I believe I have found a few bugs in stats

[ptrack mike]

OK, the only real bug here is with Full Tilt. If everyone folds to the bring-in player pre-4th streett then saw_fourth_n should = 0. Thanks for finding that. It will be fixed for the next patch.

The other things reported:
The vol_put_money_in_pot number is a whole-hand stat, not a pre-4th street stat. In the examples given the player called at least one bet later in the hand so vol_put_money_in_pot = 1.

The pot size is a small issue and is probably a result of how we calculate the pot size to get the rake. We moved away from parsing the numbers directly off the hand history for some sites because it wasn't always reliable. I'll see if I can make it more consistent, but the pot size isn't really an import stat as long as the rake and amounts won/loss for each player are correct.

[JoeTall]

The other things reported:
The vol_put_money_in_pot number is a whole-hand stat, not a pre-4th street stat. In the examples given the player called at least one bet later in the hand so vol_put_money_in_pot = 1.

Is this true for PTH and PTO as well?

[RustyBrooks]

So the concept of vol_put_money_in_pot as a "whole hand" stat does not seem consistent with what PT holdem (version 2) does. I have this holdem hand in PT holdem, for example, where I am the BB, it limps to me, and I call a bet on the turn. My vol_put_money_in_pot for this hand is 0.

select seat_number, screen_name, vol_put_money_in_pot from game
join game_players using (game_id)
join players on (players.player_id=game_players.real_player_id)
where ub_game_number='6390474053'

1 "swafjto" 0
2 "apple36" 1
3 "cjworks_com" 1
4 "unclbucc" 0
5 "Bobby Hutton" 1
6 "CHIPSDACLOWN" 0
7 "RobBar301" 1
8 "FONZY4" 1





FullTiltPoker Game #6390474053: Table Montagna - $0.50/$1 - Limit Hold'em - 18:48:38 ET - 2008/05/12
Seat 1: swafjto ($34.15)
Seat 2: apple36 ($7.85)
Seat 3: cjworks_com ($52.25)
Seat 4: unclbucc ($16.75)
Seat 5: Bobby Hutton ($12.85)
Seat 6: CHIPSDACLOWN ($5)
Seat 7: RobBar301 ($21.25)
Seat 8: FONZY4 ($22.90)
FONZY4 posts the small blind of $0.25
swafjto posts the big blind of $0.50
The button is in seat #7
*** HOLE CARDS ***
Dealt to swafjto [8s Tc]
apple36 calls $0.50
cjworks_com calls $0.50
unclbucc folds
Bobby Hutton calls $0.50
CHIPSDACLOWN folds
RobBar301 calls $0.50
FONZY4 calls $0.25
swafjto checks
*** FLOP *** [As 2d 2c]
FONZY4 checks
swafjto checks
apple36 checks
cjworks_com checks
Bobby Hutton checks
RobBar301 checks
*** TURN *** [As 2d 2c] [Td]
FONZY4 bets $1
swafjto has 15 seconds left to act
swafjto calls $1
apple36 calls $1
cjworks_com calls $1
Bobby Hutton folds
RobBar301 folds
*** RIVER *** [As 2d 2c Td] [8h]
FONZY4 checks
swafjto bets $1
apple36 has 15 seconds left to act
apple36 folds
cjworks_com folds
FONZY4 calls $1
*** SHOW DOWN ***
swafjto shows [8s Tc] two pair, Tens and Eights
FONZY4 mucks
swafjto wins the pot ($8.55) with two pair, Tens and Eights
*** SUMMARY ***
Total pot $9 | Rake $0.45
Board: [As 2d 2c Td 8h]
Seat 1: swafjto (big blind) showed [8s Tc] and won ($8.55) with two pair, Tens and Eights
Seat 2: apple36 folded on the River
Seat 3: cjworks_com folded on the River
Seat 4: unclbucc didn't bet (folded)
Seat 5: Bobby Hutton folded on the Turn
Seat 6: CHIPSDACLOWN didn't bet (folded)
Seat 7: RobBar301 (button) folded on the Turn
Seat 8: FONZY4 (small blind) mucked [9h Ts] - two pair, Tens and Twos

[ptrack mike]

It may not be the same as the flop game tracking programs (PTH and PTO) which treat this as a pre-flop stat. I honestly don't know why I did it, but I added code over 3 years ago to check all streets to see if anyone put money in voluntarily to set this stat. You'd have a decent argument to keep it consistent with the other two programs, though, but since Stud is a completely different poker game, as long this stat is consistent across poker sites in PTS and everyone knows what it means then maybe it's OK to leave it.

You also said:

This does not really affect vpip THAT much because it can only happen 1/8 times at most, and only if everyone limps. It affects vpip about 1% for me.

so it's doesn't seem to affect overall stats that much. I'll listen to opinions on this, though.

[RustyBrooks]

Well, yes, Stud and Flop games are different, but I think if you did a survey you'd find that most people think that vpip, for stud or holdem, is for the first betting round.

Stats-wise, yeah, it's not really a huge deal. It's just something I noticed that confused me.

It might be a considerable difference in hu or short games though...

[ptrack mike]

I've given it some thought and I'm going to change it to be consistent with PTH/PTO for the next patch. Given that you are the only person that noticed it in over 3 years it's likely that most people will even know it was every not the same anyway or that it is then slightly different.

[grindunumb]

The fact I never noticed this tells me I have not been analyzing my play sufficiently.