Micro Stakes Online NL Poker Forums

Just an update...

What I'm dealing with in producing a full analysis of the EV of this flop spot is recursive EV. Yes I know the lingo now after catching up on my "Running The Streets" watching.

It's really time consuming but has a number of benefits so I'm continuing on with it.

Could always post it in parts ?

Yeah it's kinda awkward because I have to work the tree through before I can finally examine the flop spot itself. It's really counterintuitive. More like a chess problem actually. Shorter than a chess problem, but more multidimensional.

I wouldn't go overboard because it's only 10NL and people sometimes do random stuff. The most important thing is to get their ranges in the right ballpark - so I'd look to filter the two players and look at their ranges in various positions and make an educated guess based on their game. If it's too small a hand sample, I'd look for some other villains that are representative in type.

I kinda kept track on this spot in my own game and found that sometimes I bet and sometimes I check, etc. depending on game flow and my impression of the villain.I was actually pleasantly surprised that its easier to get to showdown taking a bet line even when your still getting called because people are not as aggressive without a good hand and not good hand readers. So I think that betting is actually better now

I'm actually more interested in the process itself right now. It hasn't been a straight line over the last couple of days I've had to watch a couple of DC videos that Shuttle recommended, and get used to doing things differently. But I feel it's opening up an avenue of study that will be very beneficial to my game if I can find a way to do it efficiently.

Which videos did he recommend if I may ask ?

Running the streets ep 2 and 3 (his own series obviously! LOL) But it's true that the stuff in both vids is extremely relevant to this sort of spot and the kind of analysis I'm attempting.

Bet for value/protection and to protect against being bluffed off best hand on turn from oop player depending what type of villain they are. UTG looks to have overs, pot is small, make a 55 cent bet and take it down.

Bet for value/protection and to protect against being bluffed off best hand on turn from oop player depending what type of villain they are. UTG looks to have overs, pot is small, make a 55 cent bet and take it down.

What do you reckon the EV of a 55c bet is, more or less?

I'm making great progress with this now that I have CRev to work with. It's a really sucky spot to have as an intro to CRev though.

I can't write up a full analysis of the bet taking turns and rivers into consideration because in total I'm building a decision tree for 4 different turn card types vs each and both villains (12 turn spots) and then four river types on each of the four turn types vs each and both villains (48 turn spots)...

Working through those spots in the decision tree is quite intuitive when you get your momentum up... but writing it all down with ranges and such to post on DC would end up in a 50 page essay.


In the meantime, I WILL post my thoughts on the player in the CO, same as I did for the UTG raiser. I've had this in a txt doc for a while now and have been using it with Flopzilla etc.

The Loose-passive caller

To be honest I find this a lot harder than assigning a range to the original raiser for the obvious reason that a weak-tight player raising UTG has defined his range much more than a loose-passive player who flats that raise.

I would first like to think briefly about the sorts of hands he'd 3bet pre.

I don't think it's much. Like AA-KK most of the time and QQ like half the time. This kind of player is not thinking about getting it in pre. That's not how they want to spend a buy-in. They want to see a flop before committing a big chunk of their stack. I'm going to exclude all AA hands, leave in 3 combos of KK and 3 combos of QQ.

Everything else is in there with respect to having not been 3bet pre. Including all AK. “The Anna Kournakova looks good but never wins”.

So I happen to know that this guy's stats settled around 40VPIP, but without that knowledge I just started adding hands to see what I ended up with.

22-JJ, QQ50%, KK50%
All suited and offsuit broadway
All non-broadway suited Aces and Kings
A few middlingly speculative offsuit hands A8o-A9o, K9o, Q9o, and J9o
All non-broadway suited connectors, 1-gappers, and 2-gappers
All offsuit connectors



So those are the broad strokes for preflop ranges for my villains.

And next I'm gonna post my ideas of the two villains' flop ranges and tendencies.

(to DC admins, sorry for the string messages... I feel it's better to keep separate chunks of info so I'm using different messages to delineate.)

Now before we reach our decision point we can do some narrowing of these ranges based on what we'd expect them to bet on the flop.

The weak tight raiser

Overpairs I expect him to lead with half the time, and be looking to check/raise them the rest of the time.
I don't see him check/calling an overpair here. So half his overpair hands disappear.

Overpair combos AA-JJ -- 24
12 disappear
12 go into his check/raising range


I think he'd lead sets some, check/raise some, check/call some in an attempt to be deceptive.
That considered I'm going to give one combo each of TT and 77 to each of his respective flop check/calling and check/raising ranges. The rest are gone.

Set combos TT, 77 -- 6
2 disappear
2 go into check/raise range
2 go into check/call range


I think he'd lead AT almost always. It's a tiny part of the range anyway so I won't bother leaving any in.

AT combos 4
4 disappear

I think he would check/call 99-88 half the time and lead them half the time. I'll keep the case 88s and half the 99s combos in for a check/call.

99-88 combos -- 7
3 disappear
4 go into check/call range

I think he check/folds any pocket pair hands lower than 77 that he might have had up to that point.

66-55 combos -- 12
12 go into check/folding range

I think his overcards hands therefore make up the bulk of his check/calling range.
A few of those over card combos have flush draws too… but I think a weak tight player won't push the combo draw with something like a check raise so all those stay in.

Broadway overcards combos -- 68
68 go into check/call range

That's everything for the UTG raiser. Effectively we're looking at quite a weak range from this point. In summary, if we bet we'll expect to see a check/raise occasionally and a few check/folds, but mostly check/calls with hands we currently have drawing to 6 outs. His range is 100 combos (not deliberate). 12 combos c/f, 74 combos c/c, 14 combos c/r.

The Loose-passive caller

What do we expect him to do with different hands in his range when the PFR checks?

Well passive or not I would say he'd bet what overpairs he has at this point almost always.
He's going to be very happy with his hand.
He's not thinking about my range or anything... just likes his hand vs the board and is going to bet. I do still expect to be wrong about that occasionally given the generally random way people at 10NL can play, but it's a very small part of the range so I won't lose too much sleep over it.

KK-JJ combos -- 12
12 disappear

I'm not certain how this guy would play a set because loose passive players at 10NL play sets almost every conceivable way they can be played... they love the old check/call flop check/raise turn thing... but some of them bet them... some check/raise flop... 2 diamonds might make that more likely here... in any case... he has flopped a set almost exactly 2% of the time. If I allow for some betting-out and some check-raising...
lets let him keep half of his set combos for the old flop-c/c - turn-c/r move.
And 25% of them can go into his flop check/raising range. You can see here I've started dealing in fractions of combos. That's fine because underneath the combos we're really just talking about the percentages anyway.

TT, 77, 22 -- 9 combos
2.25 disappear
2.25 go into flop check/raise range
4.5 go into flop check/call range (with the intention of check/raising turn)


As a post script to the sets he flops 2 pair with his T7s hand which only has 2 combos anyway for a paltry 0.46% of his range...so I'm going to leave both of those in his range just so I can't be accused of taking liberties though I do think he'd bet them sometimes I really don't know so might as well leave them in.
There are 2 combos. Lets give 1 to his c/c - turn c/r range and 1 to his flop c/r range.
T7s -- 2 combos
1 goes into flop check/raise range
1 goes into flop check/call – turn check/raise range


I'm going to keep all his top pair hands in for a check/call, all the way from TPTK down, as well as 99 and the one combo of 88. This is due to his passivity.
AT-JT, T9, T8s, 99-88 -- 69 combos
69 go into flop check/call range


As for his middle pair hands, I'm going to deal only with middle pair hands that do NOT also have a flush draw. I deal with flush draws later and I expect all of those to call with or without middle pair.
So
1) - I expect A7(:heart)() and K7(:heart)() to call a bet always. (4 combos)
2) - I expect middle pair with a 9 or an 8 to call half the time (maybe less if I bet bigger) and fold the rest of the time. So that's 87o (5 combos) 87(:heart), and 97(:heart)() 8 combos total.
3) - I expect middle pair with an undercard to the 7 to call only about 25% of the time and fold the rest of the time. So that's 9 combos of 76o, 76( )(), 75( )(), and 74( )() 12 combos total.
A7( )(), K7( )(), 87o, 87(:heart), 97( )(), 76o, 76( )(), 75( )(), and 74( )() -- 24 combos total
11 go into flop check/call range
13 go into flop check/fold range



I think he'll check/fold bottom pair and any pocket pairs from 66-33.
33-66, A2s-K2s, 32, 42s, 52s -- 48 combos
48 go into flop check/fold range


That leaves hands that haven't flopped a pair.

Of those there are 2-overcard hands that I expect him to call a bet with. (6 with flush draws)
AK-AJ, KQ-KJ, QJ -- 96 combos
96 go into flop check/call range

There are additional flush draws (not already included in the 2 over cards hands and including the pair+flush draw of A7s-K7s). They're in for a c/c
A9s-J9s, A7s-A3s, K7s-K3s, 97s-96s, 76s-43s, 75s-53s, 74-63s -- 25 combos
25 go into check/calling range


There are OESDs… they're in for a c/c
98 -- 8 combos
8 go into check/calling range

He also has some combos of J8 and J9 that are not covered elsewhere… they're a gutshot + over card combo that I don't think he calls always… but sometimes… I'm going to include about half of them.
J9o, J9s (not diamonds), J8s (not because I have the 8 ) -- 17 combos
8 go into check/calling range
9 go into check/folding range



And I expect him to fold all the rest of his hands that didn't hit.
I won't name the hands but they're rag suited aces of non diamond suits, and lots of those small suited connectors/gappers and offsuit connectors etc.
[i]Shite -- 133 combos
133 go into check/folding range.


That's everything for the Caller. About 429 combos total (428.75) Some of his check/calling range is stronger than the UTG villain's range at this point because he has top pair hands in. So 203 combos or nearly half of his range I expect to fold to a bet here. I almost never expect to see a check/raise, and of his check/calling range, about 75 combos have me drawing to 2 outs, which is interesting.

So that's an intensely detailed breakdown of my 2 villain's ranges and what I expect to see them do with different hands if I am to bet the flop. I will have to look at the turn, but next we need to figure out what kind of turn cards there are... how they hit/miss the villains' ranges and what we will have to do if the villains do certain things.


I'm continuing with the work in CRev, but it's been hard to get time to sit down and do a solid session working on it. The TV network where I work has the broadcast rights for the Olympics and we're busy little beavers prepping.

I've discovered quite a lot along the way while doing this and I have some thoughts on the general process as well as what I expect a learning player to be able to gain from this sort of digging... and what sorts of things to look for.

Since when do 10nl players make analysis like this? I give credit, you keep this up you will get out of 10nl fast

Cheers!



LOL there's no money at 10NL, everybody's solid!

Great analysis, this is definitely a good way to improve your game.

Working through those spots in the decision tree is quite intuitive when you get your momentum up... but writing it all down with ranges and such to post on DC would end up in a 50 page essay.

I'd suggest writing it up for your own reference, the process of doing so will be greatly beneficial.

Great analysis, this is definitely a good way to improve your game.

I'd suggest writing it up for your own reference, the process of doing so will be greatly beneficial.

Cheers dude,

I've got what I've done so far, because I did a lot of it by hand using other tools before my license for CRev came through... I'm actually doing some other things like bringing percentages out of CRev into excel to graph things a certain way so I can see things more visually...

I also have quite a lot of thoughts on it that I'll be definitely chucking up here when I'm done... or even maybe making a short video on so I can show what I mean.

Oh wait wait...

@shuttle, I have a question for you if you're still around,

I'm obviously doing a tree right now that assumes I DO make the bet on the flop...

So to confirm... if I want to see if this is specifically BETTER than a check, I would have to do a second tree from scratch in which I check the flop, and then work all the turn and river spots based on the assumption now that the flop checked through... is that right?

Like... without giving away the ending, I can see from the work I've done thus far that betting is going to be better than checking, but to prove it, and to quantify how much better, I'd have to do the second tree right?

It would be a mass of extra work because the ranges are going to be different, I would have to think the turns and rivers through again... they wouldn't be enormously different... but the initiative is different, the balance of aggression is different... all enough to matter to future streets...

So to confirm... if I want to see if this is specifically BETTER than a check, I would have to do a second tree from scratch in which I check the flop, and then work all the turn and river spots based on the assumption now that the flop checked through... is that right?

Like... without giving away the ending, I can see from the work I've done thus far that betting is going to be better than checking, but to prove it, and to quantify how much better, I'd have to do the second tree right?

If you wanted to be very rigorous you'd make the second tree where you have checked back as well then compare the 2 trees. It will be time consuming but probably very much worth it if you tried to find out what it was about checking that gave you a lower EV for the hand.

It might seem like a lot of effort, but once you go to this amount of effort you'll just know the spot really well and in the future you'll be able to play this spot and similar spots much better.

If you wanted to be very rigorous you'd make the second tree where you have checked back as well then compare the 2 trees. It will be time consuming but probably very much worth it if you tried to find out what it was about checking that gave you a lower EV for the hand.

It might seem like a lot of effort, but once you go to this amount of effort you'll just know the spot really well and in the future you'll be able to play this spot and similar spots much better.

Cool. I thought so. I don't mind doing 2 trees... will just take longer.

This will be the most thoroughly investigated 40 cent bet in the history of poker.

Oh I have 1 final question, but I think I know what the answer is...

In my second tree, the check tree... I'm not like... changing all my decision points to a check as a sort of philosophical thing.. like here's what it looks like if I play it super passive...

In a sort of scientific way, I want to only deliberately change my flop action to a check... the rest of the tree beyond that is not being artificially changed to checks or anything, I'm still looking at those turn and river spots and deciding whether a check or bet is better... (though not at this point through starting 2 new trees from each turn point to examine whether a turn check/bet is better because that would be insane)

Just do the analysis as if you were playing your A+ game, because that way you can see if it's better for *you* to check back the flop vs betting it.

yeah that's what I figured.

Isolate the factor we're trying to study by making it the one variable in the experiment.

Cheers.

See you in a month!