JV gets his elbows dirty

[J_V]

I actually started to read Small Stakes Hold em by Ed Miller. Unlike the early days where I'd gobble up all the information I could find, these days it's very hard for me to sit down and look at a poker book for an extended period of time. Usually, I just skim my way through the books. I usually get suckered by one or two Sklansky quiz questions where he leads the witness with "two horrid players limp, who go to far with there hands, have money to burn, and their heads up their ass, you have J4d on the button, what do you do?" Of course the answer is fold. That dirty dog.

Reminds me of the time when we were taking derivatives in calculus class and the teacher is using all types of letters as variables - r,t,h,u,o,p and the he slips in an e^2. Of couse everyone put 2e and no one put 0. Slimy.

Anyway, today happened to be the day of a scheduled "Roadtrip Planning" meeting (a bunch of us were planning to go to Breckenridge over New Years), and the stuffy ones decided that planning now would be prudent.

Roadtrip planning meeting? Oxymoron. Call me old-fashioned, but roadtrips seem to work best when two buddies are sitting around the bigscreen drinking a Miller and one says, "Man, I'd sure like to go see Yellowstone Park someday." And then, the other buddy, says half-jokingly, "Yeah, we should go right now, what do we have to do?" A half hour later, they are in their car, rocking out to Jimi Hendrix doing 90 mph down a deserted road, with a tent on their hood.

But that's not the way this meeting was going, we had one overbearing chick, who was trying to run the show - fine by me, I was still playing poker, stuck 4 months rent, I wasn't planning on getting involved. But, as she grew more and more mettlesome and cumbersome (did we really need to go to CostCo and buy economy sized food?), I had to leave the room, before giving that AFUB a piece of my mind - and I took Ed Miller with me.

Damn good book. Probably my favorite out of all the 2+2 books. Makes HFAP look like a book for the dinosaurs. I flip it open randomly and I'm reading the chapter on River play. Mr. Ed Miller is having us decide whether to raise, call or fold on the last round, when we have a mediocre hand, a bet in front of us and a fish player behind us, salivating to overcall. You know the chapter.

Well, here's the hand.

15/30 on Party. Feel free to comment on any part of the hand.

MP limps. SB completes, I have 7h2h in the big blind and rap a knuckle.

Flop comes 9h 5h 2c. Giving me a pair and a flush draw. SB checks, I check, MP bets, SB calls, I check-raise and both call. I'm thinking there is a damn good chance I have the best pair and the best draw. 9sbs.

Turn Kc. Now the small blind bets. Both the SB and MP are pretty fishy. But the small blind is a real tool. He stinks. He always makes weird bets and weird times, the last few times it was the original top pair, but I thought that this bet was either a 9,5, or flush draw, and possibly some weird concoction of nothing. I think I fumbled the ball here and just called.
MP calls as well. 15sb's.

River Kd. Sb bets again. Your play? Can I atone for my turn error with some fancy play now? Put your own numbers or how likely you think the players are bluffing and will overcall if you are going to do the math.

I was planning on getting my elbows dirty and running my own math on this hand, Ed Miller style, but I'm gonna save that for a follow up post.

[bicyclekick]

I think you win this 5% of the time, tops. Probaqbly closer to 2-3% of the time.

Fold the river. Don't get any fancy raising ideas either.

[J_V]

I folded the river.

Let's look at some math. 8.5bb in the pot. If my opponent is bluffing 1/6 of the time, I have a clearly profitable call.

So in 6 trials:

(-1)*5 + (*8.5)*1= 3.5bb on the plus side.

Now lets assume that my opponent behind me has me beat half the time and will only overcall with better hands and will never fold any winners.

Now: (-1)*5 + 8.5(1)(.5) = -.75bb

So the presence of the third player turns this easy call into a fold.

What about raising? Assume: the opponent behind me will never call a raise and the Sb will never 3-bet bluff.

Now it's (-2)*5 + 8.5=- 1.5bb Yuck that's no good either.

So the conclusion is that these raise plays to knock out the overcaller must be made in large pots or when you think there is a larger chance your opponent is bluffing than in problem illustrated here.

To break even your opponent needs to be bluffing roughly 1/5 times. Was he bluffing 1/5 times this hand? Probably. Also, to make things easier on myself I changed some of the parameters, which might be slightly unrealistic (for instance, the overcaller will always fold to your raise, plus I could get 3-bet bluffed), so we should factor that in.

I would guess if I thought my opponent was bluffing around 35% of the time, a raise was in order.

In the actual hand, my opponent was bluffing and the opponent behind me called with J9 for a pair of nines and won the pot.

[Robk]

[ QUOTE ]
Now: (-1)*5 + 8.5(1)(.5) = -.75bb


[/ QUOTE ]

for this calculation i think you missed the half trial where you beat the sb but lose to the overcaller. so the ev per 6 trials is -1*(5) + -1*(.5)+ 8.5*(.5) = -1.25 BB, or -.21 BB per trial.

[ QUOTE ]
So the conclusion is that these raise plays to knock out the overcaller must be made in large pots or when you think there is a larger chance your opponent is bluffing than in problem illustrated here.


[/ QUOTE ]

i haven't looked at it but i'd guess that ed had in mind situations where the sb wasn't bluffing more often, but rather where you held a better hand. in this example say 66 might give you that 35%.

[J_V]

[ QUOTE ]
in this example say 66 might give you that 35%.

[/ QUOTE ]

Yes, in this specific example that is the case. I would argue though, in some cases like this (usually bigger games, where players are thinking), you will find your opponent with either a big hand or absolutely nothing.

But, I missed the point that if you could add the fact that the small blind may not be "bluffing" but you could still his hand, be it middle pair or whatever, then you'd have a clear call.

Good points, and thanks for the input Rob.

[raisins]

I came to similar conclusions, here's how I worked it out.

On the flop a check raise is clear, you've got a good bit more than 1/3 equity. I don't know if I would call it a good chance that you have best hand and best draw. On first look at the flop, before any action the chance of best hand is somewhere around 30%. After the action and with the information from that, the chance of holding the best hand must be quite a bit less.

On the river, if

1. the SB is just as likely to have bet his 5s as his 9s like this

2. would have slow played all of his strong hands going for the obvious check raise,

3. we can also eliminate high pairs as there was no raise before the flop

then we can quantify this group as 177 hands. Let's say he'll play this way with all 5's and 9's etc. Of course he won't but if he plays all of the possible hands at similar ratios to each other then we have an idea of his hands composition. It is just another way of saying he would be equally inclined to play each hand the same way.

A9 - 69; 9 hands 12 ways to make each: 108 combinations
A5 - 58; [no 2 pair] 5 hands 12 ways 1 hand 9 ways: 69

177 combinations. Another 54 combinations of medium and small pocket pairs, and the assumption that he would be as inclined to play these the same way.

If 231/x < 17/19 then it is a call, ignoring MP for the moment. There has to be more than 258 combinations that SB is equally likely to bet here for there to be a call. There are 36 combinations of flush draws and 112 combinations of Aces that don't demand a pre flop raise. If he is half as likely to bet a flush draw here as a mediocre pair and 1/5 as likely to bet with A highs as with a 9 then that is 40 more combinations of hands to make 271.

I started this problem thinking that MP had a mixture of hands that were worse and some better. So my initial guess at the possible composition is,

1. 88 or 66; 24 combinations

2. AQ - AT; 48 combinations

I'm thinking he folds the tiny pocket pairs on the turn and he is unlikely to be there with a 5. For one bet let's say he calls with pairs and folds the Aces. So, he calls 1/3 of the time.

--Call

versus the small blind 40 wins 231 losses; with the MP overcalling and taking his pots 27 wins and 244 losses.

27 W * 17 sb = 459
244 L * 2 sb = 488

-29 small bets per 242 trials -.11 sb per call.

--Raise

Say MP folds to a raise every time.

40 W * 17 sb = 680
231 L * 4 sb = 924

-244 sb; -.9 sb per raise

--

Hero is losing a high enough percentage (about 85%) of the time to SB alone to where doubling his exposure is hard to overcome. To make raising worthwhile hero needs to be winning more often than this to make the reward of MP's pots worth doubling his losses on the losers. MP's pots have to be a greater % of the whole.

The problem with the whole analysis is the MP limped and is therefore likely a passive player. An aggressive player might 3 bet the flop with top pair and a flush draw on the board but we can't eliminate a 9 from a passive players holdings.

There are a lot of possible 9s that a passive player could limp with. A9 - 89 is 84 combinations. So calling on the river is much more -ev then the above thought experiment suggests. With MP having only a 1/3 possibility of overcalling we can slightly change the assumptions and it will come out +ev. But if MP is just as likely to call the flop raise with a 9 as he is to 3-bet it then he is around 57% to have a better hand and overcall. Calling with a hand that will beat SB such a small percentage of the time plus facing overcalls by winners a majority of the time is hopeless. If you raise after running Ks and MP knows that sometimes SB gets goofy with it then the better 9s will probably overcall again screwing up the odds. So, yeah, fold, your win percentage would have to be so high against SB that he would be not much less likely to bet any ace or flush draw as a 5 or a 9. In that case, raise, it's probably worth it if MP would drop any 9.

I think raising the turn is best as hero is declaring himself to have a strong made hand and MP has to decide whether to call down with a mediocre hand looking at having to pay a probable 6 sb to get to showdown and win a 21 sb pot. Some 9s may fold. If SB happens to have a real hand you get an extra bet in when you improve.

regards,

raisins

[nummerfire]

"But, I missed the point that if you could add the fact that the small blind may not be "bluffing" but you could still his hand, be it middle pair or whatever, then you'd have a clear call."

Wouldn't you then have a clear raise.

Kim

[IlliniRyRy]

[ QUOTE ]
I usually get suckered by one or two Sklansky quiz questions where he leads the witness with "two horrid players limp, who go to far with there hands, have money to burn, and their heads up their ass, you have J4d on the button, what do you do?" Of course the answer is fold. That dirty dog.

Reminds me of the time when we were taking derivatives in calculus class and the teacher is using all types of letters as variables - r,t,h,u,o,p and the he slips in an e^2. Of couse everyone put 2e and no one put 0. Slimy.


[/ QUOTE ]

This is great stuff man, nice work.


[ QUOTE ]

Anyway, today happened to be the day of a scheduled "Roadtrip Planning" meeting (

[/ QUOTE ]

What happened to our Argentina bodyguard trip? I was looking forward to that and now you have a girlfriend and no time for RyRy.

[Robk]

i hate to nitpick at such a great post. but for the sb's hand composition you wrote:

A9 - 69; 9 hands 12 ways to make each: 108 combinations

I believe you're including K9 and 99 in this calculation, which he can't have according to your assumptions (and in any event there aren't 12 combinations of them). I count 6*12 + 9 = 81 combinations (12 for A9, Q9, J9, T9, 98, 96, and 9 for 97, since you hold a 7).

A5 - 58; [no 2 pair] 5 hands 12 ways 1 hand 9 ways: 69

For A5, Q5, J5, T5, and 85 I count 60 combos. I'm not sure what you had in mind for 9 ways here (did you intend to include 75 in the range?)

For MP you wrote

1. 88 or 66; 24 combinations

I count 2*(4c2) = 12 total combinations for 88 or 66. Although you can add 3 for 77 also.

There are a lot of possible 9s that a passive player could limp with. A9 - 89 is 84 combinations.

In general this would be true except that 99 is only 3 combos, bringing the total to 75. But taking into account the kings on board K9 is only 6 combos. But since I think we can throw out K9 and 99 in this instance anyway by assuming he would raise the turn, that would bring the total to 60.

Again sorry for criticizing these minor things as I think you did an excellent job with the important stuff.

Edit: made some mistakes of my own. /images/graemlins/smile.gif

[raisins]

Hi-

I do not consider your post nitpicking. While I agree that the errors probably do not change the conclusions, I appreciate the attention you gave my post. This helps me.

When I listed A9 - 69 for the SB hands, I intended to exclude K9 & 99 as these were strong hands where he must go for the obvious check raise almost all the time. So, [A9 Q9 J9 T9 89 69] each make 12 different ways * 6 different hands = 72 combinations. 79 makes 9 ways. 72+9=81 combinations. With the 5s the group I'm considering is [A5 35 45 65 75 85] (the two pair combinations are again tossed out) each of these make 12 ways except for 75 which is 9. So 5*12+9=69.

A questionable point is defining a group of 5s the SB would play. Seems like total guess work to me. I would probably toss the 35 - 65 and call the higher suiteds. I think the one above is a reasonable guess although he could be in there with more. Since this situation is not very close it doesn't matter if my assumption is a bit off.

So, 9s and 5s together is 150 combinations, off by 27 from the number in the first post.

Add the 54 pairs and that is 204 combinations. 204/x < 17/19 at minimum to have a call against SB ignoring MP. X=228; 228-204=24 additional combinations minimum. In my original post the minimum difference between the combination of hands where he holds winners and the total number of combinations he will bet was 27. So my guesses at his bluffing ratio of weaker hands should still work. The average sb lost at both calling and raising on the river would improve as there would be fewer losses relative to the wins. However, I did those calculations without considering MP could have a 9. That is enough to kick the ev of both of those options well into the negatives.

Hand combinations for MP, uh, yup and yup. I just spaced on 77, didn't count it and forgot to halve the combinations for 88 and 66. With the 9s, yes throw out K9 and 99, 60 combinations. So MP has fewer pairs relative to potential Aces than I thought. I guess I'll try and salvage my analysis by declaring MP as more likely to raise the Aces preflop, fold other overcards on the turn and thus the ratio of pairs to no pair is in the ballpark of the original post. That's my story and I'm sticking to it. /images/graemlins/smile.gif

Thanks for pointing these out.

raisins