Falsehood of WTSD?

[PeteX]

I got this idea that WTSD is not really a good indicator of the showdown-boundness of a player. The reason being that it includes the cases when a player has seen showdown "free", i.e. just checked through all streets. Or maybe bet flop and gotten called and then the turn and river were checked down. In other words, the player went to showdown WITHOUT facing resistance. Basically, if a player limps preflop and then gets to see free showdown when every street just gets checked through OR a player raises preflop and then calls three streets of bets to see showdown will get the same WTSD-value, although I'd consider the second player to be much more showdown-bound. Obviously we can't tell what the first player would have done on a single hand when facing a bet/raise, but after more samples the difference would become clear.

I'd like to fix this and create a better stat that would be something like "WTSD when facing resistance". Basically it would be 100% if the player is always calling all bets/raises that are asked of him. And obviously 0% when the player is never going to see the showdown when having to pay anything, maybe limiting that to postflop.

Am I making sense? Is my analysis of WTSD correct?

How would you approach making such a stat?

[kraada]

How would you handle cases of getting to showdown when that player was the aggressor?

[PeteX]

Hmm.... Why wouldn't the idea work for an agressor as well; I mean, the player can still bet/raise a lot postflop BUT he will rather fold than call when facing a bet/reraise himself. I guess one style of sensible LAGish play postflop would be to play very agro betting and stabbing constantly until meeting resistance and then giving up rather than calling. The bet/raise tendency would be revealed with AF/AFq.

Hmm... So I guess what about a conceptual formula such as: "folded postflop when facing a raise or bet" / "faced a bet or raise postflop"
or in other words
sum("folded postflop when having the opportunity to defend against a bet or raise") / sum("having the opportunity to defend against a bet or raise")
or why not:
"number of folds postflop" / "number of folds postflop + number of calls postflop + number of raises postflop"
basically meaning the proportion of folds from all such actions that can be done when facing a bet/raise. So one would just forget about bets and checks as they are not done "under pressure".

That would be a sorta of a modification of the formulas used for AF and AFq. That could pretty much give out what I'm after?

Basically the stat would be to help play weak/medium hands to see how easy it is to bluff the villain out of the pot.

[PeteX]

I suppose like this:
((cnt_f_fold + cnt_t_fold + cnt_r_fold) / (cnt_f_call + cnt_f_fold + cnt_f_raise + cnt_t_call + cnt_t_fold + cnt_t_raise + cnt_r_call + cnt_r_fold + cnt_r_raise)) * 100

So the higher the number, the easier it is to push the player out of the pot postflop before showdown.

Would this leave out any scenarios?

[WhiteRider]

That looks as if it will calculate the number you want.