Sit 'n Go Strategy study group -- Part III: High Blind Play

[Slim Pickens]

Good discussion about 22 vs. 87s. I'll add this problem. It's somewhat similar to 3-42 for those of you following along at home.

blinds 300/600
UTG (t3000)
UTG+1 (t3900)
Button (t4000)
Hero (t6100)
BB (t3000)

Hero posts SB of t300
BB posts t600

Preflop:<font color="gray"> 3 folds</font>, <font color="red">Hero raises all-in</font>

Which hand do you feel the best about pushing?
a) QJs
b) 22
c) A4o

[HajiShirazu]

Depends on what he calls with to some extent but QJs always does better than those other two if called really, I'm sure that was the point.

[Slim Pickens]

[ QUOTE ]
Depends on what he calls with to a large extent.

[/ QUOTE ]

Let's entertain the idea that he could call much wider than most players did back in Moshman's day. If BB only calls with premium hands, it doesn't matter that much what you have. As his calling range opens up, things start changing fast. The plots in SNGWhiz can be very good for understanding what is changing in this hand as his calling range opens up.

[flybe1]

[ QUOTE ]
Good discussion about 22 vs. 87s. I'll add this problem. It's somewhat similar to 3-42 for those of you following along at home.

blinds 300/600
UTG (t3000)
UTG+1 (t3900)
Button (t4000)
Hero (t6100)
BB (t3000)

Hero posts SB of t300
BB posts t600

Preflop:<font color="gray"> 3 folds</font>, <font color="red">Hero raises all-in</font>

Which hand do you feel the best about pushing?
a) QJs
b) 22
c) A4o

[/ QUOTE ]

A4o is clearly the weakest because it is dominated by a better Ace- exactly the type of hand that will call you.

Very little to choose between QJs and 22 (Eastbay 43.2%/41.4% vs. 42.9%/40.5% against typical calling hands.

Fold equity is of course king if your remaining opponents are tight, but if they really open up their calling ranges then I suppose it gets nearer to pushing against a randomn hand. Then QJs is best at 60.2%, followed by A4o at 56.7% and then 22 at 50.3%. So as always, it depends on their calling ranges. But how the heck do you estimate these in an ever changing situation, Slim?

But to answer your question, I think I feel most comfortable pushing with QJs, but I suppose I should go and check this on SnG Wizard!

[flybe1]

[ QUOTE ]
[ QUOTE ]
Good discussion about 22 vs. 87s. I'll add this problem. It's somewhat similar to 3-42 for those of you following along at home.

blinds 300/600
UTG (t3000)
UTG+1 (t3900)
Button (t4000)
Hero (t6100)
BB (t3000)

Hero posts SB of t300
BB posts t600

Preflop:<font color="gray"> 3 folds</font>, <font color="red">Hero raises all-in</font>

Which hand do you feel the best about pushing?
a) QJs
b) 22
c) A4o

[/ QUOTE ]

A4o is clearly the weakest because it is dominated by a better Ace- exactly the type of hand that will call you.

Very little to choose between QJs and 22 (Eastbay 43.2%/41.4% vs. 42.9%/40.5% against typical calling hands.

Fold equity is of course king if your remaining opponents are tight, but if they really open up their calling ranges then I suppose it gets nearer to pushing against a randomn hand. Then QJs is best at 60.2%, followed by A4o at 56.7% and then 22 at 50.3%. So as always, it depends on their calling ranges. But how the heck do you estimate these in an ever changing situation, Slim?

But to answer your question, I think I feel most comfortable pushing with QJs, but I suppose I should go and check this on SnG Wizard!

[/ QUOTE ]

Interesting results on SnG Wizard.

Firstly, all three are clear pushes in this situation. In fact, you should push with ATC except against a "very loose" BB (45%+).

There is very little difference between the three hands if the BB is very tight, tight and average. However, if the BB is loose, then A4o is somewhat better than the other two and if the BB is very loose, 22 is quite a bit worse than the other two.

Is this the point you were trying to make Slim?

And how do we apply this information in practice?

[Slim Pickens]

Here's the original hand for discussion.
blinds 300/600
UTG (t3000)
UTG+1 (t3900)
Button (t4000)
Hero (t6100)
BB (t3000)

Hero posts SB of t300
BB posts t600

Preflop:<font color="gray"> 3 folds</font>, <font color="red">Hero raises all-in</font>

Which hand do you feel the best about pushing?
a) QJs
b) 22
c) A4o

As a basic review, here's how to calculate Hero's tournament prize pool expectation ($EV). This is the same process programs like SNGPT or SNGWhiz use. Since this is blind-v-blind and all the possible action is described perfectly (at least as far forward in time as we go) by our three possible outcomes, SNGPT, SNGWhiz, a hand calculation, and any other method should get the same answer to within the numerical precision.

1) What are Hero's options?
In this case, the action has been folded to Hero on the SB. He may either push or fold. With the lone remaining opponent sitting on 5 BB, Hero really shouldn't be considering any other option outside of a few specific circumstances, all involving some fairly advanced reads and strategies. Push/fold is simple and effective with the blinds this high.

2) What are the possible outcomes that result from each option, and what is Hero's prize pool equity in each possible outcome?

2a) Hero folds
There's only one outcome. BB wins the hand and picks up Hero's t300 small blind.

2b) Hero raises all-in
There are three possible outcomes. BB can fold, BB call and Hero can win, or BB can call and Hero can lose.

3) What is the tournament prize pool expectation of each option?
This has to be divided into two parts, one for each option Hero has.

3a) Hero folds
If Hero folds, the hand is over. The stacks will look like this for an instant before the button moves and the next hand is dealt.

UTG (t3000)
UTG+1 (t3900)
Button (t4000)
Hero (t5800)
BB (t3300)

Putting all five of those stacks into an ICM calculator gives Hero's prize pool equity as 26.19% of the total prize pool. This is EV_fold.

3b) Hero raises all-in

3b-i) Hero pushes, BB folds
UTG (t3000)
UTG+1 (t3900)
Button (t4000)
Hero (t6700)
BB (t2400)
Hero's $EV in this case, $EV_push/nocall, is 28.88% of the total prize pool.

3b-ii) Hero pushes, BB calls and Hero wins the hand
UTG (t3000)
UTG+1 (t3900)
Button (t4000)
Hero (t9100)
BB (t0)
Hero's $EV in this case, $EV_push/call-win, is 35.34%.

3b-iii) Hero pushes, BB calls and Hero loses the hand
UTG (t3000)
UTG+1 (t3900)
Button (t4000)
Hero (t3100)
BB (t6000)
Hero's $EV in this case, $EV_push/call-lose, is 16.65%.

4) How likely is each outcome within each of the available options?

4a) Hero folds
BB will win the pot 100% of the time, obviously.

4b) Hero raises all-in
Here's where your skill as a poker player comes out. It is necessary to determine a range of hands with which BB will call. For simplicity, assume BB will call with 100% of the hands in his range and 0% of the hands not in his range. Say Hero was dealt Q[img]/images/graemlins/diamond.gif[/img]J[img]/images/graemlins/diamond.gif[/img]. Let's also say, based on what Hero has seen BB do over the course of the tournament, that he thinks BB will call with this hand range: {22+,A7s+,A9o+,KJs+}. Now, we go to PokerStove or any accurate holdem EV calculator and see that...

{22+,A7s+,A9o+,KJs+} constitutes 12.9% of possible starting hands given Hero holds Q[img]/images/graemlins/diamond.gif[/img]J[img]/images/graemlins/diamond.gif[/img] and QJs will win 40.5% of the time against {22+,A7s+,A9o+,KJs+}.

4b-i
BB will fold 87.1% of the time (P_push/nocall)

4b-ii
BB will call and Hero will win 40.5% of the 12.9% of the time BB calls, for a total of 5.2% (P_push/call-win).

4b-iii
BB will call and Hero will lose 59.5% of the 12.9% of the time BB calls, for a total of 7.7% (P_push/call-lose).

Check that 100% of the time, something happens... good.

5) What is the $EV value of each option.

5a) Hero folds
$EV_fold=26.19%

5b) Hero raises all-in
$EV_push=(p_push/nocall*$EV_push/nocall)+...
(p_push/call-win*$EV_push/call-win)+(P_push/call-lose*$EV_push/call-lose)
=(0.871*0.2888)+(0.052*0.3534)+(0.077*0.1665)
=28.27%

6) Which option has the higher $EV value?
28.27&gt;26.19 therefore pushing&gt;folding

OK, this is all background to help answer the original question of which hand would you rather have and why. This is the process that we'll have to do many, many times, and display graphically.

[flybe1]

Thanks Slim for a superb explanation of how to work out tournament prize pool expectation.

However, since you can not do these calculations at the table, presumably you get a feel for what hands are right to push through experience and by studying critical hands later with an ICM calculator?

In the meantime, any tips for estimating opponents all-in calling ranges from the stats collected by "sniffers" and how these change as your and their M diminishes?

[Slim Pickens]

[ QUOTE ]
Thanks Slim for a superb explanation of how to work out tournament prize pool expectation.

[/ QUOTE ]
Make sure you understand how that all happened. The real interesting part of a SNG is taking that operation as a whole and looking at how the results change as the stacks, calling ranges, and your hand changes. Also, I forgot to mention that push came out as +2.08%, and that is a very favorable number as far as pushes go. +0.5% would have been a no-brainer.


[ QUOTE ]
However, since you can not do these calculations at the table, presumably you get a feel for what hands are right to push through experience and by studying critical hands later with an ICM calculator?

[/ QUOTE ]
Yeah. That's pretty much it. Even if you could do the calculations at the table, you would still have to feed your calculation engine the hand ranges. Overall, learning how to quickly estimate the whole process by yourself will allow for the fastest, most accurate, decisions.


[ QUOTE ]
In the meantime, any tips for estimating opponents all-in calling ranges from the stats collected by "sniffers" and how these change as your and their M diminishes?

[/ QUOTE ]
This is a bit trickier than it might seem. You might see a player who has been playing very tight through the first four levels. The first instinct is to think that when he finally wakes up in Level 5 and pushes from the button into your BB that he must have a very good hand. You look down at ATo and think that it must not do well enough against his very tight pushing range to be a call. Maybe, instead of being a nit, he's actually a decent player, who like you, understands the value of playing tight and avoiding marginal situations early, then being aggressive and stealing blinds late. His range may very well be quite wide despite not having played a hand up to that point. Looking at stats is best for identifying types of players. Then you have to decide what that player is doing based on how that type of player plays at your current stage of the game. That's not necessarily some sort of extrapolation of their previous play.

[flybe1]

Thanks again Slim.

Any response to my belated comments about the last question you asked in the Part II: Mid Blind Play study group?