A Whole New Series: GTOW AI
Posted by Nuno Alvarez
Video Player is loading.
Resume Video
Start from Beginning
Watch Video
10
Train with the best
Learn from the Run It Once Pro Team with two new instructional videos every day.
Loading 6 Comments...
At 18.00 you were looking at the probe sizing options on KT4ss5.
I did some work on probing in 6 max BB vs BTN. I know it's a different spot but I'm sure there are some similarities as it's also a wide range vs wide range scenario. In this spot, BB probes with only polar sizings when there are 2 "big" cards on the board - in this case a card 9 or above (given that BTN opens 9X with broadway kicker pure as it's weakest pure opens in 6-max). Adding a block size to the probing strategy doesn't add EV in these scenarios as BB doesn't bet for protection when BTN can have some many hands that are 2nd pair+. Not sure whether it will be exactly the same in HU, but it's definitely something to look out for, and quite easy to test now that we have GTOW AI to help!
Hey mat thanks a lot for the insights!
I tested this and doesn't seem to show as clearly in HU in general. I think there are too many nuances to grasp given how wide the ranges are.
In this particular board the 25% sizing almost doesn't get used but if we go to K945, KJ45 or even KQ45 it seems to be needed.
If you leave the block sizing in it will usually get used, and often with a decent frequency of the betting volume which makes it look like an important part of the probing strategy. But if you compare the EV of the overbet/block mixed strategy to the overbet only strategy, you might find that they are in fact the same/very close, and that the block sizing doesn't add any/much EV but increases the complexity of the strategy a lot.
The difference in EV is pretty small, and comes out at about 0.7BB/100 for the added complexity of adding the block sizing (assuming you are able to execute the mix).
34.00
Looks 8x 2-pairs are more or less the only 2 pairs going for the river check raise. I think that blocking IP's 88 which rivered a set is very important too here, especially when using a 6x raise size. It also blocks IP's rivered 2-pair hands that might have you beat, if you have a lower 2 pair like 82s for example. Having that 8 blocker allows OOP to get thinner value it seems?
And for the bluffs being pulled mainly from Q8 and J8, is there also an element of IP calling the raise with bluff-catchers that block the straight, e.g. KJ, or does that not make much of a difference?
Good point. I think the main heuristic to chekc raise here is still to block the check back range for IP with our 2p hands but blocking the rivered set is also a nice extra.
Correct. We pull the bluffs from Q8 and J8 not because we want to block their QJ hands but because we want to block their better bluffcatchers against our value range that contains QJ at frequency.
Be the first to add a comment
You must upgrade your account to leave a comment.