PokerTracker

Hey,

I created the stat coldcall vs position, and filtered it in the HUD profiles again after position. So I would get a popup that shows me coldcall MPvUTG, COvMP etc.

This works partially.


The stat compares the custom columns below against the default cnt_p_3bet_opp_vs_[pos]_open -> (cnt_p_coldcall_vs_[pos]_open / cnt_p_3bet_opp_vs_[pos]_open) * 100


Custom columns:

note: i used the "cash_hand_player_statistics.flg_p_ccall" originally, but it resultet in zeros for all positions where the coldcaller is oop. Now it works for BBvBU and SvBU.

cnt_p_coldcall_vs_btn_open

Code: Select all

sum(if[cash_hand_summary.str_aggressors_p LIKE '80%' and
cash_hand_summary.str_actors_p LIKE '0%' and
cash_hand_player_statistics.flg_vpip and
not(cash_hand_player_statistics.flg_p_3bet or cash_hand_player_statistics.flg_p_squeeze_opp), 1, 0])

cnt_p_coldcall_vs_co_open

Code: Select all

sum(if[cash_hand_summary.str_aggressors_p LIKE '81%' and
cash_hand_summary.str_actors_p LIKE '1%' and
cash_hand_player_statistics.flg_vpip and
not(cash_hand_player_statistics.flg_p_3bet or cash_hand_player_statistics.flg_p_squeeze_opp), 1, 0])

cnt_p_coldcall_vs_mp_open

Code: Select all

sum(if[cash_hand_summary.str_aggressors_p LIKE '82%' and
cash_hand_summary.str_actors_p LIKE '2%' and
cash_hand_player_statistics.flg_vpip and
not(cash_hand_player_statistics.flg_p_3bet or cash_hand_player_statistics.flg_p_squeeze_opp), 1, 0])

cnt_p_coldcall_vs_utg_open

Code: Select all

sum(if[cash_hand_summary.str_aggressors_p LIKE '83%' and
cash_hand_summary.str_actors_p LIKE '3%' and
cash_hand_player_statistics.flg_vpip and
not(cash_hand_player_statistics.flg_p_3bet or cash_hand_player_statistics.flg_p_squeeze_opp), 1, 0])

What is wrong?

regards,
madsquare

That looks like it should to me. The places without data can't happen - you can't cold call a button open from the CO - the CO acts first and has to put money in then after he does the button can't open. Basically you can't cold call out of position because there's no way for you to get to a spot where you haven't put money in and the person in position has raised.

I think there is a misunderstanding. Here are a picture and a screenshot to clarify.




http://i.imgur.com/ACUfY.jpg



what I dont understand is why its working BBvBU but not BBvCO, even though the stats are made up the exact same way.

Do you never get data for those specific stats that show 0 in the screenshot? I see no reason why they would always be 0 - I thought you were confused why the rest of the graph had no data.

yeah, the problem are the stats that show zero.
I saved the the stats, clicked ok on the configure stats window, reloaded the reports, tried different players, tried replayer HUD and actual table HUD, but they always show something like 0/83, 0/152.
I have zero idea why, and weirdly enough, it doesnt affect the BLvBU ones, wich are 99,9% similar to BLvCO.

Looking for an alternative or solution so I can use them ingame. Also, curious about reasons and a solution, just for the sake of it.

regards.

Well, the reason flg_p_ccall didn't work like you expected is that it requires you to have not put any money in the pot previously - and blind money counts.

So shifting to flg_vpip was a good idea.

Assuming you used the exact column definitions pasted, I don't see why vs button would work but vs co would not.

Could you export these two stats, zip them up and send them to us via a support ticket? That would help us look at these in greater detail for you and we could run tests on our end to see what is going on here.

For some reason its working now.
I just copied the text here in the forum and saved, now they show numbers.
I validate all my stats before saving, but I remember that the text i postet is not the exact same as the one copied from my stats. There was oringially a line break missing. I thought its not relevant and would only be due forum formatting, since I validate my stats. But it seems like it was in fact relevant.

Thanks for your help.

Hey,

it appears there is something wrong with the stats. The numbers seem off.

1. my own stats are over large samples 3-10% off my default coldcalling ranges (frequencies), all positions/possibilities. (too low)

2. this custom cc v pos stat and the corrosponding PT-default 3b v pos stat dont add up to the PT-default fold v pos stat [cc + 3b < (1-fold)]; it should be equal.

so I conclude
a) cc v pos stat is broken
or
b) the discrepancies from 1) are normal and instead the PT-default 'fold v pos' stat is broken


would appreciate some help with this.

to be clear what I want from coldcall vs pos
preflop call vs openraise, no other caller - where the openraiser is in a specific position (3, 2 , 1 , 0, 9)


greetings.

(1) Just because you would cold call a certain percentage in a vacuum doesn't mean that you actually do that. The actual hands you are dealt aren't going to match up exactly to the theoretical random distribution because of that very randomness. Also you may well skew based on opponent tendencies -- cold calling less or more depending who is doing the raising. And that of course will affect your percentages. Without a specific example of a case that you think should match but doesn't, I can't say what is going on here.

(2) Are you looking at fold vs open or fold vs 2bet? The former will eliminate some hands and might be why you're confused. For example, if EP limps and the CO raises that's a CO 2bet not a CO open, but you can cold call on the button.

about 1
Im aware of variance and skewing.
The stats have been consinstently off for about 400k hands.
Wich becomes suspiscious exactly at the time similar % are missing as described in 2.


about 2
Here are the fold v pos stats. Columns are default / original PT defined

Code: Select all

((cnt_p_3bet_opp_vs_btn_open - cnt_p_call_vs_btn_open - cnt_p_3bet_vs_btn_open) / cnt_p_3bet_opp_vs_btn_open) * 100

Code: Select all

((cnt_p_3bet_opp_vs_co_open - cnt_p_call_vs_co_open - cnt_p_3bet_vs_co_open) / cnt_p_3bet_opp_vs_co_open) * 100

Im mostly interested in verfiying / correcting the coldcall v pos stats. Instead of chasing rogue %s.