EV almost the same? GTO WIZARD
GTO wizard say a move was -91 GTOW score but the EV was pretty marginally worse for that move.
What even is GTOW score and why does it not follow how big the EV difference is in different moves?
I'm doing spin and go exercises.
9 Replies
How do you define "marginal"?
Is a 10bb/100 mistake marginal?
1bb/100?
0.1bb/100?
Where is the line?
It's difficult to objectively define it.
0.17 difference in this spot and -91 gtow score.
I mainly don't understand why the score can differ widely and seemingly not much related to the EV difference, sometimes its equal or less than a 0.1 difference but big score and sometimes the positive or negative score is not greater than + - 91 but the EV difference is a lot bigger.
0.17 EV difference is 17bb/100, which most people would consider a significant mistake.
GTOW score is normalized to a 0–100 scale and decreases nonlinearly based on how much EV your decision lost compared to the other available options.
I see I didn't raise it meant that much.
I do find the gtow interesting sometimes but sometimes it says to do something like call more often than raise or go all in oop despite EV being the same or like 0.02 difference preflop but could be costly if not played perfect postflop and it doesn't take in to account that many might fold a bit more to a shove making it more plus EV in many situations.
Wish there where somewhere with both solvers and more explanations and nuanced thoughts on spots.
GTO Wizard has a youtube channel where you can learn poker theory for free. It gives you insights from different coaches.
I see I didn't raise it meant that much. I do find the gtow interesting sometimes but sometimes it says to do something like call more often than raise or go all in oop despite EV being the same or like 0.02 difference preflop but could be costly if not played perfect postflop and it doesn't take
What you're talking about is exploitative play. A good coach can help you identify these spots. Using your own reasoning is also necessary as a poker player. Like in your example, GTO wizard might tell you that raising to a smaller sizing is slightly higher EV than jamming in a given spot. However it requires you to perfectly balance a complex range in various postflop spots. You might decide that the slight amount of EV you give up by jamming is more than made up for by the reduction in postflop errors.
I just played a final table the other day with an amateur who was basically always jamming or folding. He was certainly giving up some theoretical EV that way, but it was actually a great strategy given that he was playing against several professionals who would make better postflop decisions. His strategy effectively minimized our edge and increased the variance which is exactly what you want to do when you're the less experienced player.
The GTOW score is basically a normalized penalty. At 20bb, a 0.17bb EV mistake is actually a bigger deal than it sounds because your entire stack is only 20bb. The score scales the error relative to what was available to win, so a small raw EV gap can still be a large score penalty in shallow-stack spots.On the solver preferring call over shove when EV is basically equal: at 20bb it's often accounting for how your range plays postflop across the full tree, not just the immediate spot. Two moves can have near-identical EV in isolation but one leaves you with a cleaner range to continue with.
