PFR doubts!

[andrenalina]

I was reading about isoating impers and the autor said to look at "Fold to PFR" of the players behind. I didnt see this at PT3. I know that you have 2 chances of facing a PFR, when you limp or when you didnt act yet. The stats about limp and fold we have in PT3, but what about the second? How do we get the "Fold to PF raise" stat refering to only case when the player didnt limp.

Another question. When we talk about PF raise is it included 3bets and 4bets? When we talk about PFR its only about single raise or all raises? I mean, when we use the stat "call PF raise" its computed call 3bet or only call single-raise.

Thanks for the attention!

[kraada]

PFR counts any preflop raise, no matter what type of raise it is. Call PFR is true when you call any raise, and do not reraise yourself. 3Bet Preflop is when you raise exactly a single raise preflop.

The Fold to PFR stat doesn't exist by default in PT3, and though you could build it, I suppose, I'm not certain how useful it would be. We already have Call PFR, and 3 Bet Preflop, and since calling, raising and folding are your only options, folding is the amount that's left.

[andrenalina]

Hello, you said fold = 100 - call - raise, but I run a sampe of hands and the number was "Call PFR" = 7.33 (719/9809) e 3bet = 2.42 (219/9049). That 760 times? I faced a raise PF 9809 and cant 3bet probabily was the times I faced a all-in, right? Is there a problem if I simple subtract both of then from 100%, because that 760 times the 3bet was 0?

With respect to single raise, I woud ike to know how can I separate the action fold to PFR preflop that not include 3bets or 4bets, nor fold to PFR when limped, the second is less important, because we have the stat "fold to PFR when limped", so we can figure the ratio between them.

[kraada]

Hrm, you're right.

However we can build the stat you want in PT3.

The following columns can be used in conjunction with this tutorial to create a stat which will show you how often a player folds to a single preflop raise.

cnt_p_face_pfr_fold: sum(if[holdem_hand_player_statistics.flg_p_face_raise and NOT(holdem_hand_player_statistics.flg_p_3bet or holdem_hand_player_statistics.flg_p_4bet or holdem_hand_player_statistics.flg_p_3bet_def_opp or holdem_hand_player_statistics.flg_p_4bet_def_opp or holdem_hand_player_statistics.flg_p_limp) and enum_folded = 'P', 1, 0])

cnt_p_face_pfr: sum(if[holdem_hand_player_statistics.flg_p_face_raise and NOT(holdem_hand_player_statistics.flg_p_3bet or holdem_hand_player_statistics.flg_p_4bet or holdem_hand_player_statistics.flg_p_3bet_def_opp or holdem_hand_player_statistics.flg_p_4bet_def_opp or holdem_hand_player_statistics.flg_p_limp), 1, 0])

Put the former over the latter and multiply by 100 and that will give you your stat.

[andrenalina]

Thanks for the attention, after that I startet reading the article. I want a stat the say the percentagem of time the player first action is to fold to a PFR. So i made the follow 3 columns:

cnt_p_noaction_faceraise_fold : sum( if[ holdem_hand_player_statistics.flg_p_face_raise AND lookup_actions_p.action = 'F' AND NOT holdem_hand_player_statistics.flg_p_limp , 1 , 0 ])

cnt_p_noaction_faceraise_call = sum( if[ holdem_hand_player_statistics.flg_p_face_raise AND lookup_actions_p.action LIKE 'C%' AND NOT holdem_hand_player_statistics.flg_p_limp , 1 , 0 ]) (this includes 4bets calls when someone raises and another 3bets before you act)

cnt_p_noaction_faceraise_3bet = sum( if[ holdem_hand_player_statistics.flg_p_face_raise AND lookup_actions_p.action LIKE 'R%' AND holdem_hand_player_statistics.flg_p_3bet AND NOT holdem_hand_player_statistics.flg_p_limp , 1 , 0 ]) (I included the p_3bet because if we not do this, the first raise of us could be a single raise then face a 3bet)

So, to make this stats we just need to defnine it

Preflop No Action Face Raise Fold = (cnt_p_noaction_faceraise_fold / ( cnt_p_noaction_faceraise_fold + cnt_p_noaction_faceraise_call + cnt_p_noaction_faceraise_3bet)) * 100

Is that right? I tryied it in a report and it seems nice.

You said

cnt_p_face_pfr_fold: sum(if[holdem_hand_player_statistics.flg_p_face_raise and NOT(holdem_hand_player_statistics.flg_p_3bet or holdem_hand_player_statistics.flg_p_4bet or holdem_hand_player_statistics.flg_p_3bet_def_opp or holdem_hand_player_statistics.flg_p_4bet_def_opp or holdem_hand_player_statistics.flg_p_limp) and enum_folded = 'P', 1, 0])

My doubts are

holdem_hand_player_statistics.flg_p_3bet: If its true its not counted. Does this mean that if theres some 3bet from anyone in the hand its not verified?
holdem_hand_player_statistics.flg_p_3bet_def_opp: Who faces a 3bet PF? Me or anyone?
enum_folded = 'P': It means that if we fold preflop its verified, right?

[WhiteRider]

Your columns look OK.
Just to note, though, if the player's only action is to fold then they can't have limped, so you don't need the flg_p_limp check:
lookup_actions_p.action = 'F' AND NOT holdem_hand_player_statistics.flg_p_limp

With respect to your doubts: all these checks refer to the player the stat is being viewed for. If you're looking at a stat value for yourself then it checks that you faced a 3-bet (or whatever); if you're looking at it for VillainA then it checks that VillainA faced a 3-bet.

[andrenalina]

Hi, I think I should change the "cnt_p_pfr_1st_action_3bet_or_higher" to

cnt_p_pfr_1st_action_3bet_or_higher = sum( if[ holdem_hand_player_statistics.flg_p_face_raise AND (holdem_hand_player_statistics.flg_p_3bet OR holdem_hand_player_statistics.flg_p_4bet) AND lookup_actions_p.action LIKE 'R%' , 1 , 0 ])


Because the other way we would not count 4bets raises first Action, and the part of limp we dont need, because of lookup_actions_p.action LIKE 'R%' and the action R must be 3bet or 4bet, right?

[WhiteRider]

You could also check for
NOT (holdem_hand_player_statistics.flg_p_limp OR holdem_hand_player_statistics.flg_p_first_raise) to rule out having limped or 2-bet, although if you're checking for the first action being raise then you don't need to check for limps at all since then the first action would be a call.

[andrenalina]

Hello, I think if we put just theese it will be counted the times the player called ou raised a pfr, ok? But With your help I managed to make te consition shorter:

cnt_p_pf_1st_action_fold_raise
sum( if[ lookup_actions_p.action = 'F' AND holdem_hand_player_statistics.flg_p_face_raise , 1 , 0 ])

cnt_p_pf_1st_action_call_raise
sum( if[ lookup_actions_p.action LIKE 'C%' AND NOT holdem_hand_player_statistics.flg_p_limp , 1 , 0 ])

cnt_p_pf_1st_action_reraise
sum( if[ lookup_actions_p.action LIKE 'R%' AND NOT holdem_hand_player_statistics.flg_p_first_raise , 1 , 0 ])


Is that right?

Its only needed to change the description of the stat.

[andrenalina]

Its obvious that "cnt_p_pf_1st_action" is not correct, we need to write "cnt_p_1st_action".