls (20)
5♣ is revealed.
uhbw checks
Cepheus bets
uhbw calls (20)
uhbw shows pair of eights with ace, nine, and five
Cepheus shows pair of eights with ace, nine, and seven [Cepheus won w/ A7o]

what i'm most interested in is cepheus's river bet. and more interestingly, how it approaches the river?

is cepheus bluffing? or is it value betting? i'm like virtually certain this is a value bet due to the insane # of bluffs it can have.

buuut, some of those value bets will check the turn. cepheus has a very balanced range and puts some bet/bet/bet hands in either the check/bet/bet bucket or the bet/check/bet bucket (though admittedly this one is small). so not ALL his value bets are ALL there on the river.

as a result, he has to include more value bets than i guess we'd think since it has to make up for the fact that there's a loootttt of bluffs that fall into the 11% bluffing frequency it needs to hit on the river.

i'll be honest, i was surprised to lose this pot. it seemed most likely he had any number of busted draws rather than value and i assumed he'd check any better ace behind vs. bet for value. ofc, i shouldn't be thinking like that and instead thinking about what my range of hands is and how far up A4o is in it. offhand it seems like it's way too far up to fold even though he's only bluffing 11% of the time.

last thought on this part-> aren't there SOOO many bluffs he can have here (oesd, gutshots +overcards, etc.) that he'd have to be checking back the river a lot? of the 89%, he has to be checking like 50% of those, right? does this mean if i play my draws mostly aggressively that i can fold even a hand this high up in my call/call/call range? EDIT: no i can't ever fold this. even if my c/c/c range is made stronger and smaller by playing draws generally aggressively, he's still bluffing lots. he'd have to check like 2:1 check/value bet the river for me to start considering folding hands like this imo

secondarily, cepheus doesn't follow the fundamental theorem of poker since WHEN CALLED, cepheus loses more than 50% for sure. instead, it's developing ranges solely based on pot size and the decision trees that eliminate the possibility of being exploited vs. getting the most money out of each bet.

in order to follow the FTOP, it would have to reduce the value bet frequency, which would then also reduce the bluff frequency, which means it would be giving up on too many bluffs.

what do you guys think here? agree it's a clear value bet? agree it's not following FTOP? other thoughts?
After seeing results this may sound result oriented but A7o by Cepheus/villain is for sure a value bet HUHU.

TAG or Fish, too many players chk call down too light on these boards, including as low as Qhi+ based on meta and play history.

Cepheus/villain should have a lot of bluffs though as you note but so should most competent HUHU players I think in Cepheus's shoe as aggressor
maka2184 is offline   Reply With Quote
Old 01-13-2018, 09:42 AM   #68
maka2184
adept
 
Join Date: Oct 2006
Posts: 748
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by DeathDonkey View Post
Sorta random but relevant:
https://youtu.be/tRiaGahlyy4
Thanks for this.
maka2184 is offline   Reply With Quote
Old 01-29-2018, 02:08 AM   #69
ninefingershuffle
veteran
 
Join Date: Aug 2009
Posts: 2,008
Re: hand vs. cepheus. let's try to understand this one

What I don't seem to get from cepheus is how light it calls down.

It opens, I 3 bet and barrel AA9JJr board and it calls down with T7.

It opens and I three bet and it calls down with Q5s on a AKJJ4hhh board.
ninefingershuffle is offline   Reply With Quote
Old 01-29-2018, 01:33 PM   #70
UpHillBothWays
adept
 
Join Date: Nov 2013
Posts: 976
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by ninefingershuffle View Post
What I don't seem to get from cepheus is how light it calls down.

It opens, I 3 bet and barrel AA9JJr board and it calls down with T7.
i bet if you changed an ace to another low blank and the jack to another low blank it'd fold. i think the removal here factored into the call.

Quote:
It opens and I three bet and it calls down with Q5s on a AKJJ4hhh board.
i'm guessing another card removal decision. it seems to base a bunch of its decisions on card removal based on my review of various hands.
UpHillBothWays is offline   Reply With Quote
Old 02-02-2018, 08:47 PM   #71
jdr0317
Pooh-Bah
 
Join Date: Jan 2012
Location: Boston, MA
Posts: 5,957
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by UpHillBothWays View Post
i bet if you changed an ace to another low blank and the jack to another low blank it'd fold. i think the removal here factored into the call.



i'm guessing another card removal decision. it seems to base a bunch of its decisions on card removal based on my review of various hands.
Ya basically. Like if he doesn't hold a card critical to your barrels, he is way more likely to call
jdr0317 is offline   Reply With Quote
Old 02-03-2018, 09:37 AM   #72
maka2184
adept
 
Join Date: Oct 2006
Posts: 748
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by UpHillBothWays View Post
i bet if you changed an ace to another low blank and the jack to another low blank it'd fold. i think the removal here factored into the call.



i'm guessing another card removal decision. it seems to base a bunch of its decisions on card removal based on my review of various hands.
With assumption that Cepheus is GTO and plays correctly mathematically/statistically, do you think it is practical/beneficial to take card removal to the extent that Cepheus does?

Specifically, as a HERO playing against a solid human GTO LAGTAG HUHU, should we be doing the EV analysis off the table often enough to make a decision like Cepheus does in real time a default play (NOTE: Default as in X% frequency if board and hand requires)?
maka2184 is offline   Reply With Quote
Old 02-03-2018, 05:12 PM   #73
ninefingershuffle
veteran
 
Join Date: Aug 2009
Posts: 2,008
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by jdr0317 View Post
Ya basically. Like if he doesn't hold a card critical to your barrels, he is way more likely to call
I won 1.5 of these hands.
ninefingershuffle is offline   Reply With Quote
Old 02-04-2018, 10:08 AM   #74
Bob148
Carpal \'Tunnel
 
Bob148's Avatar
 
Join Date: May 2012
Location: takin it to the streets
Posts: 9,676
Re: hand vs. cepheus. let's try to understand this one

Quote:
Originally Posted by maka2184 View Post
With assumption that Cepheus is GTO and plays correctly mathematically/statistically, do you think it is practical/beneficial to take card removal to the extent that Cepheus does?

Specifically, as a HERO playing against a solid human GTO LAGTAG HUHU, should we be doing the EV analysis off the table often enough to make a decision like Cepheus does in real time a default play (NOTE: Default as in X% frequency if board and hand requires)?
yes otherwise you could play a pure strategy and exploit me for not playing a properly mixed strategy.
Bob148 is offline   Reply With Quote
Old 02-04-2018, 10:46 AM   #75
Bob148
Carpal \'Tunnel
 
Bob148's Avatar
 
Join Date: May 2012
Location: takin it to the streets
Posts: 9,676
Re: hand vs. cepheus. let's try to understand this one

More specifically:

If I don't randomize my 0ev river calls and folds, and if I only call exactly the top portion of my range which makes my opponent indifferent to bluffing or checking the river, then you could bluff with any decent blocker in addition to the neutral ev bluffs from the bottom of your range, for much more profit than those hands would receive from the pot vs cepheus.
Bob148 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 03:35 PM.


Powered by vBulletin®
Copyright ©2000 - 2018, Jelsoft Enterprises Ltd.
Copyright © 2008-2017, Two Plus Two Interactive
 
 
Poker Players - Streaming Live Online