Pure 3-bet and Fold to 3-bet stats?

[Dan_Jazz_Man]

Is it possible to create these? By a pure 3-bet I mean excluding situations when a player re-steals form the blinds, and
by a pure Fold to 3-bet I mean excluding situations when a player steals from CO and BU and folds to a re-steal from the blinds.

[kraada]

Yes, this shouldn't be too hard.

The columns you'll need are:

cnt_3bet_pure:

sum(if[ flg_p_3bet and not(flg_blind_def_opp), 1, 0])

This will sum up all times a player 3bet preflop that wasn't a blind defense situation.

cnt_3_bet_pure_opp:

sum(if[flg_p_3bet_opp and not(flg_blind_def_opp), 1, 0])

This will sum up all of the times a player could 3bet preflop that wasn't a blind defense situation.

cnt_3bet_pure_def_opp_fold:

sum(if[enum_p_3bet_action = 'F' and not(flg_steal_att) and flg_vpip, 1, 0])

This sums up all of the times someone folded to a 3bet when they hadn't stolen themselves but they'd put money in. If you want it to be where they've raised themselves, make it cnt_p_raise > 0 instead of flg_vpip.

cnt_3_bet_pure_def_opp:

sum(if[flg_p_3bet_def_opp and not(flg_steal_att) and flg_vpip, 1, 0])

This will sum up all of the times a player faced a 3bet when they weren't stealing and they put money in themselves. (See the above note if you want them to have done the raising).

Then create the stat as (times / opportunities) * 100 on the stats tab, and don't forget to fill in the Format tab.

[Dan_Jazz_Man]

Thanks, Kraada!

I will try to create these stats. Your post also made me think about duplicating and modifying the standard Fold to PF 3-bet stat
with cnt_p_raise to reflect the percentage of times a player RAISES and folds to a 3-bet, while the standard stat shows the
TOTAL percentage of folding to 3-bet (including times where a player hasn't VPIPed yet, but is faced with a 3-bet).

[kraada]

Yeah, that's been requested several times. I think that stat is already in the set in my Advanced Hud FAQ (located here). If it isn't, though, it will be in the update to that FAQ which I anticipate I'll have up later this week or early next week.

[Dan_Jazz_Man]

Actually, ignore my previous retarded post, I don't know what I was thinking. I'm going to create a "pure fold to 3-bet" stat, so I won't need the standard one anyway.

[kraada]

Have fun, if you need help with anything, you know how to find me.

[Dan_Jazz_Man]

I created the "pure" stats and after some initial testing the results are eye-opening. The difference between
"pure PF 3-bet" and "standard PF 3-bet" is around 0.5-1 % for most players (this is for FR), not a whole lot, but it is still
an indication of a tighter "pure 3-bet" range (and in FR 1 % can sometimes sway the preflop decision the other way around).
Now, the difference between "pure fold to PF 3-bet" and "standard fold to PF 3-bet" is HUGE, which makes me think that
the standard stat I've been using up to this point is pretty much worthless. For example, my "standard fold to 3B" was 82%,
so I thought I was raising and folding to 3-bets too much. My "pure fold to 3B" is 58%, so maybe I was CALLING a little bit
too much. Of course, a big part of it is that the standard stat uses total times we're faced with a 3-bet instead of only those times
we raised and are faced with a 3-bet. As you wrote in a previous post, the latter makes a lot more sense, that's why there are
requests to have the standard stat modified.

[kraada]

I know, and I'm aware of that, which is why in my personal HUD, I've built Fold to PF 3bet When Raised - which is the % of the time that you fold to a 3bet after you've raised.

Since I have the stat handy, I've gone ahead and attached it to this thread.

[Dan_Jazz_Man]

Thanks, Kraada! I'll take a look at your stat to make sure mine are correct, since they have many things in common.

Where should I post my "pure" stats for people to test? I believe they are very powerful, because we should differentiate
between:
1. "folding to 3-bet, when not stealing" and "folding to a re-steal from the blinds" and
2. "3-betting in the field" and "re-stealing from the blinds"

As for #1, we agreed that the standard stat is useless, everyone should use yours at least, but I took it one step further.
I mean, we can only get an average "raise and fold to 3B" percentage if we don't look at "stealing" and "not stealing" situations
separately. For example, if a player's "pure fold to 3-bet (when not stealing)" is 50%, his "fold to a re-steal" might be 85%, and then we can draw
conclusions about how he plays and what kind of hands he continues with in each case, based on the PFR and ATS respectively.
As for #2, again my tests show that most people (correctly) re-steal from the blinds quite a lot wider than the standard "3-bet" stat
indicates. For example, a player's "standard 3-bet" is 3 %, while his "Re-steal" might be 10 %, and his "pure 3-bet (not including re-steals)"
might be 1 %! Again, we get a much clearer view of how he plays, while with the standard stat showing 3 % we're unlikely to notice anything
unusual.

[kraada]

Hrm, I think you're right.

The repository is a good place to add stats, as I said earlier I'm working up a second edition of the Advanced HUD (which will be much prettier, and has position stats among other things). I already have fold to resteal as the color code on my attempt to steal stat, so I think you're right that I'd much rather have the "pure" fold to 3bet stat when raised (which I otherwise have as the color code on PFR), so I'll go make those changes.

The positional 3bet stats though negate the need for a pure 3bet stat, though as you can just see where he's making his 3bets from :)