Page 1 of 2

Postional stat values

PostPosted: Mon Mar 28, 2016 2:21 pm
by vankammen
Hi,

Why does: tourney_hand_summary.str_aggressors_p NOT LIKE '89%' represent Late Postion ? I tought that the % could be empty or any string. Therefor i thought this expression would
result in any position except the blinds.

i had this:

tourney_hand_summary.str_aggressors_p = 1 OR tourney_hand_summary.str_aggressors_p =1 (for late position players, b and co)

Re: Postional stat values

PostPosted: Mon Mar 28, 2016 3:52 pm
by kraada
[Mod EDIT] It looks as if kraada missed the 'NOT'. This expression means that the first raiser was NOT the small blind, so was therefore late position because we're testing for a steal not from the small blind.

tourney_hand_summary.str_aggressors_p NOT LIKE '89%' means that the first aggressor in the hand was the small blind - where are you getting that it has anything to do with late position?

tourney_hand_summary.str_aggressors_p = 1 will always be false - the string always starts with 8 as the first aggressor is always the big blind.

Re: Postional stat values

PostPosted: Mon Mar 28, 2016 4:14 pm
by vankammen
found it here:

https://na2.pokertracker.com/forums/vie ... 1&p=284057

so for LP aggressors it should be:

tourney_hand_summary.str_aggressors_p = '81' OR tourney_hand_summary.str_aggressors_p ='80'

Right?

I have now this for all possible lp min raises (raise smaller then 2.5 times the bb) when the raiser has between 11 and 30 big blinds:

Code: Select all
sum(if[tourney_hand_player_statistics.flg_blind_def_opp AND

( tourney_hand_summary.str_aggressors_p = '80' OR tourney_hand_summary.str_aggressors_p = '81')  AND


((tourney_hand_player_statistics.amt_p_effective_stack / tourney_blinds.amt_bb) BETWEEN 11 AND 30)

 AND


(      ( tourney_hand_player_statistics.position =8 AND tourney_hand_player_statistics.amt_p_2bet_facing < tourney_blinds.amt_bb*1.5 ) OR
 ( tourney_hand_player_statistics.position =9 AND tourney_hand_player_statistics.amt_p_2bet_facing < tourney_blinds.amt_bb*2.0 ) )



, 1, 0])



And this for every time the player in the blinds folds to such a raise:

Code: Select all
sum(if[tourney_hand_player_statistics.flg_blind_def_opp AND lookup_actions_p.action = 'F' AND

( tourney_hand_summary.str_aggressors_p = '80' OR tourney_hand_summary.str_aggressors_p = '81')  AND





((tourney_hand_player_statistics.amt_p_effective_stack / tourney_blinds.amt_bb) BETWEEN 11 AND 30)

AND

(      ( tourney_hand_player_statistics.position =8 AND tourney_hand_player_statistics.amt_p_2bet_facing < tourney_blinds.amt_bb*1.5 ) OR
 ( tourney_hand_player_statistics.position =9 AND tourney_hand_player_statistics.amt_p_2bet_facing < tourney_blinds.amt_bb*2.0 ) )



, 1, 0])


Is this correct?

Re: Postional stat values

PostPosted: Tue Mar 29, 2016 3:07 am
by WhiteRider
You're close, but you'll need to use "LIKE" instead of equals when checking the start of the aggressors strings otherwise it will not count hands where the blinds re-raise:

tourney_hand_summary.str_aggressors_p LIKE '81%' OR tourney_hand_summary.str_aggressors_p LIKE '80%'

..but otherwise your expressions look OK.

Re: Postional stat values

PostPosted: Tue Mar 29, 2016 6:41 am
by vankammen
WhiteRider wrote:You're close, but you'll need to use "LIKE" instead of equals when checking the start of the aggressors strings otherwise it will not count hands where the blinds re-raise:

tourney_hand_summary.str_aggressors_p LIKE '81%' OR tourney_hand_summary.str_aggressors_p LIKE '80%'

..but otherwise your expressions look OK.


Thanks. But i do not fully understand it yet. I want to count all the times the blinds fold to a LP min raise. So why do i need to count the hands where the blinds re-raise?

The first code just counts all the hands where the blinds face a CO or B minraise. The second piece counts all the times that the sb or bb calls such a raise. It is not clear to me why i need to
count the hands where the blinds reraise. Could you please clarify this for me? :)

Re: Postional stat values

PostPosted: Tue Mar 29, 2016 7:19 am
by kraada
You want hands where the blinds do reraise to be part of the opportunity sample - it's fine for the folding side, but if you exclude 3bets from the opportunity you're going to think people are folding more than they actually are because you're only counting folds / (folds + calls) instead of folds / (folds + calls + raises).

Re: Postional stat values

PostPosted: Tue Mar 29, 2016 8:04 am
by vankammen
check. thanks guys!

Re: Postional stat values

PostPosted: Sat Dec 17, 2016 1:39 pm
by alexrjl
kraada wrote:tourney_hand_summary.str_aggressors_p NOT LIKE '89%' means that the first aggressor in the hand was the small blind - where are you getting that it has anything to do with late position?

tourney_hand_summary.str_aggressors_p = 1 will always be false - the string always starts with 8 as the first aggressor is always the big blind.


Why is the (built in) column for 3bet opportunity versus late position open like this then?
cnt_steal_def_3bet_opp_vs_lp
sum(if[tourney_hand_player_statistics.flg_blind_def_opp and tourney_hand_summary.str_aggressors_p NOT LIKE '89%' AND tourney_hand_player_statistics.flg_p_3bet_opp, 1, 0])

Does this mean this is actually wrong and it's counting 3bet opportunities versus a SB raise?

Re: Postional stat values

PostPosted: Sat Dec 17, 2016 2:56 pm
by WhiteRider
tourney_hand_summary.str_aggressors_p NOT LIKE '89%'

..means that the 2bet was NOT by the small blind, and since the player faced a steal it must therefore have come from either the cutoff or button.

Re: Postional stat values

PostPosted: Sun Dec 18, 2016 6:53 am
by alexrjl
WhiteRider wrote:tourney_hand_summary.str_aggressors_p NOT LIKE '89%'

..means that the 2bet was NOT by the small blind, and since the player faced a steal it must therefore have come from either the cutoff or button.


that's what I thought, which I think means kraada's above post is incorrect (looks like he missed the NOT)

highfalutin