Percent Chance of Raise at This Table?

[kingfisher]

Hi,

I'm trying to figure out how to create and display a PT3 statistic indicating the percentage likelihood that a hand will be raised preflop at this table, given the number of hands I've been seated at the table. So for example, if I've been seated at the table for 100 hands, and 65% of those hands have been raised preflop (by any player at the table), this statistic should display "65%".

I can piece this together in raw SQL but doing it within PT3 eludes me. Any suggestions would be appreciated!

-Chris

[WhiteRider]

Your own stats in the HUD are always for the current session/table, so if you build a custom stat to count hands which are raised and display it in a 'show on table for hero only' group it should work.

You could build the stat to count hands where you faced a raise or made a raise; e.g. a column like this: cnt_p_raised_hands =
sum(if[holdem_hand_player_statistics.flg_p_face_raise OR holdem_hand_player_statistics.cnt_p_raise>0, 1, 0])

If you divide that column by cnt_hands you'll get what you want.
Remember it will only work for hero, though - for other players it will cover all the hands they have played.

[kingfisher]

Hi, thanks for the response. Your solution seems pretty good, that's definitely along the lines of what I was looking for, but there's one problem. If we key off flg_p_face_raise OR cnt_p_raise > 0, we miss all the hands where (for example) I fold in early position, and a player raises behind me. flg_p_face_raise is false for me if I fold prior to the raise happening.

I'm really trying to get at the "true" raise frequency percentage for the entire table, so I can say, "hmm, 75% of the hands at this table are raised before the flop".

Is there any way we can infer this from the table's average PFR (on a shared statistics panel)? If the average PFR is say 20% and there are N players, then we can compute chances that at least one player raises as 80% to the Nth power, where N is the number of players (including me) at the table?

[kraada]

I can't think of a way at the moment to get the number of times a hand was raised if you folded before the raise was made.

With regards to your overall thought process, it depends on how nitpicky you want to be; there are a lot of situations in NLHE which change people's opening ranges. Most simplistically, what position they are in. You might get someone who raises UTG only 2% of hands but who raises 20% on the button. So if that player gets AJo UTG he is going to fold it 100% of the time, but if he got it on the button and he was folded to he would raise it 100% of the time. But when he gets it UTG, card removal concerns make it marginally less likely for players behind him to have a raisable hand. If you're going to try using positional statistics to infer the actual value you want, though, you're going to need to have something like 200 hand samples on everybody at the table - which is probably an unlikely occurrance. Probably using PFR in a table averages group is going to be your best overall chance and just figure that the EP and LP considerations average out. But it still won't be perfect. If the PFR of the table were something like 10, I'd figure that somewhere between 8 and 12% of the time a hand will be raised by someone else before I look at my cards. After looking at my cards, if my cards are better, it'll be towards the lower end of that scale, if they're worse it'll be higher. But I haven't done any research to back up my gut on this one.