https://forumserver.twoplustwo.com/1...12/?highlight=
Yeah that thread was ok but I'm not bumping it. I was playing with math this morning in the beginners forum in an attempt to explain turn bluffcatching stuff. Here's the gist of what I got going through my head now:
There are two ev sources for bluffcatchers in equilibrium on the turn:
a) free showdown ev
b) draw value
"bluffcatching ev on the river" is not an ev source, because hands that can only beat a bluff are 0ev by definition in equilibrium, with the exception of hands with good blocking qualities, which are slightly +ev.
so to calculate (a) when facing a bet:
opponent's river check frequency * pot * my equity vs checking range = (free showdown ev)
to calculate (b) when facing a bet:
% hit hand * equity when hit hand * pot = (draw value)
the sum of (draw value + free showdown ev) = total ev of bluffcatching the turn in bb. If we set (pot = 1) then the value will be a fraction or a percentage of the pot.
Quote:
I raise on the button with A7o, small blind folds, big blind calls.
J45r
Big blind checks, I bet, big blind check raises, and I?
Let’s assume I call.
Turn 4o
Big blind bets.
free showdown equity vs 40% river check frequency: (40% * 8bb * 50%) = 1.6bb for 1 invested? This is a clear call with those assumptions.
different free showdown equity: (20% * 8bb * 50%) = 0.8bb showdown ev, we need some draw value to go along with it to make the call given these assumptions.
draw value with A7o: Maybe it's about 10% if we're lucky? debatable imo. I'll find the breakeven point given 0.8bb showdown ev: need to make up 0.2bb in draw value. 0.2bb = (8bb*%hit hand * equity when hit)
50% equity when hit and 5% hit hand result in 0.2bb draw value.
30% equity when hit and 8.3% hit hand result in 0.2bb draw value.
20% equity when hit and 12.5% hit hand result in 0.2bb draw value
so basically it doesn't take much to make this a call provided the 40% check frequency and 50% win frequency with A7o. I don't necessarily think those numbers are correct. I'm just looking for feedback, tweaks to the process.
I guess it's going to depend on the assumptions, but I think the process is correct.
thanks for reading.