Daniel Negreanu

But it tells you what you would do with any hand, right?

It gives you the full output.

And the solvers do things that are really interesting.

You've seen AlphaGo, I would imagine.

Brilliant film, right?

I thought.

And I thought what was interesting is there was, you know, accepted theory from all the top Go players that this is what you do.

But...

The AI was doing things way different and they're like, this has to be wrong, but it really, it wasn't.

So for example, a solver may say this, right?

You, let's say you bet on the end and you bet a lot.

And a solver may say, you should fold here with a pair of Kings and a queen kicker, which is, you know, a pair of Kings, but call with a pair of fours and an ace kicker.

So it's essentially telling you that you should fold this hand that is much better than this.

So it begs the question, why?

Because what the solvers do is they use the information of your own cards to formulate all the possible hands your opponent can have.

So if your opponent is...

So basically, if you had the king-queen, you know, it may say, for lack of a better nerdy term, it blocks potential bluffing hands that your opponent can have.

So let's say if your opponent would bluff with queen-jack, but you have a queen, so there are less combinations of queen-jack.

So it will find a better bluff catcher, if you will.

So that's what's really not intuitive to poker players.