M = effective Stack or Hero Stack?

[js2002]

wtsd_yz

sum(if[tourney_holdem_hand_player_statistics.flg_showdown and

((tourney_holdem_hand_player_detail.amt_before / ( tourney_holdem_blinds.amt_bb + tourney_holdem_blinds.amt_sb + (tourney_holdem_hand_summary.cnt_players * tourney_holdem_hand_player_detail.amt_ante)) <= 20)
and
(tourney_holdem_hand_player_detail.amt_before / ( tourney_holdem_blinds.amt_bb + tourney_holdem_blinds.amt_sb + (tourney_holdem_hand_summary.cnt_players * tourney_holdem_hand_player_detail.amt_ante)) > 10))


, 1, 0])

1. what is tourney_holdem_hand_player_detail.amt_before?
2. Why isnt a effective stack expression inside? Is this column only about the "real" hero stack and not the effective one?

[WhiteRider]

Is this one of Kraada's stats?
1. That is the amount of chips the player had before the hand started.
2. Kraada would have to answer that, but yes, it looks like it is just the player's M, not effective M. I guess that's a personal choice - if there was one very short stack at the table, would you want that to affect things if everyone else was relatively deep?

[js2002]

Is this one of Kraada's stats?
1. That is the amount of chips the player had before the hand started.
2. Kraada would have to answer that, but yes, it looks like it is just the player's M, not effective M. I guess that's a personal choice - if there was one very short stack at the table, would you want that to affect things if everyone else was relatively deep?

2. is releavant 4 HU and shorthanded tables. So, the M doesnt help for HU because I decide on the effective stack and never on my stack. Ok sometimes I am the effektive stack^^

[WhiteRider]

Yes, I guess for heads up the effective stack would be a better measure. That should be pretty easy to change for a man of your skills, using holdem_hand_summary.amt_p_effective_stack?

[js2002]

thx^^

u are right it easy easy I guess but its much work because we would have to chance many stats.
I think I ll build the most usefull 4 HU first and others maybe later...

[preparac]

As I know it's at the moment NOT possible to do "effective stack" stats in the sense, that they would measure allways your stacksize AND (if you are NOT the shortstack) the second biggest stack still active in the hand at any point of the action tree ... You CAN do it for HU play using holdem_hand_summary.amt_p_effective_stack though ... but as this measures allways the SMALLEST stack at the table REGARDLESS if he is still active in the hand, you will be way off sometimes at tabels with more than 2 players depending on the concret action (as for example: small stack folds the BTN, you are in the SB with a big stack and have another big stack in the BB - here effective stack would indicate you have the opportunity to push allin wide because the effective stacks are small, which obv. would not be realsitic).

I think this limitation is actually one of the bigger flaws - if not THE biggest flaw - of PT3 for advanced users, as effective stacksizes are absolutely crucial for realistic ranges especially (but not only) in tournament play (as for example push ranges, raise/fold ranges and so on...). I already requested this feature a while ago but didn't get a clear answer if it is planned to introduce it in the feature.

... anyway, I would be happy to be corrected in this point by the PT3 staff.

[WhiteRider]

Yes, I made that request too when the effective stacks were introduced; however adding that information would be pretty complicated as you'd need effective stacks at several different points in a hand (potentially).
e.g. If a short-stack limped and you limp (deep) and another big stack then makes a small raise behind you and the short-stack folds the effective stacks change. You also have to consider whether you even care about the short-stack - maybe they have 3BB and you and the other villain have 50BB. Would the short stack affect your play?

[js2002]

Well I think the best we is to start it with PT4 or at the beginning ==> the parsing of a hand -> We see the stacks and can store them into db fields aka columns

[preparac]

Yes, I made that request too when the effective stacks were introduced; however adding that information would be pretty complicated as you'd need effective stacks at several different points in a hand (potentially).
e.g. If a short-stack limped and you limp (deep) and another big stack then makes a small raise behind you and the short-stack folds the effective stacks change. You also have to consider whether you even care about the short-stack - maybe they have 3BB and you and the other villain have 50BB. Would the short stack affect your play?

... it obv. would in some situations. But ok, I could live with it, you probably cannot have everything in every situation. :?
And yes I agree, it could become complicated, because effective stack sizes can change on the same street (with every player folding). Nevertheless: it would be a crucial feature, because with wrong stack sizes you could get wrong information about the EFFECTIVE tendencies/ranges of a player ... and wrong information is probably worse than no one. So ... Can you tell me, if for PT3 it's is only complicated OR not at all possible to integrate this feature in the future?

[WhiteRider]

I can't give you a definitive answer, I'm afraid. I think that this is unlikely to be added in the sort of detail you'd need to have everything fully accurate.