Really phenomenal post Mpethy, thanks. Not sure which I liked better, the OP or the responses to the first few comments / questions.

More to come I hope

read the first paragraph only. looks golden. will read the rest later. nice job matt...

Would pay money for this information, wait....

superb COTW, so much win in it.

That's a good idea. It will eliminate all of the situations you describe, and leave you mostly with the hands where you decided not to c-bet, and then you can proceed with the hand history review to see if you are leaking.

Great post. Wrt excessive 3bet calling, what would you say is a reasonable 3bet call %? And how does this change in 6max?

For those of us still on PT2 (ie, just me), can you explain what the green/blue/red line are supposed to signify? And are the graphs produced with PT3/HEM only? Didn't see this in sticky...

Too cheap/busy (lazy) to upgrade...

red line is money won w/o showdown, blue is money won w/ showdown, green is total money won (red + blue line)

Your stats are somewhat unusual in that you c-bet 69% on the flop and 60% on the turn. If you look at the graphs you made of the flop c-bet (graph 1) and turn c-bet (graph 2) it tells a coherent story about your style of play.

Your red line on the flop is saying--I c-bet and get a lot of folds. But because I am c-betting a wide range, I usually don't have a hand, and this wide range that I am betting doesn't do so well at showdown, though I am up a bit.

When you move on to graph two, it is saying, "here we are one step closer to showdown, and I am still usually firing. I bet most of the hands I bet on the flop. I'm REALLY hoping you fold to this turn barrel, sir, because that is where all my profit comes from. If you call, this range is too weak to win at showdown, and I expect to take a small average loss at showdown."

Your high double barrel frequency lowers the average strength of the hand that you take to showdown and raises the average strength of the hand your opponent takes to showdown. In a case such as yours, I would expect for you to have a low W$SD%, maybe even below 50%.

The problem that you are encountering in your existing break even stretch (graph 3) is that your double barrels are losing at showdown (which is part of the reason that your blue line is flat toward the end of the graph) COMBINED with the fact that your red line has nose-dived for some reason.

As graphs 1 and 2 show, you can tolerate a flat blue line in your c-betting game and show an overall profit. Graph 3 shows that you can tolerate a flat red line when you are winning on your blue line.

At the moment, both of them are nose diving.

Unfortunately, this just describes the mechanism of your existing b/e stretch, it does not expose any potential leaks.

Clearly, from graph 3, both your red line and blue line have changed dramatically in the past 35k hands.

All I can say at this point is that c-betting is not the source of your recent red line nose dive. That means you need to be looking at all the other red line spots to find out where the losses are concentrated.

The blue line being flat over the last 35k hands IS related to your high turn c-bet%, as graph 2 shows. But like I said, your game can tolerate a flat blue line, provided that you improve your red line back to flatness. And this could easily be variance; a 35k hand stretch in which you happen to be double barreling at villains with the top of their ranges. If it is not variance, and is a recent change in your game, then I would suggest a hand review of the air/marginal holdings you double barreled to see if there are many that you should have checked instead.

Your results could be the result of variance. But your graphs make it look like you muddled up your game recently. It looks like your game changed at hand 105k from a solid TAg game to a failing LAg game.

Check the other red line spots in the OP and see if you can find the red line losses.

best post ever

Outstanding post Mpethy. I'm always amazed when I talk to someone and they don't know how to use HEM and/or don't bother spending a couple hours a month running some filters to review their play.

I'm interested in hearing your opinions on how sample size affects your stats. For example, in my review last month I ran some of the filters you mentioned when I discovered I was getting destroyed in my blind play this year. The problem was that there didn't seem to be a clear distinction in which hands were losers and which ones were winners when I looked at w/rs. KJs may be a big loser for me, but KTs was a winner. 66 was a loser but 55 was very profitable. I had to review the actual hands and try to adjust the wrs for luck factors. Thoughts?

You're not going to like my answer.

I was doing a database review with a winning $400 6 max grinder the other day. We were talking a lot of strat and I was learning more than I was teaching. When we got to his calls of 3 bets I decided to float a trial balloon I have been thinking about for several months now. I said:

The problem most people have calling 3 bets is this: Other than slow-played premiums, there simply is not any range that you can call a 3 bet with and profit from it by playing fit or fold.

He instantly agreed with me, and it got us talking about how to play these spots, and, more importantly, whether to play these spots.

The bottom line here is that calling 3 bets is a play where you have to have and use your skill edge to make a profit. This means that you have to win a significant fraction of the hands you call with without making a hand.

I am not a poker theoretician. I am a poker empiricist. I'm not going to sit here and say, "here is an optimal % to call 3 bets with, and here is an optimal range to be using."

As an empiricist, I am going to say: If you have a big sample of calls of 3 bets and your losses calling are smaller than they would be if you folded your preflop raise, keep doing what you are doing, and maybe loosen up your calling standards to see if you can change some -3bb folds to a 3 bet into -1.5bb losses by calling some more and winning a few of them. If your losses exceed the loss you would have taken by folding to the 3 bet, then man up, admit you don't have a sufficient skill edge to play that spot profitably by calling, and move those hands into the 4 bet or fold category.

That's not the end of the process, though. Every leak that gets plugged by folding hands you are currently playing should be considered as a temporary fix that you implement while you are spending your poker homework time studying how to play that spot profitably. So if you are calling 3 bets from MP with a range of mid pockets and broadway and losing more than if you had folded to the 3 bet, start folding them while you play detective and try to figure out what you are doing wrong. Once you think you have that figured out, add them back in to the extent that your study suggests it is correct to do so.

Most people call too many 3 bets because they are looking at their cards. They call based on the following thought and emotional process:

"Oh, that dirty so-and-so just 3 bet me AGAIN. I have to make a stand. I have KQo. That's a good hand. But I can't 4 bet it because I am dominated by his shoving range. OK, I will call."

Blech.

You need a plan to exploit the 3 bettor.

"Oh, that dirty so-and-so just 3 bet me again. That puts his 3 bet % at 9%. His flop c-bet% is 72% and his turn c-bet% is 36%. If I call and float any flop, he'll turn his hand face up on the turn <looks at card> I have some back up equity--call."

so much information to process.. but thanx for the great COTW!

[x]OMGItsmpethybridge

Sorry, I missed this question earlier. It's a good one.

This may very well be related to your sample size. But a lot of it is that the quantity/quality line is not so bright as I maybe suggested. It is definitely a continuum, not a bright line. A flop c-bet of 70% is sort of borderline quantity/quality. The last guy I saw with a clear quantity streak was c-betting 82% of flops, lol.

Similarly, your high turn c-bet stat is a quality stat, not a quantity stat.

If I had to analyze what is going on, I would say this: based on your stats, I would guess that you fall more into the quantity category in some situations (attempts to steal) and more in the quality camp in others (EP raising).

This is actually an excellent way to play most stakes, and can be hugely effective.

Matt,
One thing I have been thinking about is how multiway pots can drammatically affect non-showdown winnings. Seems like the amount we are risking vs. reward goes up but the success rate goes down, but its hidden from the filters. Off the top of my head with no calculator.

If we always bet 2/3 PSB for a cbets and have a success rate of 67% we should show a nice profit. But if 25% of those times are multiway were we only have a 25% success rate, then we could be about BE with our overall flop cbets.

Anythoughts?

P.S.

Quick comment on the 3bet analysis, folks seem way to eager to play 3bet pots OOP both with and without initiative.

Ok my turn! My flop cbet % over this large sample is 69 and my turn cbet is 59. Does this make me a quantity player? I thought of myself as more of a quality player because it feels like I c/f a lot of flops.

PFR + Cbet flop = true:


Cbet turn = true:


And overall:


Also, I found that I am losing slightly when 3betting light out of the blinds, but its still a little bit better than -1bb per hand. I just find that nobody folds once they've flatted a 3bet because their continuance range is so tight, or they were only calling you with the intention of floating/shipping a lot of flops in the first place.

Just a few disjointed thoughts in no particular order, and with no conclusion drawn:

1. I heard a quote from one of the top online coaches, I don't recall which, who said, "you can make a nice living at uFR and SSFR just betting to collect the dead money in multi-way pots." This quote stuck with me, because at the time i heard it, my c-bet frequency in mw pots with air was close to 0.

2. I'm going to try some workarounds to see if I can isolate the multi way pots I raise. Right now I am thinking of PFR = true, faced 3 bet = false, size of pot on flop is 8bb and see if that works. If you get to it before I do, let me know if it works (I have a db analysis staring now).

Excellent coverage of this topic. I'm usually not much of a stats guy but I found this well written and easy to follow. Thanks for taking the time to write it up.

Thank you very much for your indepth analysis. The nosedive over the last 35k hands is due to a long breakeven stretch due to variance and then I spewed badly in two sessions, which I never do. Frustration got the better of me. My mental part of the game is what I have always been most proud of.

I have been looking at lowering my cbet recently particularly against good regs as they for good reason can read the higher cbet% as a weakness. Do you see a lower cbet% than TheBowlBoy and mine as necessary to be successful at 100nl and 200nl?

Thank you again.

Please disregard TheBowlBoy's post, he doesn't need any help.


My W$SD% is 51.6

Can't you filter for "Hands with # - # of players seeing the flop?" -- It's an option in general filters on pt3...

Your graphs were pretty. No problems there.

I find your second paragraph difficult to fathom for NL $50. My recollection from $50 was that people were folding to c-bets a ton. You might want to look at some of your 3 bet decisions to make sure that you are making the correct decision preflop.

Oh, and don't get hung up on the quantity/quality thing. a. it is descriptive, mostly, and b. as I said in a previous post, it is a continuum, not an either/or. You are a quality guy, but definitely at the aggro end of the quality spectrum bleeding into quantity.

No, definitely not. The whole range of flop and turn c-bet stats can be effective at $100 and $200.

This is an important point I want to make over and over: No set of stats (within a very broad range of playing styles) is more likely to produce winning results than other stats. There is no "right way," to play and there is no "optimal," stat for any given situation (although a few have very narrow acceptable parameters).

The important thing is that you be making good decisions based on the available evidence. So when you are holding JJ and the flop comes down A98 and the guy who cold called you from the button is a 33/7 with a 1.0 AF and a fold to c-bet of 38%, you need to make a good decision.

In this spot, against this player, I could probably convince an open-minded listener that check/folding is right and I could probably convince that same listener that betting is right. It's a marginal spot. Whether your tendency is to bet this flop or c/f this flop doesn't really matter very much in the long run. None of us really can prove that betting is superior to check/folding here, because the theoretical "best," play is going to be best by only a tiny amount, and will depend COMPLETELY on things we cannot know for certain.

What matters is that you:

a. recognize this as a marginal spot.
b. Are able to articulate the reasons betting might be correct.
c. Are able to recognize the reasons you might consider check/folding.
d. At the table, you debate c/f vs. bet here, and aren't thinking about something like a check/raise.

One thing I can guarantee you is that anybody who can make the argument for betting or check/folding here understands poker well enough to be making a profit from the game. Which argument you prefer is, in my opinion, more descriptive of your tendencies than anything else. Maybe BalugaWhale could come in here and convince me that one of these is theoretically clearly better than the other (as I said, I am an empiricist, not a theorist). But I'm not sure of that.

The fact of the matter is, to most of us, making a small mistake in a marginal spot is no big deal for the simple reason that we are occasionally making huge mistakes in clearly incorrect spots. That doesn't mean you shouldn't work on getting the marginal spots as right as possible. It just means that a broad range of styles are "good enough," to beat all levels at which I have experience playing or reviewing databases, and to do so for a win rate any of us would be pleased to have. And, for the record, I have done a non-trivial number of reviews for players at $5/10.

Mythbridge, thank you for your efforts and another excellent COW. It, in many ways, was very relieving on following your analysis of my database.

<bbv></bbv>

The gem that I dug up from the discussion was from about Sammy where you describe gaining a skill advantage and opening up your range. Of course, you would need to close up your range when you move up. The question I have is: are you making individual hand changes (i.e. KJ) or are you making changes to a range of hands like 89s+? If you are making change for specific hands it would take a number of hands just to witness the hand and playing it. How many hands are you using to make your evaluation to determine it is a success verses variance?

great cotw mpethy.

Have spotted a couple things I need to think about with my game just by following a few simple steps, which is always nice.

You are single handly making the games tougher.