Sometimes the force is with you

[peterbirks]

It would be foolish of me to deny that the last month (since May 6, to be precise) has been good. Take the good times with the awareness that they are just that, unusually good times. And bear them in mind when the inevitable bad times come.

I've noticed a pattern in my winnings in that, after a particularly good streak, I get a short period where the wins continue, but at a slower pace, before a bad run starts. I think I have worked out why this is.

Basically my head knows that I am in the middle (or coming to the end) of a good run. So I subconsciously prepare for the bad run that must inevitably come. Now, mathematically, this is nonsense. Although you know you WILL get a bad run, eventually, it is no more nor less likely to start tomorrow because you are in the middle of a good run. But never mind the maths ... this is how my brain behaves. As a result I start "protecting" my good run ... the standard "manufacture of a winning streak" as Caro terms it. This causes me to win less than I would have won if I had carried on playing the same way. (Of course, there is a danger that you can move too far in the opposite direction -- the "I'm hot, I can play anything" attitude, which also brings any good run to an inevitable crash and burn).

So today, in a sense, I manufactured a win. OK. I sat down and decided to play for three hours, and I did play for three hours. And I won $36.50. But this was from a deficit of $260 at one point. And I suspect that if I had got significantly up at any point, I would have stopped. It just so happened that I moved into profit in the last couple of minutes of the three hours.

So we see the first signs of the "protecting a winning streak" emerging, characterized by more conservative play. I must be aware of this and not fall into the trap this time. The cards have no memory!
-------

Below is an Omaha hand from PokerStars. I'd had a blinding couple of hours at $5-$10 (two tables) and needed only another 12 points to qualify for my bonus. The cards had "gone cold" at the limit tables and I had given back $100. I also didn't feel good about the tables. So I moved over to the Omaha.

A couple of hands: I lost the first, and won the second. In the latter case, it was definitely one of the force being with me.

Game was 25c-50c ($50 max buy-in)

SB: Ann&Ben ($67.45)
UTG: Sniff1003 ($17.15)
MP2: PJB ($45.05)

My hand: Ah As 4s 2d


Sniff called 50c. I called. Button called Ann&Ben raised to $2. Sniff called. I reraised to $9. Button folded. Ann&Ben and Sniff both call. $26.60 in pot.

Flop comes Qc 5h 3h

Checked round to me. I have top pair PLUS the backdoor straight draws. I think that I am looking pretty good here. I bet $26.60. Ann&Ben folds. Sniff calls all-in for $8.15

Turn Ks, River Qh

Sniff shows 8d 8h Qc 7s. Sniff wins $42.20 with three Queens.

I mutter "good grief" to myself. In the words of Paul Samuel, how big a favourite am I? My instinct is about 75%. Running it through Pokercalc. Hmm, not bad. In fact it's 77.44%.

But there was also a little smile there. If there are Omaha players this bad even on Stars, the renowned home of tough ring games, then things are looking up.
------
The next hand was a funny one. I suspect that in a bigger stakes game I would have folded. But the stacks of cash held by the players are of some importance. We were not "too deep in" on the flop bet. But I still feel a bit guilty. My only excuse was that I was sitting on more than $400 profit from the ring games!

BB Ann&Ben ($20.60)
UTG: Scumbag Joe ($49.25)
CO2: PJB ($55.50)
Button: teach jr. ($77.25)

Dealt to PJB: Kd Qc Tc 3d

Scumbag flat calls. I like this hand better with as few opponents as possible, given the lack of nut-flush possibilities. I also fancy buying the button. I raise $1.75 to $2.25. teach jr then reraises $1.75 to $4. Another limit player! Ann&Ben calls. scumbag joe calls and I call.

Flop brings Js 9h 4d.

scumbag joe bets $15. This leaves him with $30. I had a bit of a think about this. Assuming scumbag has a set, I suspect that I should fold. And yet, he still has $30 left. If I HIT on the turn, will he be able to walk away from a 7/2 shot, even though he would only be getting 2/1? This is a classic implied odds dilemma and I really don't have the Omaha experience to know whether callling would be profitable. And I also have a raiser behind me? Will HE decide to call?

Sod it, I'm $400 up. This board is a rainbow. I have 13 outs. I call. teach jr also calls. I suspect that this makes things better for me rather than worse. I just hope that I'm the one drawing to the straight and the other two are drawing to a paired board.

Turn brings the 8s. Woohoo!

scumbag goes all in for $30. Did HE have the draw? I have $36 left, but a raise is pointless. A call is much more likely to get teach to play. But he doesn't, he folds. River is an irrelevant 6s.

scumbag joe shows Jh Jd 7h 3s for trip jacks.
I have top straight and win $118.75 pot.

Running it through pokercalc, I see that I am not as bad a dog as I thought. 39% to 60%. With the implied odds available, that would seem to make the flat-call on the flop correct. Obviously if the turn misses me and scumbag bets the pot on the turn, then I fold.

I gave teach jr a possible holding (given his betting on the hand) of Ac Kc Kh Qd
That puts scumbag up to 65% and me down to 28.8%. However, for that 11% drop in equity I am getting an extra $15. But the heads-up scenario is nicer.

Comments

When you are 39%

[iadams.livejournal.com]

you have an automatic call, unless you have a reason to think you'll get a raise behind you. Even if you miss on the turn, I make you around 32.5%, which makes a call not far wrong, and certainly right if your opponent underbets the pot.

On the first hand, your straight draws are NOT back-door, although they are to non-nut straights, any A,2,4, or 6 gives you a straight, although some make flushes possible. To be really picky, you also have an overpair, not a top-pair. Against a a good player (which Sniff clearly isn't), your Ah gives you extra outs, since you can bet a H turn too.

Are you using the Hendon mob poker calcuator? You keep getting numbers that are different from the calculator I wrote 4 years ago. I tried your first hand (putting Sniff with Qd since I assume there was only one Qc in your game), getting 74.0244% against HendonMob's 77.55%. Interestingly, putting the Ks turn into each I get 82.50%, HM gives 82.58% (same result from each of several tries). However there are only 40 possible final cards now, and if 33 give you a win, then it's exactly 82.50%, so I've no idea where this extra .08% comes from. I'll be checking my code on the flop analysis to see if there's a bug in my code - will let you know the result.

Doh!

[iadams.livejournal.com]

When you get a raise behing you, it's only a problem if the raiser is taking away a large proportion of your equity. For a single pot raise and call by Sniff, you still only need 31.25%, tending towards 33.33% as the number of raises increases.

Calculators

[iadams.livejournal.com]

Worked out that I'd mistyped a card when entering the hand into my calculator, which accounts for most of the 3% error, I now get 77.439%, which agrees with your result. Looks like the HendonMob calculator works with a random sample, which is fine for wide-open situations (e.g. KQs vs Ax), but is just silly when there are a small number of combinations, and if your random number source isn't good enough, will give wrong answers.

Glad to find my code isn't as far off as it might have been.

Re: Calculators

[peterbirks.livejournal.com]

I actually use pokercalc, a Finnish piece of freeware.

Your point that "if you are 39%, it's an automatic call" is of course true. The trouble is, that situation only arises if, for example, there was $40 in the pot, my opponent went all in for $40, and I call.

That is not the situation here, where he bets $15 into a $15 pot, and has $30 left, which he can still bet on the turn.

I am not 39% to hit the straight in one card. And, if I fail to hit it in one card, I have to fold to the $30 bet on the turn, (although the additional player behind me complicates things slightly)

My absolute odds on the flop bet are not good enough, so I am going for the implied odds that I will hit on the turn, that my opponent will call my all-in bet (as it happens, he made the all-in bet, but that's irrelevant for the maths) and that he will not hit the paired board on the river.

If all of those things come true (19%), then I am betting $15 to win $73.
If I miss the turn (75% chance) then I am minus $15
If I hit the turn but am outdrawn on river (6%) then I am minus $45.

This assumes that my opponent will call the turn bet.

So we get an EV of $14.50 minus $11.25 minus $2.70, or plus 65 cents.

If opponent folds the turn bet, we get

75% minus $15
25% plus $48

= $12 minus $11.25, or plus 75 cents.

So the play is marginally correct.

----

On the opponent's call when I had the Aces and the straight draw, yes, it is of course a straightforward straight draw that I have (insofar as anything can be straightforward in Omaha) and I have the overpair, which does make a difference. On the opposite side, my opponent's short stack actually makes it correct for him to call! He is putting in $8.15 at odds of more than 4-to-1. So, clearly against short stacks such as this (if it is a heads up situation) you should try to manoeuvre it so that your opponent has enough chips left for his call to be mathematically incorrect.

Re: Calculators

[(Anonymous)]

Hi--

Where do you get $14.50 in the first set of EV calcs?

Re: Calculators

[peterbirks.livejournal.com]

About 19% of my expected profit on the hand. Did it in my head and I see from the calculator that it actually comes to $13.87. However, since a lot of the figures are "rough and ready" (as they need to be if you are playing in a live situation), including the 19%, I don't think that it impacts the overall analysis, which is that a call is probably marginally correct. For example, there is a chance that player three will call the $30 on the turn (which didn't transpire). And I haven't factored this in. Ther eis also a chance that the original bettor actually has two pair, and I haven't factored this in. I might even have an outdraw on player 1 (or 3) which leaves them drawing virtually dead compared to me (although I accept that this is an unlikely scenario).

My actual post was to correct the idea that I was 39%. I knew that I was less than that and that the bet related to implied odds (which is necessarily subjective). I just did a rush job to try to see if the call was still worthwhile. I reckon that it probably is (marginally).

I reckon you're better off than you think.

[iadams.livejournal.com]

I'm doing the sums for the heads-up situation; I haven't done them for the third player, but my feeling is that he'll give give more money than he'll take EV.

If you call the flop and fold the turn you lose $15. If you call the flop and turn you win $60, or lose $45.

When you call the flop, you will fold the 7/41 turns that pair, of the other 34 turns: 10 put you ahead, leaving you 75% to win on the river; 15 don't help and leave you 32.5% to win on the river; 8 give you a backdoor flush, leaving you 42.5% to win on the river.

Assume in each case your opponent bets his last $30 and you call. I make this +$7.50. If you hit the turn, bet and he folds, your overall EV comes down to $3.75 - although your opponent got his money in here anyway.

I don't call +$7.5 for a $15 marginal.

The point you are missing is that you get a +EV call when you both miss, you are now getting a +EV call: folding a blank turn is actually a mistake.

Interestingly if there is enough money for the turn bets to be $45, I still make it +$7.5 if your opponent always calls, but it now becomes a -$3.75 EV if your opponent folds when you hit.

Must do more (careful) sums before posting.

[iadams.livejournal.com]

The above forgot to include a turn-card weighting, so the real figures are actually:

As the cash and play were: +$1.35
As the cash was, but opponent plays optimum: -$2.30
Deep cash, fool opponent: +$1.13
Deep cash, optimum opponent play: -$4.35

You still missed the value of the back-door flush draw, and the +ev call you get with the shallow money when you miss when doing your sums

So it is indeed very marginal, and against someone who won't call when you hit, probably not worth playing.

The third player will probably only have a call when his hand duplicates yours, which kills both your hands. Give him Kh Qs TD 3C, and the lead player's ev on the flop rises from 61% to 65.8%.

Looks even more like a pass when you factor that in.

Re: Must do more (careful) sums before posting.

[peterbirks.livejournal.com]

Ah yes, the diamond-diamond. Say a diamond comes and I miss my straight. That gives me an extra, what five or six outs. That pushes me to over 2/1 for my money, so makes the call okay. Good point.

The thing is, my instincts that this call was marginal were correct. If I run it through the figures and get a +EV of 50%, then I would run the figures aagin, because they wouldn't feel right. This is the amazing thing about Paul Samuel, that he seems to have no feel for the correct odds.

I based my call on the fact that opponent would not be able to lay down his 7/2 shot. In fact, it was better than that. because he was out of position, he does not know that I have already hit. He is therefore scared of giving the free card which, in fact, I already have. We are hardly double-dummy here (it's a low-stakes game, he knows nothing about me). Perhaps he thought I had called on a ropey two-pair.

So, my gut feeling, that a call in this situation is probably marginally plus EV, whereas in a higher-stakes game it would probably be a fold, look pretty good. Then again, marginals like these are only of technical interest. What really matters is that you do not make mistakes in the 80/20 situations, where you are the 20%.

BTW

[iadams.livejournal.com]

My comment about my blog was serious, I'd welcome some feedback on what I posted.

Re: BTW

[peterbirks.livejournal.com]

I have posted a brief comment. Perhaps more later.