Flat PokerEV Greenline and Raised Redline

[Troll_Inc]

I've started trying to examine portions of my Pt data where I have the dreaded flat or decreasing greenline with red/blue lines(ala what Roundtower showed a few days ago in the low content thread). For those who don't know PokerEV, the green line is basically just a average winrate line. And when the greenline runs below your redline (adjusted all-in winnings), then it basically means that you are losing money overall in hands that don't go to showdown.

What sort of leaks to check for during these periods?

My impression after these sessions was that I was playing a little "loose", i.e. I felt that I was pushing in marginal situations.

But I guess this data argues that I'm either calling too much on the flop (e.g.) and then releasing when I don't hit or getting bluffed out of pots. I don't think the latter is the case since, if anything, I call a little bit on the side of too much. If I have anysort of hand and someone bets 1/3 on the river and I can put them with correct frequency on a range I beat, I usually call.

I don't see any correlation to changes in VPIP/pfr. I am about 30-33/10-12 for both times when greenline increases or diverges/decreases from the redline.

If I look at the PokerEV tab with the histogram, at all streets my green(+)/red(-) ratio is positive.

Where to look next? Has anyone already investigated this sort of thing? It seems that some of the things I can think of for smaller hand sections will be limited by distribution. (passive play on different streets

[abscr]

I have been having the same problem and I'm trying to adjust my game.

I think some of my leaks are playing just a little too loose when someone raises. I think I was calling with too many hands like T975ss or J956ds, being too optimistic about hitting a huge hand to stack someone with AAKQ when I usually just fold on the flop, so I'm losing 1-2ptbb a ton while not stacking them enough to make up for it (if at all . . .).

Another leak I'm trying to plug is c-betting too much into calling stations. I'll c-bet, and then check it down and usually fold to their river bet when I have basically nothing. A corollary to that is not raising trash in the CO and button based on the idea that I can put money in the pot pf just to steal it on the flop with a c-bet against 1-2 opponents (this sometimes works, but based on that, I probably do it way too often and try to force it, making it way -EV).

I understand you probably know this stuff and don't (and I'm probably just a bad player that needs to plug these), but I figured I'd put some ideas out there for brain storming.

[cmyr]

I hope to come back to this topic in a few hours, but I'll point out that I have spoken privately with a few players I respect, and none of us have, over any large sample-size, kept our non-showdown winnings positive. I have a few ideas for why this might be (dealing with folds in multiway pots, primarily) but I haven't really drawn any conclusions.


I would be interested in seeing if two players playing heads up create graphs that correspond, i.e. if one player is losing in non-showdown hands if another player is winning a corresponding amount.


If someone with PokerEV installed wants to play some low (or not low, whatever) stakes hu on stars or something today drop me a PM, and we can see how it looks.

[abscr]

I asked in the LC thread but since this thread exists, would people be willing to post their non showdown ptbb/100? I'm at work so I can't now, but I can later. I'd be mostly interested in 6max, but whatever is cool.

[Troll_Inc]

[ QUOTE ]
I hope to come back to this topic in a few hours, but I'll point out that I have spoken privately with a few players I respect, and none of us have, over any large sample-size, kept our non-showdown winnings positive. I have a few ideas for why this might be (dealing with folds in multiway pots, primarily) but I haven't really drawn any conclusions.


[/ QUOTE ]

Interesting, so I guess you are saying that playing a winning style of PLO* (6max or full) will end up with losing money at non-showdowns and ergo money is one by forcing/creating/being part of showdowns. And it is in those showdowns where you will win.

Also, one thing that is somewhat confusing about PokerEV is when it "does it's analysis", it of course is only analyzing the hands that went to showdown. But then when overall winnings are posted, they of course include all hands.

If someone hasn't taken you up on your experiment by the time I get home tonight, I'd play Stars HU with you. Another thing would be to later try it at a 6max game, which might be necessary to test out the idea of multi-pots. You could even pretty easily test different styles of play, which might more quickly identify leaks.


*I wonder if this is true for NL, or something PLO specific?

[jbird]

I deleted PokerEV from my computer but I may have to reinstall it -- this is a really interesting thread, but your summary of CMYR's post sounds about on the money Troll.

Oh, and:


[ QUOTE ]
If someone hasn't taken you up on your experiment by the time I get home tonight, I'd play Stars HU with you.

[/ QUOTE ]

ZOMGAIMMELETMEGETSOMEPOPCORNANDBEER

[CrushinFelt]

I definitely had a green line positive when I played about 10k hands a few months ago. I think the reason for that is primarly that I was playing at weak-tight tables as much as possible, and I made a lot of money without going to showdown.

It could be a matter of table selection. Before I decided to hunt for these tables I was playing lower stakes which were loose-passive most of the time which definitely led to a terrible green line. I think a "standard", solid, LAG style will inherently lead to negative $ when not showing down.

I think it is most applicable when there are more people at the table and more multi-way pots. Playing HU, though a good way to see if it is calculating it properly, will not have the same results (ie negative winnings when not showing down) as playing 6-max or FR.

[cmyr]

i'll caution that afaik losing money in non-showdown hands is only a feature of a few particular players styles of winning poker, styles which are pretty similar. That said, they're also the sort of styles i would expect to be winning money in non showdown hands.


I just ran my hands since august, and I'm running at -3bb/100 in nonSD pots, which seems extreme... I def. didn't think it was that bad.

[Troll_Inc]

There is something fishy about RT's graph:

http://forumserver.twoplustwo.com/sh...age=0&vc=1

The blue line seems to mirror the green line. As I said before, if someone folds hands, then there will be fewer hands that make up the data that is grpahed in the blue line. So they shouldn't have the same detail. (also of course if everyhand went to showdown the two lines should be equal?)

If he folds a lot of hands, then there should be some decreases in the green line, while the blue line stays unchanged.

When I get home, I'll post and look more closely at my graph. But if I recall, I was analyzing this months hands (~4,000) and only about 1/10th of them went to showdown. So in that case, there really should be no way that the blue line should mirror the green line.

[cmyr]

i think it's the other way around; the green line (total winnings) echoes the blue line, which is a subset of total winnings. It would be easy to check (i might in a second) but i'm willing to bet the green line has more data points then the blue line does.


edit: yeah, it does.