Hey my newly found friends,

My understanding of all-in-adj stats is that it accounts for showdown equity before the remaining cards are dealt, when it is an all-in situation. Meaning that if someone goes all-in for 100bb has an 80% chance of winning the hand, they will gain 80bb in their all-in-adj stats, regardless of who actually wins in the end (the true result will either be +100b or -100b).

This is to help account for luck over the course of the sample size. It seems imperfect, though, for example when two players go all-in after the river has already been dealt, and the losing player discovers the winner sucked-out because the river happened to give the opponent a two-pair. He got unlucky, but that isn't shown in the data.

Anyway, my question is... how useful are all-in-adj stats? Is my all-in-adj bb/100 strictly superior information to my actual bb/100?

Is there even much point to looking at my actual bb/100 besides trying to tell if I'm running hot or cold? I wouldn't think so, with my current perception of how these things work.

Thanks in advance,

~BAS

It's more important.

I don't even display my bb/100 on my tracker. I only display the all in adj equivalent.

I'd go so far as to say bb/100 is useless statistic when the all in adj version is available

It's a "variance reduction technique". It's not the same thing as your true expected value. It's a way of clarifying your variance-riddeled win rate.

Your "all-in-adjusted" Win Rate is more accurate and will converge to your true win rate more quickly.

Looking at the difference between your actual and "adjusted" win rate can provide insight. You can see, for example, if you're on a heater (running over EV), or running bad (running under EV).

Your actual bb/100 winrate doesn't reflect this as being unlucky, either.

Yes, it's basically your actual bb/100 with one specific source of variance (variance after allin and call) filtered out. It's strictly superior.

Once you have enough hands. How else would you suggest evaluating whether you're winning or losing?

Your EV is actually 60bb in that situation.

80% of the time, you make 100bb of profit.
20% of the time, you lose 100bb.

EV = (80% * 100)-(20% * 100) = 80-20 = 60.

So on your graph, you'll either be 40bb above EV (when you win 100bb), or a horrific 160bb below (when you lose 100bb).

Assuming your software uses an orange EV line, like HEM does, it's a fairly good indicator of your theoretical winrate (better than actual winnings), particularly once you've played a decent volume. I tend to use EVbb/100 and the orange line on graphs when analysing databases.

Thanks everyone!

Good point about the math, Arty. For some reason my head assumed that when I lose I just win 0bb rather than actually losing 100bb. So I had to remove another 20% of equity to get 60bb. I won't forget that one.

The answer would be different considering how much there is to win after V calls. You're risking 100 to win 100+ V's Call. I don't know why you even bothered calculating EV of a scenario without enough info to run the proper calculation.