Hi, I like the math pack very much and I've took some time to learn all the math in the series but have trouble putting into practice the math from the last episode. I would like to know if we can use this in-game or just for postgame analysis.
In the last episode we use custom filters in PT4 (6-8bb) to get an estimate of our opponents 7bbs calling range. After that we construct our shoving range in ICMIZER for average, loose and tight opponents. Then we build the bayesian estimator excel chart to help us determine if he is average, loose or tight player.
I understand all this and have no problem with it. It's perfectly fine for post game analysis and to check our errors, but in-game I have no way to get an accurate read for our opponents 6-8BBs calling percentages. If we use the COS stat from the CoffeeHUD it is filtered from 1-8BBs and the calling percentages from 1-8BBs are much higher than 6-8BBs calling percentages on average as they include short depth from 1-5BBs when people call much wider. So our final calculation will be way off. In fact, all players would appear much looser than they actually are…
If you want to use the HUD (as opposed to just paying attention/counting) the best plan would be to make your estimator for the entire 1-8bb, and use it to get a a loose/tight read on villain's entire push/fold game plan. If you get a loose/tight read from his overall 1-8bb game plan you can apply that read to the more specific individual stack depth regions as it is fairly likely that if he shoves looser than the population overall at 1-8bb, he will do it both at 5-7 and at 3-5 as well.
Hope this makes sense :)
Thanks very much for the quick answer,
Paying attention/counting seems very difficult, probably impossible if we play an opponent more than once. I think HUD is our best option, and making an estimator for the entire 1-8bb range looks like a great idea to get a loose/tight read. I'll definitely try this in next few days ...
"If you want to win, you must not lose!"