I would be interested to know how far ahead the AI looks.
It suggested today that a fold I had made to a check/raise was in fact a call. My knowledge of the player would lead me to believe that HGP got this wrong but that is not the point I am making and I have no problem with it. I expect to be able to make better reads than an AI program. I could be wrong
Let's say that HGP's assumptions about his check/raise range was true. Does HGP now do hot+cold simulations assuming no further betting and calculate that my equity is +$x and so I should call?
If that is true, it is not looking forward far enough.
If my opponent is fairly competent, then he will bet the turn in a manner that is close to GTO. That bet makes me indifferent to calling or folding so let's say we call.
On the river he does the same and as I am indifferent and curious I call again.
On this basis we have done bothing wrong. We took each decision as it came; the 1st one being slightly +EV for the FLOP and the next 2 streets being neutral.
However, when he makes his GTO bet on the turn, even though he is making me indifferent, he is actually stealing some of my EV with this bet. When he fires the river he does the same again.
If my EV on the flop is "close" and he will "steal" some on 2 more streets, the flop call becomes wrong.
I may have totally misundertood GT or I do not see what HGP is doing but I'd sure like to know
Trevor