confused about WR for filtered situations

[d0zer]

hey,

i am trying to filter for a certain situation (25bb hypers) to compare the postflop EVs of either action

say 10-25bb limped pot we are SB, and have Q2o+ Q2s+ (Q high)

i have thousands of hands of a sample for both SB check and SB bet with Q high, and the overall EV of checking is say -5bb/100, where as betting is around +40bb/100, similar results are true for all sorts of different high card hands.

yet obviously our strategy should not be to bet everything when checked to in spots like these, so my question is - what am i missing ? does the flop X vs BET decision point EV already not include all the possible outcomes on later streets ?

using similar EV comparison lines for preflop seems to work fine, but gets us completely lost in postflop situations and has me completely confused ?

hopefully you can help

[WhiteRider]

PT4 doesn't calculate "EV" as that requires estimations of villain ranges. It calculates net adjusted results for all-in hands, so for any hands you don't get all-in the net adjusted results will match your actual results.
What exact stat are you looking at? If you're looking at any ICM tournament stats I think those will be for the overall tournaments including those hands, but net adjusted chip results would be for those hands only, but again only those where you were actually all-in.

[d0zer]

PT4 doesn't calculate "EV" as that requires estimations of villain ranges. It calculates net adjusted results for all-in hands, so for any hands you don't get all-in the net adjusted results will match your actual results.
What exact stat are you looking at? If you're looking at any ICM tournament stats I think those will be for the overall tournaments including those hands, but net adjusted chip results would be for those hands only, but again only those where you were actually all-in.

does the EVbb/100 stat not reflect my actual results for hands then and only calculates it for hands that ended up all-in ?

i am trying to compare winrates for different lines, for example limped HU, BB checks, and checks the flop to us (so we can c-bet in a limped pot), and i want to compare my results for betting vs checking in this spot, as well as many others

not looking at ICM stuff, i am focusing on spin & go format (3H + HU, winner takes all, ICM isnt a factor)

[WhiteRider]

does the EVbb/100 stat not reflect my actual results for hands then and only calculates it for hands that ended up all-in ?

Correct. There is no "EV" in PT4, only "all-in equity adjusted" which is only calculated for all-in hands. Guide (There is also this guide about the tournament ICM side of equity adjusted calculations)

[d0zer]

does the EVbb/100 stat not reflect my actual results for hands then and only calculates it for hands that ended up all-in ?

Correct. There is no "EV" in PT4, only "all-in equity adjusted" which is only calculated for all-in hands. Guide (There is also this guide about the tournament ICM side of equity adjusted calculations)

sorry i must have misused the terms, the stat in the screenshot that i am talking about is "all-in adjusted bb/100", which reflects the winrate of the hand at that decision point, right ? my initial confusion arose from me using some incorrect filters and arriving to wrong conclusions

additionally, how big of a sample do i need to draw conclusions for preflop winrates of different lines / how quickly do they converge ? how do you arrive to the number, of say if its 500 hands ?

[WhiteRider]

sorry i must have misused the terms, the stat in the screenshot that i am talking about is "all-in adjusted bb/100", which reflects the winrate of the hand at that decision point, right ? my initial confusion arose from me using some incorrect filters and arriving to wrong conclusions

Yes, you're using the right stat, I just wanted to make sure you understood the distinction. This stat only varies from your actual win rate for hands where you were all-in, against the actual hand you were up against each time and not some theoretical or estimated range, and is calculated at the point of the all-in not the point of the actions you've filtered for.
If your bet at this point is always all-in then this stat is somewhat relevant for that.
If you check, it isn't any different to using your actual win rate (because any all-in would come at a later point).

additionally, how big of a sample do i need to draw conclusions for preflop winrates of different lines / how quickly do they converge ? how do you arrive to the number, of say if its 500 hands ?

That is beyond my knowledge I'm afraid, but given the multitude of different variables (relative stack sizes, board textures, etc) I'd say it would be pretty hard to draw any concrete conclusions from a few hundred hands. You may want to discuss that kind of thing further in a poker strategy forum like those on twoplustwo.