Can I create a stats like this?

[Caxero]

I want to create a report to analise some players frequency of raise flop cbet, and bet turn. Can I do it with a stats, or its only possible with a Filter?

[Flag_Hippo]

There is a 'Bet Turn After Raise Flop CBet' custom statistic available for download here.

[GregXX]

There is a 'Bet Turn After Raise Flop CBet' custom statistic available for download here.

Trying to create a little bit different stat like this: I need Bet Turn after Any Raise (vs both Cb, Float or any other bet of any kind regardless of limped/raised/3b pots) on the Flop. Can you please help me with the column?

Initial column of built-in Bet Turn After Raise Flop Cbet is

sum(if[tourney_hand_player_statistics.flg_t_bet and tourney_hand_player_statistics.enum_f_cbet_action = 'R' and not(tourney_hand_player_statistics.flg_f_3bet_opp or tourney_hand_player_statistics.flg_f_4bet_opp), 1, 0])

I assume I need to change red part, but a bit unsure what is correct expression?

P.S. For the first view it looks like I just need to delete C letter in word "cbet", but it feels too simple ))

tourney_hand_player_statistics.enum_f_cbet_action = 'R'

[Flag_Hippo]

You cannot remove the 'C' because enum_f_bet_action does not exist in the database schema but you can substitute the following:

Code: Select all

tourney_hand_player_statistics.amt_f_bet_facing > 0 AND lookup_actions_f.action = 'R'

[GregXX]

You cannot remove the 'C' because enum_f_bet_action does not exist in the database schema but you can substitute the following:

Code: Select all

tourney_hand_player_statistics.amt_f_bet_facing > 0 AND lookup_actions_f.action = 'R'

I most likely made something wrong, because as a result I get a new stat (Bet Turn After Raise any F bet), which significantly less than initial Bet Turn After Raise F Cbet. I had 5 samples for new stat VS 24 samples for initial (UPD for Large samles stay the same 666 samples vs 1670 for example) . Sure Cb happens quite often, but instead new stat should for example include limped pots or XR vs Float-bets (I assume that flop action R in PT4 includes any flop raises IP as well as OOP, right? i.e XR)

I just post my expression and columns, please check it (Validation is fine for all of them though)

Expression:

(cnt_t_bet_f_bet_def_action_raise / cnt_t_open_opp_f_bet_def_action_raise) * 100

Column cnt_t_bet_f_bet_def_action_raise

sum(if[tourney_hand_player_statistics.flg_t_bet and tourney_hand_player_statistics.amt_f_bet_facing > 0 AND lookup_actions_f.action = 'R' and not(tourney_hand_player_statistics.flg_f_3bet_opp or tourney_hand_player_statistics.flg_f_4bet_opp), 1, 0])

Column cnt_t_open_opp_f_bet_def_action_raise

sum(if[tourney_hand_player_statistics.flg_t_open_opp and tourney_hand_player_statistics.amt_f_bet_facing > 0 AND lookup_actions_f.action = 'R' and not(tourney_hand_player_statistics.flg_f_3bet_opp or tourney_hand_player_statistics.flg_f_4bet_opp), 1, 0])

[GregXX]

Working with custom stats is always a challenge))) Reading some guides now...

May be this operator is more suitable for my case: lookup_actions_f.action LIKE '%R'
???

[GregXX]

Have to update this thread given I'm learning on the fly)

1) I have replaced = "R" with LIKE '%R' to include both OOP & IP situations
2) I have made Custom cache rebuild and now numbers look fine.

But I still have a small question on how tourney_hand_player_statistics.amt_f_bet_facing > 0 affects and why it's necessary to use it in combine with lookup_actions_f.action LIKE '%R' . For the first glance it seems like lookup_actions_f.action LIKE '%R' is already describing any spot itself where Hero either raises flop or XR. Can Hero actually Raise/XR Flop when not facing a bet>0 ?

[Flag_Hippo]

1) I have replaced = "R" with LIKE '%R' to include both OOP & IP situations

Apologies I made an error not changing the pattern match query but given the original definition of the statistic it would be more accurate to use this:

Code: Select all

lookup_actions_f.action SIMILAR TO '(XR|R)'

But I still have a small question on how tourney_hand_player_statistics.amt_f_bet_facing > 0 affects and why it's necessary to use it in combine with lookup_actions_f.action LIKE '%R' . For the first glance it seems like lookup_actions_f.action LIKE '%R' is already describing any spot itself where Hero either raises flop or XR. Can Hero actually Raise/XR Flop when not facing a bet>0 ?

Yes but the definition of the original statistic excludes hands where there was further flop raises by the player and using '%R' would include hands where the players flop action was 'RR'. If you don't want that you should use the SIMILAR pattern matching I gave above but if you do want those hands included then you can leave things as they are. Also in multiway pots it's possible for a player to raise/check raise without facing a bet so tourney_hand_player_statistics.amt_f_bet_facing > 0 should be used to not count hands like that.

[GregXX]

Also in multiway pots it's possible for a player to raise/check raise without facing a bet so tourney_hand_player_statistics.amt_f_bet_facing > 0 should be used to not count hands like that.

Can you please elaborate on this?

Example: Hero on SB and he checks; If UTG bets and MP raises, then (if SB makes reraise) PT4 will count this spot as a suitable for my conditions without adding tourney_hand_player_statistics.amt_f_bet_facing > 0 (because SB faced no bets, just raise straight away)?

But if I use combination of, for example, lookup_actions_f.action SIMILAR TO '(XR|R)' and and not(tourney_hand_player_statistics.flg_f_3bet_opp or tourney_hand_player_statistics.flg_f_4bet_opp) do I still have to add in adition tourney_hand_player_statistics.amt_f_bet_facing > 0???

I.e. from the logical standpoint, it looks like this denial of 3b/4b situations properly replace requirement to face bet>0 and not any other raise.

[Flag_Hippo]

Can you please elaborate on this?

Example: Hero on SB and he checks; If UTG bets and MP raises, then (if SB makes reraise) PT4 will count this spot as a suitable for my conditions without adding tourney_hand_player_statistics.amt_f_bet_facing > 0 (because SB faced no bets, just raise straight away)

The original statistic you were using as a base was counting raises vs flop CBets only and was being adjusted from that base (vs Flop Bets instead of just Flop CBets) so it depends on whether you only want to count flop raises vs bets or not - it's your statistic. The SB is not facing a bet here so this wouldn't be counted so if you wanted to count this for SB you can remove the tourney_hand_player_statistics.amt_f_bet_facing > 0 condition.

But if I use combination of, for example, lookup_actions_f.action SIMILAR TO '(XR|R)' and and not(tourney_hand_player_statistics.flg_f_3bet_opp or tourney_hand_player_statistics.flg_f_4bet_opp) do I still have to add in adition tourney_hand_player_statistics.amt_f_bet_facing > 0???

I.e. from the logical standpoint, it looks like this denial of 3b/4b situations properly replace requirement to face bet>0 and not any other raise.

You could use not(tourney_hand_player_statistics.flg_f_3bet_opp or tourney_hand_player_statistics.flg_f_4bet_opp) instead of tourney_hand_player_statistics.amt_f_bet_facing > 0 but I don't understand why you would want to replace a shorter piece of code with a longer piece. In other words you can have two expressions count the exact same thing but one can be much longer than another depending on how the efficient the database schema has been used but if your more comfortable/familiar with a different piece of code you can use what you want.