Try these PT filters

[bennyhana]

I guess I suck really bad in the blinds compared to you guys. When I get to 100k again, I'm gonno do this again.

[Aaron W.]

[ QUOTE ]
I did it but I still don't get it. [img]/images/graemlins/confused.gif[/img]

[/ QUOTE ]

The idea is that I'm trying to isolate a specific situation, namely 3-handed limped pots when you're in the BB (so you see the flop for free).

PT doesn't have a filter that will do that for you, so to get information, you've got to do a little work.

The baseline filter: This just takes a look at all of the situations where it's 3-handed to the flop just see how frequent these situations are in your database.

Filter #1: Looking at just the BB hands.
Filter #2: Looking at the BB hands where you called a late position raise.

The computations are then all about removing those defense hands from the results. Unfortunately, it doesn't get rid of the times you call against an EP raise, which is why VPIP > 0. If you could get only limped pots where you check your option, then VPIP would be 0.

[Bona]

Yes. I actualy did get that much of it. What I don't get is your value expectation.

In other words what kinds of changes in default play do you suspect might lead to better long term results? Why are you isolating 3 handed situations as opposed to any number of limpers? Do you think 3 is different from four is different from 5 etc?

Or do you intend to write up these conclusions after looking at a broad base?

[Aaron W.]

[ QUOTE ]
In other words what kinds of changes in default play do you suspect might lead to better long term results? Why are you isolating 3 handed situations as opposed to any number of limpers? Do you think 3 is different from four is different from 5 etc?

[/ QUOTE ]

I'm isolating 3 because in 3-handed limped pots, I'm almost always betting the flop from BB after SB checks to me. Or at least, I tell myself that I should be doing that. I would like to do a more in depth study on that particular situation (and perhaps the similar extension of betting from SB whenever you complete it in a 3-way pot), but such a study will require more time than just playing around with filters. This was just sort of a preliminary feeler to see if there's a lot of variation in these numbers or if it's pretty flat. If there's a lot of variation, that might indicate that the value of various strategies varies quite a bit. But if everyone looks like they're in the same ballpark, then it's not going to be easy to say anything definitive.

I would not autobet hands in 4 or 5 handed pots, which is the major distinction.

[CrMenace]

This will really help you out:

3,300 (LOL) hands since my hard drive crashed

Filter1: 72 hands @ -0.48BB/100, 25VPIP
Filter2: 1 hand @ -1BB/100, 100VPIP

Computation1: -34.56BB with 18 VPIP hands
Computation2: -1BB with 1 VPIP hand

Computation1 - Computation2: -34.56 -(-1) = -33.566BB @ 17 hands

Big Blind = 72 - 1 = 71

Big Blind: -33.56BB + (0.5BB/hand)*71= 1.94bb

Big Blind: 1.94BB/71 = 0.0273BB/hand @ 17/71 = 23.9VPIP

[shuinthehouse]

Aaron - I've learned a lot from your posts so I'd like to give you my data if it will help your analysis even though I'm not sure what I can learn from it. I use pokeroffice (surprised how few people use v. pokertracker, I was relatively new when I bought it, seemed better to buy one program than 2 (PAHUD for PT)). In this case, pokeroffice is better, there is a filter option "checked as BB (free)".

Baseline data: 35k hands .5/1 FR, overall -26BB loser (but + 3.6 BB/100 in 4500 hands .25/.50 [img]/images/graemlins/tongue.gif[/img]). 4153 hands in BB, -$826 ~= -20BB/100 (in the SB I am 14.35BB/100 loser overall).

Filtered for “checked as BB (free)”:

3 players to flop: 490 hands -23BB/100 (worse than average)
4 players to flop: 340 hands -15BB/100
5 players to flop: 149 hands +10BB/100
6 players to flop 40 hands +64BB/100
7-10 players only 3 hands.

[RemyXO]

Here is mine trackermathdonkfest. Thanks for asking. I puked.

40K hands

Filter #1: 1237 hands @ -0.36BB/hand, 17.78 VP$IP
Filter #2: 41 hands @ -1.09BB/hand, 100 VP$IP

Computation #1: -79BB with 220 VP$IP hands
Computation #2: -45BB with 41 VP$IP hands

Computation #1 - Computation #2 = -34BB in 179 VP$IP hands

Big Blind: 1196 hands

[b]Big blind: -34 BB + (.5 BB/hand)*(1196 hands) = 564 BB

Big blind: 564 BB/1196 hands = 0.471 BB/hand @ 179/1196 = 14.97 VPIP

[NIX]

Here's mine Aaron: 75k hands at .5/1 FR/6max, 1/2 FR/6max, and 2/4 FR, probably in a 30/45k FR/6max ratio. I can redo them to split it up if you'd prefer.

Filter 1: 2794 hands at -0.08BB/100, 18.93 VPIP
Filter 2: 180 hands at -0.44BB/100, 100VPIP

Computation #1: -219.92BB with 529 VPIP hands
Computation #2: -79.2BB with 180 VPIP hands

Computation #1 - Computatation #2: -140.72BB in 349 VPIP hands

Big blind: 2794 - 180 = 2614

Big blind: -140.72BB + (.5BB/hand) * (2614 hands) = 1166.28BB

Big blind: 1166.28BB/2614hands = 0.446BB/hand @ 349/2614 = 13.35VPIP


Our expectations look pretty similar if I understand this right, although it looks like I'm a huge nit. And FWIW, I tell myself I should be close to autobetting in these spots, but just about every flop I see scares me into checking with air.

[Bona]

[ QUOTE ]


Our expectations look pretty similar if I understand this right, although it looks like I'm a huge nit. And FWIW, I tell myself I should be close to autobetting in these spots, but just about every flop I see scares me into checking with air.

[/ QUOTE ]

I'm not calling you out here Nix, as a matter of fact we agree. Just to try and advance the disourse my obseravtion is the only 2 possible reasons for a raise is for value, that is we think our hand is better than either limpers range. And quite a bit better since there are 2 of them.

OR we have reads on the players that they would both fold marginal hands on the flop if we raised here and again on the flop. Thats not gonna happen a lot at the low stakes I play.

So I'm not seeing "near autobetting" here in my future.

[Aaron W.]

[ QUOTE ]
I'm not calling you out here Nix, as a matter of fact we agree. Just to try and advance the disourse my obseravtion is the only 2 possible reasons for a raise is for value, that is we think our hand is better than either limpers range. And quite a bit better since there are 2 of them.

OR we have reads on the players that they would both fold marginal hands on the flop if we raised here and again on the flop. Thats not gonna happen a lot at the low stakes I play.

[/ QUOTE ]

I might be misreading this, but I think you're talking about something different. Here's the situation I'm considering:

Somebody open-limps from CO, SB completes, and we check our two cards from BB. The flop is K49 and SB checks. My assertion is that we should be betting and it doesn't matter a whole low what our cards are.

If the action comes down any differently (CO open-raises preflop, SB donks the flop, or we raise preflop), then we are no longer in the situation under consideration.