The stat is done outside of PT4 in a program I'm writing. I read all the river bets* from a player and the board from the DB (with an SQL query). Next, for each river bet, I compare the bettors equity to random cards** on that board, using the 2+2 hand evaluator under the hood. So if I have 100 riverbets from a player, I get 100 equities. Those are ordered and arranged to 10 bins (like 0-10%, 10-20%....90-100%). The more bets in a bin, the bigger value I give to the bin.These 10 values are normalized and arranged into a string (using digits 0-9), so in the pic you see the underlying string is something like 000111597. He basically never bluffs with missed stuff, maybe makes some merged range type value bets / bluffs but not enough to try to raise him off those.
Once the string is done, the program inserts (via SQL) the tags and the string to the players's notes. If a note does not exist it is created, if there's already a note, I insert to the top. For the actual data string I use a font I designed where digits 0..9 are represented by the different height bars. The colors are handled during the same process (and are mainly for fun).
Repeat for all the players in the db.
(To do: repeat for river raise, river 3!, maybe look at turn stuff...***)
* all the bets where I know his holecards. This may introduce a small bias, but since it's river and I mainly play FL, most bets get called anyway so I believe this should be fine-ish.
** random because I can't construct a range for the villain(s) in hand. Random should be fine to get this distribution, though... it's used to figure out his betting polarization, not exact value.
*** I can see some biasing problems here as there's one more street where the player can muck his weaker cards. Ideas appreciated.
I can publish the sql queries I used to put data into the notes if there's public interest and the officials here don't mind it... they seem to be a bit wary about using the db in unconventional ways
Using the notes for stuff like this was more simple in PT3, although the rich text formatting does give some liberties (once you figure out how it's used). The only side effect I've noticed so far (doing all this so far in a backup db obviously) is that the notes tooltip doesn't synchronize with the data I insert before I actually click save in the notes box or restart the replayer window... PT4 seems to do some caching there. This doesn't really matter as the tool is run "offline" anyway.
As for my C source code,that would take serious cleaning before publishing... it was built on the go, starting from fooling around with the hand evaluator.
Was this worth it to get a few river stats? From the game perspective, most certainly not. For academic purposes, certainly yes... had to relearn a lot of C and learn a lot of SQL to get here.
Also, if anyone knows how to change the expression below to return 'asomas', not 'as', i.e. finding the
first occurrence of 'om', I'd appreciate the help.
select substring('Thomasomas' from '%om#"%#"%' for '#')