Short Stack Play From the Small Blind Dennis Bragg

[sabre170]

I enjoyed this article. It simplifies and clarifies a lot of the discussion on the single table forum.

However, it glosses over an important point. All of the analysis is done based on the assumption that you can estimate what percentage of hands the BB will call when you push from SB. The implication is that this percentage depends only on what BB holds.

At the very end of the article, the author mentions that BB is more likely to make a loose call if SB pushes a small stack. This is critical to understand! According to one table, SB should push 32o if he judges BB to have tight calling standards. But if BB knows that SB should push with 32o (meaning any 2 cards), what should BB call with? Almost anything? Top 60%? Top 50%?

Personally, if I have 8BB after posting the BB on the bubble, and SB pushes 4BB after it is folded to him, my call depends on my read of him. I will certainly make a much looser call than he probably estimates I will. If I lose, we will have traded stacks. If I win, he is out and I have much improved my chip position.

Sabre170

[JC_Saves]

is this article missing one of the tables ie table 2? or is it missing table 5?

[Piers]

[ QUOTE ]
However, it glosses over an important point. All of the analysis is done based on the assumption that you can estimate what percentage of hands the BB will call when you push from SB.

[/ QUOTE ]

Not really, if you just assume that all Big blinds will defend with 50% you will be doing better than if you did not use the charts. If you see him defending with hands outside the top 50%, just change tables. The idea is not necessarily to get the exact right answer this time, but just make best decision with the information at hand.

Here are some points I would like to make about the article.

I think this article is doing the right things, however:

I think the difference between suited and non-suited is bigger than implied, more like over 10%. Still I think the tables are a plausible attempt to reduce the memory overhead.

Edit. Just reread the article, Ok the difference in EV between suited and non-suited might be only about 3% as stated in the article, however this leads to stack size differences in excess of 10%. Note that 3% is about 10% of 30.

At a full table in a tournament, any antes will make a big difference. For instance if you’re opponent is folding 50% of his hands I don’t think you should be folding any hand with 5BB.

Again in a tournament there is a factor that money won needs to be discounted. This is going to depend on the number of players left, and what proportion of the average stack you have. With less than 20% of the average stack, or with only 20 players left I would reduce the stack sizes by around 10%.

If you are prepared to favour expediency over rigor, I believe you can fairly easily extend the stack size equation to handle all positions.

[Jim C]

I thought this was a particularly useful article.

On a further thought, another article is the reverse of this coin.

If you know that SB is moving in with X% of his top hands, then what is it correct to call from in BB dependant on stack sizes? X is of course critical since many players do not change their move in requirements enough with a shortstack.

Any chance of u putting this info into an article Dennis?

Oh, Sabre, I think Dennis covered your question in the article.

He said your estimated value of his him to call. Soo... you have to take into account all things when considering whether he is going to call or not.

[wreckem]

[ QUOTE ]
However, it glosses over an important point. All of the analysis is done based on the assumption that you can estimate what percentage of hands the BB will call when you push from SB. The implication is that this percentage depends only on what BB holds.

[/ QUOTE ]

Yes, the article is based on the assumption that you can roughly estimate the calling standards of your opponent. In fact, that is stated in the assumptions section.

Table 1 is useful in estimating the percentage of calling hands. If you think your opponent will call with any queen you better use the 75% table. If you think he will call with any king, use the 50% table. At the lower buyin tables, I suggest starting with the 50% table. Then adjust based on how loose your opponent appears to be and how big is your stack.

Dennis Bragg

[wreckem]

[ QUOTE ]
is this article missing one of the tables ie table 2? or is it missing table 5?

[/ QUOTE ]

I can see all the tables. There should be 5 in all.

Dennis Bragg

[wreckem]

[ QUOTE ]
Edit. Just reread the article, Ok the difference in EV between suited and non-suited might be only about 3% as stated in the article, however this leads to stack size differences in excess of 10%. Note that 3% is about 10% of 30.

[/ QUOTE ]

I am not sure what you mean by this statement. Having suited cards is not really significant when you are heads up, so I ran the simulations using unsuited hole cards to simplify the tables. If your cards are suited, it might turn a borderline decision into a push.

Dennis Bragg

[wreckem]

[ QUOTE ]
If you know that SB is moving in with X% of his top hands, then what is it correct to call from in BB dependant on stack sizes? X is of course critical since many players do not change their move in requirements enough with a shortstack.

Any chance of u putting this info into an article Dennis?

[/ QUOTE ]

Have patience, grasshopper, and you will be rewarded.

Dennis Bragg

[Jim C]

(hop hop)

Thats good news. Oh, while i was sleeping last night, I was thinking about the variables regarding a call.

I'm sure distance away from money should be a factor while considering a call. Also other short stack size.

Soo.. by this i mean:

5 ppl left in a SnG. Top 3 positions get paid.

Relevant factors.

Our cards Top %? of cards
Their cards What will they push with?
Chipstacks - how many chipstacks are lower than me?
How many positions away from money and structure of payout
General aggressiveness of table-likelihood of them knocking each other out
etc.

After thinking about it, I think that calling is a much more complicated decision than moving in. What do you think?

[Piers]

My statement was intended to indicate how a 3% difference in EV can turn into a much larger difference in stack size.

The stack size equation

S = (C-1.5)/C(2W-1) gives

S inversely proportional to (1-2W)

So assume we have some hand that unsuited has EV .35 and suited .38 then

W = 0.35 = > 1-2W = .3

W = 0.38 = > 1-2W = .24

.3/.24 = 1.25

In other words a 3% difference in EV has lead to a 25% difference in stack size. Basically I think you underplay the difference suitedness makes.

I believe a much better way of doing it is to give the Stack sizes for unsuited hands, and say that for suited hands add one (or 25%) to the maximum stack size. This has the same memory overhead but I suspect is more accurate. (Maybe an extra processing overhead but I don’t find that a problem.)

Edit corrected signe.