PokerTracker

Hello,

I am trying to use global filters to specify node and texture while using expression filters to specify player profile.

For example:

Global filters specify BBvBU SRP, river cbet

Expression filters specify vpip, pfr, and 3bet frequency bounds to define a player profile.

Custom stats populate the report, for example, fold to XYZ sizes vs river bet.

My first iteration with this workflow resulted in a null result, just a blank report which presumably meant it turned up 0 hands.

I'm assuming that the way filters work is that the global filters are applied first, and then the expression filters are applied. So in this case, the global filters removed all 3bet hands from the report, and then the expression filter removed all players with 0% 3bet, resulting in an empty set.

(1) Is this assumption correct?

(2) Is there a more elegant solution to this problem than coding the custom stats themselves to specify the node and texture? I'd like to avoid all of that coding if it can be helped.

It's difficult to comment on this without more specific details such as the exact expression filter(s) being used.

#Hands# > 300
#VPIP# >= 23
#PFR# > 18
#VPIP# <= 29
#3Bet Preflop# > 7

The first global filter is Maximum Raise 2bet.

Simple expression filters like that just work with the stat values in the report so if you are filtering then that's not going to be based on overall stat values anymore. If you have a filtered report and want to filter data based on a players overall statistics then that will require a different expression filter - see this thread for an example.

Thank you, I used that format with success.

Future readers may find this info useful - you can stack these filters by simply adding " AND " in between instances. I am saving my player profiles to text files to copy and paste all of the stat filters in one easy go.

However, I realized that for my purposes I need to filter out HU hands.

Taking this VPIP example:

player.id_player in (SELECT p.id_player from player p, cash_cache cc where p.id_player = cc.id_player group by p.id_player HAVING (( sum(cc.cnt_vpip) / (sum(cc.cnt_hands) - sum(cc.cnt_walks)) >= .23)))

Is there a way to modify this language in order to only count hands with 3 or more players dealt in? Would this require creating a new VPIP stat coded for >2 players?

You will need custom columns to calculate that for 3+ handed play before querying the column values in a new expression filter.