View Plans & Pricing

If you are signed in and are seeing this message, please be sure you have selected a user name in My Profile. The forum requires it.
A A A
Search

— Forum Scope —




— Match —





— Forum Options —





Minimum search word length is 3 characters - maximum search word length is 84 characters

sp_Feed
Topic Rating: 0 Topic Rating: 0 Topic Rating: 0 Topic Rating: 0 Topic Rating: 0 Topic Rating: 0 (0 votes) 
sp_TopicIcon
What determines river bet sizing?
Avatar
Carlos
Atlanta, Ga
Member Moderator
Forum Posts: 778
Member Since:
October 21, 2012
sp_UserOfflineSmall Offline
1
May 23, 2017 - 3:23 pm
sp_Permalink sp_Print
0

What determines the size of a river bet when you are attempting to play GTO?

Here are my guesses.

Lots of value hands and few bluffs? Small bet

Lots of bluffs and few value hands? Big bet

Few value hands and few bluffs? Big bet (not sure about this one, but it’s what I just heard in a twitch video)

Lots of value hand and lots of bluffs? No idea

The Riceman
London UK
Hitting The Circuit
Members
Forum Posts: 731
Member Since:
February 5, 2015
sp_UserOfflineSmall Offline
2
May 23, 2017 - 4:53 pm
sp_Permalink sp_Print sp_EditHistory
0

Playing GTO?

Hey Carlos. But my feeling is you are off here. I have imbibed my evening whiskey, so beware the possibility I am skewed in my thinking. But I am fairly certain about this principle: if we are attempting to “play GTO” then we are trying to be balanced i.e. unreadable. 

This means that if we have lots of value hands and few bluffs, and  if we have lots of bluffs and few value hands, in order to remain balanced/ unreadable, we will need to be betting the same amount with both ranges.

I am sure Andrew or Matt can put it more eloquently!

(Nice to see you in the forums more by the way…pleased to make your acquaintance. I always enjoy listening to you on the podcasts. You get the Riceman seal of approval, fwiw!).

I started an extensive thread around GTO bet sizing in the MTT Strat. forum six months or so ago. It is called the “Riceman Unexploitable GTO Bet Sizing Dice TM” thread. Although based around a few misguided notions and fundamental misunderstandings around game theory on my part, the thread became fascinating by the end. Andrew and Matt contribute a lot.

Anyway, GTO bet sizing is discussed at length in that thread. If you haven’t already seen it, check it out!

kntz
Playing Freerolls
Members
Forum Posts: 12
Member Since:
April 28, 2017
sp_UserOfflineSmall Offline
3
May 23, 2017 - 5:42 pm
sp_Permalink sp_Print sp_EditHistory
0

Hey guys, I’ve just recently started taking studying poker more seriously, so I have a lot to learn, especially in this area of GTO play. From what I’ve gathered watching Andrew’s and some other educational videos, you want to size your bet to make the opponent indifferent about calling or folding. Meaning, if you have lots of value hands and only few bluffs you need to be betting a smaller amount in relation to the pot, that also correlates to the value/bluff ratio that you have in that given spot. From Andrew’s range construction video on AKQ game: if the hero bets 1,000 into a 3,000 pot, he wants to have 80% of value hands and 20% bluffs. This way, when the opponent calls 1,000 to win the pot of 4,000, he will lose a bet of 1,000 four times out of five (1,000*4=4,000) and will win 4,000 one time out of five (4,000*1=4,000). Meaning, he won’t be making any profit there and calling will have the same EV as folding.

Now, if the hero had the same amount of value and bluffs (80/20) and bet 500 into 3,000, then the opponent would have to call 500 to win 3,000. He would lose 500 four times out of five (500*4=2,000) and would win 3,000 one time out of five (3,000*1=3,000). Meaning he would have a profitable call (+1,000).

If the hero bet 3,000 into 3,000 and the opponent called, the opponent would lose 3,000 four times out of five (3,000*4=12,000), and win 3,000 one time out of five (3,000*1=3,000) making it a really losing call (-9,000).

The last one is an exploitation example, and if the opponents know this is what you’re doing, they can re-adjust and exploit you even more profitably. This is why if you play a GTO strategy, you want to have a lot of bluffs in your range when you bet BIG (33% of bluffs when you bet pot). And while this means that you’re not going to be exploiting anyone, they won’t be able to exploit you either.

Now, as for the ‘lots of bluffs and few value hands’, it sounds a bit strange to me, because I can’t think of a situation where you would have your range constructed in such a way where you have more bluffs than value bets. If that does happen, I assume it’s better to not bluff at all. Because if you think about it mathematically, you can never have a GTO strategy where you have more bluffs than you have value hands. For example: we know that we should bet big when we have a lot of bluffs. For argument’s sake, let’s say you bet 10,000 into a pot of 1,000. The opponent would need to call 10,000 to win 11,000. If you are playing a balanced strategy, you would have to have 47.6% of bluffs there. And the more you overbet the pot, you will still never reach 50% or more (for example, if you bet 100x the pot, you would need to have 49.75% of bluffs, and if you bet 1000x the pot, you would need to have 49.975% of bluffs and so on).

The Riceman
London UK
Hitting The Circuit
Members
Forum Posts: 731
Member Since:
February 5, 2015
sp_UserOfflineSmall Offline
4
May 24, 2017 - 1:30 am
sp_Permalink sp_Print sp_EditHistory
0

Gee whizz, “just started studying poker more seriously “…?

If our two answers were people, mine would be an unwashed dead beat dad with 2 bucks in his pocket, with a drink and gambling problem and a broken leg in the emergency ward, and yours would be David Hasslehoff in his prime running down the beach surrounded by hot chicks.

Of course, the principle of indifference is key. 

Really, it is a huge subject Carlos.

kntz
Playing Freerolls
Members
Forum Posts: 12
Member Since:
April 28, 2017
sp_UserOfflineSmall Offline
5
May 24, 2017 - 2:42 am
sp_Permalink sp_Print
0

The Riceman said
Gee whizz, “just started studying poker more seriously “…?

If our two answers were people, mine would be an unwashed dead beat dad with 2 bucks in his pocket, with a drink and gambling problem and a broken leg in the emergency ward, and yours would be David Hasslehoff in his prime running down the beach surrounded by hot chicks.

Of course, the principle of indifference is key. 

Really, it is a huge subject Carlos.  

Haha, that analogy made me smile. To be honest, I basically just regurgitated what I’ve learned there, by no means does it mean that I can do that stuff in practice. When I’m playing, especially if I’m playing more tables, it just becomes too overwhelming to keep track of all the hands and my ranges at that given moment.

The Riceman
London UK
Hitting The Circuit
Members
Forum Posts: 731
Member Since:
February 5, 2015
sp_UserOfflineSmall Offline
6
May 24, 2017 - 2:27 pm
sp_Permalink sp_Print sp_EditHistory
0

Nobody can do it. In multiway pots apparently GTO solutions may not even exist. You’ll be a genius if you can get most Nash preflop shove ranges right, let alone balancing bet sizes and ranges post flop vs multiple villains.

This is why poker won’t be solved and everyone should relax around GTO IMO.

I’ve been experimenting with HoldemResources and there is a cathedral of information within that “simple” pre flop calculator, much of it unintuitive. 

For sure it cannot hurt to attempt to unlock and study the solution to the game though.

Foucault

TPE Pro
Members
Forum Posts: 2067
Member Since:
December 6, 2012
sp_UserOfflineSmall Offline
7
May 25, 2017 - 8:25 am
sp_Permalink sp_Print
0

You’re on the right track talking about bluff:value ratio, but that applies only when the Villain is capped. In other words, if his range consists only of bluff-catchers, then you just want to bet as big as you can with all of your value hands and then as many bluffs as you can balance (which doesn’t usually mean every hand you could possibly bluff with).

Realistically, though, this isn’t often the case. Here’s an example from a different TPE thread: Hero raises AK from SB, BB calls. Flop K73hh. Bet, call. Turn 7. Hero bets (he didn’t but should have, in that thread), Villain calls. River 2. Villain’s river range probably consists primarily of Ax, Kx, 7x, 3x, busted draws, and perhaps some pocket pairs.

If Hero bets huge (say, 5x pot) with a polarized range, then he can’t include AK in that range. This is because, facing such a large bet, Villain doesn’t have to call very often to make Hero’s bluffs unprofitable. He can probably just call trips or better and fold everything else. So although this could be the most profitable way to play 33 and some bluffs, it’s not the most profitable way to play AK.

Hero needs to find a bet size such that Villain must call with many hands worse than AK or risk folding too often to bluffs. If Hero bets 1/2 pot, Villain needs to call with 2/3 of his hands that can beat bluffs. He probably won’t have trips or better nearly often enough, which means he’ll have to call with Kx and weaker to prevent Hero from bluffing profitably.

So it is about making V indifferent to calling, but you need to focus on hands weaker than yours (value targets, if you’ve seen my Getting Paid series), not just on “bluff-catchers” generically.

Avatar
Carlos
Atlanta, Ga
Member Moderator
Forum Posts: 778
Member Since:
October 21, 2012
sp_UserOfflineSmall Offline
8
May 26, 2017 - 3:35 am
sp_Permalink sp_Print sp_EditHistory
0

kntz said
Hey guys, I’ve just recently started taking studying poker more seriously, so I have a lot to learn, especially in this area of GTO play. From what I’ve gathered watching Andrew’s and some other educational videos, you want to size your bet to make the opponent indifferent about calling or folding. Meaning, if you have lots of value hands and only few bluffs you need to be betting a smaller amount in relation to the pot, that also correlates to the value/bluff ratio that you have in that given spot. From Andrew’s range construction video on AKQ game: if the hero bets 1,000 into a 3,000 pot, he wants to have 80% of value hands and 20% bluffs. This way, when the opponent calls 1,000 to win the pot of 4,000, he will lose a bet of 1,000 four times out of five (1,000*4=4,000) and will win 4,000 one time out of five (4,000*1=4,000). Meaning, he won’t be making any profit there and calling will have the same EV as folding.

Now, if the hero had the same amount of value and bluffs (80/20) and bet 500 into 3,000, then the opponent would have to call 500 to win 3,000. He would lose 500 four times out of five (500*4=2,000) and would win 3,000 one time out of five (3,000*1=3,000). Meaning he would have a profitable call (+1,000).

If the hero bet 3,000 into 3,000 and the opponent called, the opponent would lose 3,000 four times out of five (3,000*4=12,000), and win 3,000 one time out of five (3,000*1=3,000) making it a really losing call (-9,000).

The last one is an exploitation example, and if the opponents know this is what you’re doing, they can re-adjust and exploit you even more profitably. This is why if you play a GTO strategy, you want to have a lot of bluffs in your range when you bet BIG (33% of bluffs when you bet pot). And while this means that you’re not going to be exploiting anyone, they won’t be able to exploit you either.

Now, as for the ‘lots of bluffs and few value hands’, it sounds a bit strange to me, because I can’t think of a situation where you would have your range constructed in such a way where you have more bluffs than value bets. If that does happen, I assume it’s better to not bluff at all. Because if you think about it mathematically, you can never have a GTO strategy where you have more bluffs than you have value hands. For example: we know that we should bet big when we have a lot of bluffs. For argument’s sake, let’s say you bet 10,000 into a pot of 1,000. The opponent would need to call 10,000 to win 11,000. If you are playing a balanced strategy, you would have to have 47.6% of bluffs there. And the more you overbet the pot, you will still never reach 50% or more (for example, if you bet 100x the pot, you would need to have 49.75% of bluffs, and if you bet 1000x the pot, you would need to have 49.975% of bluffs and so on).  

This seems to be somewhat inline with what I was thinking, except I maybe erroneously tried to take the number of bluffs into account before choosing a size.

Maybe the bet size should be determined by the number of value hands and the number of bluffs follow afterward based on the ratio. So…

Lots of value hands = Small bet

Few value hands = Big bet

Then, either way, include a proportional number of bluffs to balance

I’ll need to go back and review some of my old materials on the subject to clear up the fog in my head.

Thanks all.

Foucault

TPE Pro
Members
Forum Posts: 2067
Member Since:
December 6, 2012
sp_UserOfflineSmall Offline
9
May 26, 2017 - 8:53 am
sp_Permalink sp_Print
0

Lots of value hands = Small bet

Few value hands = Big bet

Then, either way, include a proportional number of bluffs to balance”

No, it’s really the ratio of bluffs:value bets (in scenarios where your value bets beat all of Villain’s potential calls – see my post above). If you are betting a perfectly polarized range less than all in and also not betting all of your possible bluffs, then you’re doing it wrong. There are plenty of cases where you could have lots of value hands and lots of bluffs, and in that case betting big is correct.

Forum Timezone: America/New_York

Most Users Ever Online: 2780

Currently Online:
22 Guest(s)

Currently Browsing this Page:
1 Guest(s)

Top Posters:

bennymacca: 2616

Foucault: 2067

folding_aces_pre_yo: 1133

praetor: 1033

theginger45: 924

P-aire 146: 832

Turbulence: 768

The Riceman: 731

duggs: 591

florianm1: 588

Newest Members:

CSerpent

KJ

Tillery999

sdmathis89

ne0x00

adrianvaida2525

Forum Stats:

Groups: 4

Forums: 24

Topics: 12705

Posts: 75003

 

Member Stats:

Guest Posters: 1063

Members: 12010

Moderators: 2

Admins: 5

Administrators: RonFezBuddy, Killingbird, Tournament Poker Edge Staff, ttwist, Carlos

Moderators: sitelock, sitelock_1