Amazing!!! Though a lot of what is being said goes over my head this is the type of a video I like the most. Not saying I want you to stop with the low stakes zoom videos though :-))
Ben is there a way to explain what about OBORRA/40and7/etc's play makes so many people think they're using some sort of real time assistance that isn't too revealing/speculative? Why do they get all the speculation and not Otb or LLinus or whoever?
Because they are from east europe and everyone are racists...
On a more serious note I'm pretty optimistic that otb crushes competition harder than a gto bot ever could so clearly he's very talented or no software in the world could make him this good I think (Without him having a very deep understanding of the game)
Llinus is a bit too crazy to be using software or anything. He has a big heart and is very talented/smart, that's all.
Ben in your first PIO model where you mentioned a bunch of his turn bets were converging out I believe that was the case and they show up at almost no frequency because when allowing him to bet a small flop sizing I believe if you look back PIO will have him betting most those hands on flop near 100% which won't allow him to get to the turn with a check flop and delay cbet turn line.
I'm wondering why you prefer CREV which feels like way more time consuming to set up and easier to make mistakes when setting up a model then using PIO with node locking function to find EVs based on strategies that sway from equilibrium?
If you nodelock in Pio it will find the equilibrium given the constraints you put on it. The result is that Pio will play around the imbalances you lock into its strategy to the best of its ability. In this case, both players were playing moderately far from equilibrium strategies on multiple nodes, so CREV seemed like a better tool.
You can lock multiple nodes on PIO up to the points you want to evaluate though and then get a full range construction vs those strategies. So in your example you could lock the assumed strategy he was playing preflop and on flop before you made your turn decision of raising A8o to see what combos you can call/raise/fold. I would think PIO would do an excellent job of giving you solid range construction given your assumptions and then river would just come down to assumptions on how he plays river and if you have correct price to call as you don't think he is anywhere near equilibrium at this point. Or even see what his equilibrium looks like for bluffs based on his assumed strategy up until that point of the hand where he shoves river.
That would be fine too I think. I prefer CREV for certain situations like this because I want to make many assumptions and look at the EV against a human strategy, not a computer strategy. The CREV tree also lets me work through my assumptions and evaluate them in a way I'm familiar with from years of practice.
Thanks for your comment though I look forward to playing with locking multiple nodes in Pio trees in the future.
Hey Ben, what an awesome video! Usually the examples we have of exploitative play are more like "I think this sizing means he has X", but it usually seems like there is too much guessing/assumptions going on, and not enough "science" to those types of reads, which makes peolpe wrong more often than they should, I think, to make such assumptions and big deviations... and that has made me kinda struggle when relating theory with exploitative play, but this was a great example of how things can correlate, and exploitative play can actually be quite "scientific" and not have to depend on "feeling" or "magic" as some like to call it... Thanks a lot!
one of my fav videos. made a lot of sense thx.
so about the turn raise, u wanna make it bigger if he jam polarized on flush rivers to make it more expensive for him to put u in a near zero EV spot on 50%ish of rivers?
Loading 17 Comments...
Amazing!!! Though a lot of what is being said goes over my head this is the type of a video I like the most. Not saying I want you to stop with the low stakes zoom videos though :-))
Maybe a dumb question, but what exactly is a tempo bet? Or rather why is a small blockish type bet called a tempo bet?
I call it a tempo bet because it forces oop to xr or ip gets 2 cards
Cliffs: I put him on clubs
Ben is there a way to explain what about OBORRA/40and7/etc's play makes so many people think they're using some sort of real time assistance that isn't too revealing/speculative? Why do they get all the speculation and not Otb or LLinus or whoever?
Because they are from east europe and everyone are racists...
On a more serious note I'm pretty optimistic that otb crushes competition harder than a gto bot ever could so clearly he's very talented or no software in the world could make him this good I think (Without him having a very deep understanding of the game)
Llinus is a bit too crazy to be using software or anything. He has a big heart and is very talented/smart, that's all.
They say OBORRA is better than otb tho..
Dont forget to pay your canadian bills
Ben in your first PIO model where you mentioned a bunch of his turn bets were converging out I believe that was the case and they show up at almost no frequency because when allowing him to bet a small flop sizing I believe if you look back PIO will have him betting most those hands on flop near 100% which won't allow him to get to the turn with a check flop and delay cbet turn line.
I'm wondering why you prefer CREV which feels like way more time consuming to set up and easier to make mistakes when setting up a model then using PIO with node locking function to find EVs based on strategies that sway from equilibrium?
If you nodelock in Pio it will find the equilibrium given the constraints you put on it. The result is that Pio will play around the imbalances you lock into its strategy to the best of its ability. In this case, both players were playing moderately far from equilibrium strategies on multiple nodes, so CREV seemed like a better tool.
You can lock multiple nodes on PIO up to the points you want to evaluate though and then get a full range construction vs those strategies. So in your example you could lock the assumed strategy he was playing preflop and on flop before you made your turn decision of raising A8o to see what combos you can call/raise/fold. I would think PIO would do an excellent job of giving you solid range construction given your assumptions and then river would just come down to assumptions on how he plays river and if you have correct price to call as you don't think he is anywhere near equilibrium at this point. Or even see what his equilibrium looks like for bluffs based on his assumed strategy up until that point of the hand where he shoves river.
That would be fine too I think. I prefer CREV for certain situations like this because I want to make many assumptions and look at the EV against a human strategy, not a computer strategy. The CREV tree also lets me work through my assumptions and evaluate them in a way I'm familiar with from years of practice.
Thanks for your comment though I look forward to playing with locking multiple nodes in Pio trees in the future.
Pretty sad you didn't go over PIO river play
Hey Sauce, whats your IQ? You are a mad genius.
Hey Ben, what an awesome video! Usually the examples we have of exploitative play are more like "I think this sizing means he has X", but it usually seems like there is too much guessing/assumptions going on, and not enough "science" to those types of reads, which makes peolpe wrong more often than they should, I think, to make such assumptions and big deviations... and that has made me kinda struggle when relating theory with exploitative play, but this was a great example of how things can correlate, and exploitative play can actually be quite "scientific" and not have to depend on "feeling" or "magic" as some like to call it... Thanks a lot!
'oborro explo fun times'
:D
one of my fav videos. made a lot of sense thx.
so about the turn raise, u wanna make it bigger if he jam polarized on flush rivers to make it more expensive for him to put u in a near zero EV spot on 50%ish of rivers?
Be the first to add a comment
You must upgrade your account to leave a comment.