HU SNG theory question 1

[jon_1van]

I recently started playing the HU SNGs on PS.

For those who aren't aware of them, each player starts with 1500. Blinds levels are 10/20-->15/30-->25/50-->50/100

In my expereince (of about 20 low buying) sng's it is rare for the SNG to last to the 50/100 round.

So most of the SNG is played pretty deep.

Question 1
How shallow does your opponent need to be for you to risk doubling him up with a most likely inferior hand. For instance, you have 57s, you think your opponent has something like KJ. He pushes _____ into a pot with 75 in it. What does that blank have to be for you to call with any hand you think is live.



Thought process A.

Well, if you think your hand is live you should call with all sorts of hands. The reason is if you think you are 40/60 then by calling you will win the whole SNG 40% of the time. And the remaining 60% of the time you'll win (say you have a slight edge) 60% of those times. So in total you'll win 40% + .6*60% = 76% of the time.


Thought process B.

It would be retarded to call when you think you are a 40/60 dog. Just ship him the 75 and move on.


Now, I see flaws in both thought processes.

In process A we might not be maximizing the possible % win when auto calling a dog. It should be easy to contruct a situation where we win a large portion of the time when we already have a large portion of the chips. So just showing that we win alot does not show that we win the highest % possible from that point in the SNG.

The problem with B is that using this logic too much when the opponent starts trying to push themselves out of the basement might really hurt your chances. If you allow an opponent to push 3 consecutive times they could easily go from 600 to 600 + 50 + 25 + 50 = 725. Now if you decide to take a stand your opponent goes to 1450 instead of 1200.


So how many chips does your opponent have to have for you to call the 1st push irregardless (3k chips in play)? Currently my answer is 700 chips. From here on out I push every playable hand. I also call any push if I think my cards are likely to be live.

[Indiana]

Very very good post. Give me some time to digest. I am working on becoming a heads up specialist so this is great for me. I do think there is a point where you can take the worst of it and get him all in for the win. Personally, for me if I can still retain a 2:1 chip lead after losing the hand then that's my threshold for considering the 57 vs KJ stuff that you discuss.

Comments from other good HU players?

Indy

Indy

[valenzuela]

I personally think its retarded to call knowing the play is -CEV.

Well if ______ is 80, then of course it's a call.

[Irieguy]

It depends on your opponnent and how he's playing his short stack.

A lot of players get nervous/frustrated/pissed, etc. when they get short and push frequently. If that's the case, you have plenty of time to wait for a good racing hand that rates to be well ahead of him with which to win the tournament.

If the player is better than you, and is stubborn with his short stack, you would do well to take any 2 live cards against him for a chance to win.

If the player is passive, has been passive, and will continue to be passive even with a short stack... then you don't ever have to race until he is all-in for his BB.

So, this isn't a math problem. Nor are most HU strategy and decision dilemmas.

Irieguy

[HesseJam]

[ QUOTE ]
I personally think its retarded to call knowing the play is -CEV.

[/ QUOTE ]

This is not true. HU, you cannot judge a play isolated from the play before or from the play you plan for next hand. HU is a dance. It is one whole story. You can very well make a -CEV play that gives you a greater return two hands later. Of course, you should not make a huge -CEV move. But a slight -CEV move is OK.

My HU opponents at Stars 15+1 tend to be not too good. I tend to sacrifice a couple of chips in terms of -CEV here and there just to get a feel how they play and to create an impression how I play.

[jon_1van]

[ QUOTE ]
I personally think its retarded to call knowing the play is -CEV.

[/ QUOTE ]

What about $EV?

[HesseJam]

Oh, everybody already read this months article? Well worth it for anybody who feels lost at HU battles.

[valenzuela]

the example you described above would be a +ev play. Any play that will bring chips on the long run is +ev, even if its not +ev right away.
to the other dude:Cev is the same as $ in Hu

[jon_1van]

[ QUOTE ]
If the player is passive, has been passive, and will continue to be passive even with a short stack... then you don't ever have to race until he is all-in for his BB.

[/ QUOTE ]

Really so your opponent pushes 200 into a 75 pot and trying to bust him doesn't at least cross your mind?


[ QUOTE ]
If the player is better than you, and is stubborn with his short stack, you would do well to take any 2 live cards against him for a chance to win.

[/ QUOTE ]

So how much will you call here for a chance to bust this guy? 400? 600? 1000? If you do double up your skilled opponent now you put yourself in dramatically different situations. Big lead/evenish/Big Dog.

Come at least try to come up with some number.

[jon_1van]

[ QUOTE ]
Cev is the same as $ in Hu


[/ QUOTE ]

This assumes each chip yields the same chance to win. I don't think this is strictly true. It might be a good aproximation.....but I've always hated this assumption.

[jon_1van]

[ QUOTE ]
HU is a dance. It is one whole story

[/ QUOTE ]

I think this is very true.

[the shadow]

[ QUOTE ]


Quote:
--------------------------------------------------------------------------------

I personally think its retarded to call knowing the play is -CEV.


--------------------------------------------------------------------------------



What about $EV?

[/ QUOTE ]

In a HU freezeout, CEV = $EV.

The Shadow

[the shadow]

[ QUOTE ]
In my expereince (of about 20 low buying) sng's it is rare for the SNG to last to the 50/100 round.


[/ QUOTE ]

Despite the fact that I just won a 4-player HU freezeout in 3 hands, it's not unusual for me to have the games reach 50/100 or even occasionally 100/200.

The answer to your question depends on a lot of factors, including your read of the villian (villian's prior play plus your expectation of villian's future play), your expectation of what edge you have over villian (or vice versa), and the stack sizes, which I didn't see in your OP.

The Shadow

[ QUOTE ]
[ QUOTE ]
I personally think its retarded to call knowing the play is -CEV.

[/ QUOTE ]

What about $EV?

[/ QUOTE ]

In HU, $EV is proportional to cEV. You're playing for cEV HU.

[el_dusto]

[ QUOTE ]
The answer to your question depends on a lot of factors, including your read of the villian (villian's prior play plus your expectation of villian's future play), your expectation of what edge you have over villian (or vice versa), and the stack sizes, which I didn't see in your OP.

[/ QUOTE ]

uh...

maybe i misread, but i thought it *was* a post about stack sizes.

the poster is talking about instacalls, right? i tend to call any push when they're down to about 500; i don't want to accidentally walk into AA, KK, or QQ with any 2 and make our stacks even again... i'd at least like to keep a 2:1 advantage, especially when it's only 5bb or so.

this sounds like a job for gigabet's "chip chunk" idea... link?

[jon_1van]

[ QUOTE ]
[ QUOTE ]


Quote:
--------------------------------------------------------------------------------

I personally think its retarded to call knowing the play is -CEV.


--------------------------------------------------------------------------------



What about $EV?

[/ QUOTE ]

In a HU freezeout, CEV = $EV.

The Shadow

[/ QUOTE ]

Not strictly true.

Assume BB = 3000

If I have 10 chip and you have 2990 you will not win this tourney 2990/3000 times.

In order for me to win I will need to be dealt the winning hand 9 times in a row
20 (1)
40 (2)
80 (3)
160 (4)
320 (5)
640 (6)
1280 (7)
2560 (8)
WIN (9)

I will win 1 in 512 times, or 6 times out of 3000 not 10.

Here are the factors that I think matter and how I consider them. They are in the order in which I weight them:

1) Skill level If you're of equal or lesser skill than your opponent and your opponent doesn't play a short stack like an idiot, call when straight pot odds dictate a call. You gave the numerical answer to that in your post.

2) Villain's Short-Stacked Play. If villain can't play a short stack well, especially if villain's overly tight as a short stack, fold much more frequently.

3) $/hour v. ROI. If you're much more concerned with $/hour than ROI (this requires that you be psychologically comfortable with the idea of a lower ROI and that your ROI is high enough that you don't risk it becoming negative), you should take all +cEV situations when either you or your opponent is short-stacked and all +cEV situations that allow you to beat the rake when you're about even.

4) Meta-game stuff. (This is low here because you're talking about a situation in which you're very very likely to win in a few hands no matter what, so meta-game's very unimportant. This factor becomes more important as the effective stack to BB ratio increases.) At my stakes ($10+.5), most players tend to play better after I call them with 23s; overly loose players realize that I'm going to call with junk and find a more reasonable pushing range, and overly tight players realize that I'm going to call with junk and start pushing with more hands that beat junk. So, obviously, I usually call pushes less often than straight 1-hand cEV dictates. However, there are certainly some players who will respond by giving me walk after walk and blinding themselves off, and I try and find these players and call them with 23s as much as possible.

5) Variance. Variance in HU SnGs is really low, so it shouldn't be much of a factor. I probably shouldn't list it here, but if everything else has you completely unsure of what to do, fold to lower variance.

[Mr_Oog]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]


Quote:
--------------------------------------------------------------------------------

I personally think its retarded to call knowing the play is -CEV.


--------------------------------------------------------------------------------



What about $EV?

[/ QUOTE ]

In a HU freezeout, CEV = $EV.

The Shadow

[/ QUOTE ]

Not strictly true.

Assume BB = 3000

If I have 10 chip and you have 2990 you will not win this tourney 2990/3000 times.

In order for me to win I will need to be dealt the winning hand 9 times in a row
20 (1)
40 (2)
80 (3)
160 (4)
320 (5)
640 (6)
1280 (7)
2560 (8)
WIN (9)

I will win 1 in 512 times, or 6 times out of 3000 not 10.

[/ QUOTE ]

Your example is incorrect as after winning 8 hands in a row you can now lose a hand and then still win the tournament. I am not certain how to calculate the probability of victory from that point off the top of my head.

But this is minutae and most likely unimportant as I agree with Irie.

-Mike

[jon_1van]

[ QUOTE ]
But this is minutae and most likely unimportant as I agree with Irie.

[/ QUOTE ]

Yes, it is unimportant.


Ok your agree with Irie, great. Now provide some numbers.

Against an opponent who is way too TP you'll call _____
Against an opponent who is way too LAG you'll call______
Against an opponent who is way too LP you'll call_____
Against an opponent who is TAG you'll call_____

[The Student]

which article are you talking about? the one about the battle of the blinds?

[jon_1van]

[ QUOTE ]
and the stack sizes, which I didn't see in your OP.


[/ QUOTE ]

Since its heads up and there are 3000 chips in play if your opponent has x chips you have (3000 - x) chips.


I really haven't played a whole lot of these. And I'm not seeking the best of competition right now.

Right now my "autocall" line is around 650 vs opponents that are too tight (IMO).

If I double them up....it isn't the end of the world because I feel like I have good control over an TP opponent. And given a double up, I feel like I can whittle them down alot faster this time because the blinds are higher.

I'm more cautious about shipping chips to an aggressive player. This is for 2 reasons.

1. I don't feel like you can have nearly as big an edge over an aggressive player than you can over a passive player. Therefore many more times when I double up the opponent I go on to lose.

2. Playing an aggresive opponent when the stacks are deepish (~30BB), can be dramatically harder than playing that player with shallow stacks (~13 BB)

[Mr_Oog]

Perhaps you misunderstood that I agree it is not a numbers question.

Remember, half the time I get to act frst. How is my opponent playing his BB? If too tight then I do not have to call with trash as I get to take the blinds right back every hand. If too loose I get to simply choose the hand I think has a decent chance to knock him out. (I won't say top 50% or such as each opponent has their own personal hand preferences. For example some would rather play A3o than KTs. Some would rather play 87s than Q8o etc. This must be factored into your hand range.)

Here is another question regarding this. How afraid am I of doubling my opponent up? Is he better than me or will I have an easy time getting the chips back?

Way too complex for a mathematical formula.

-Mike

[Onaflag]

[ QUOTE ]
5) Variance. Variance in HU SnGs is really low, so it shouldn't be much of a factor.

[/ QUOTE ]

I'm tempted to start a new thread about this, but please explain this in further detail. Enquiring minds want to know.

Onaflag.............

[HesseJam]

[ QUOTE ]
which article are you talking about? the one about the battle of the blinds?

[/ QUOTE ]
2+2 Magazine November issue (http://www.twoplustwo.com/magazine/current/streib1105.html)