Heads up Theory

[microbet]

You had just pushed a bunch. I think that calling range is fair. I might have used a bit broader range though.

I don't have any tools with me now - what's the result with that range?

BTW, what part of the world are you in? I'm in Southern California.

[Jman28]

[ QUOTE ]
You had just pushed a bunch. I think that calling range is fair. I might have used a bit broader range though.

I don't have any tools with me now - what's the result with that range?

BTW, what part of the world are you in? I'm in Southern California.

[/ QUOTE ]

That range made a push +.3%.

I'm in the midwest, so central time zone. Usually not awake at this hour (I like to wake up at 3:30 pm and go to bed at 4 am) but I got two phone calls while sleeping and couldn't go back to bed. I assume that's what this question was about. right?

[A_PLUS]

Well, I get it, so thats a start.

I agree when you have more chips. Basically, it takes a situation where against a perfect opponent it would be EV neutral and puts him ina spot to makea decision which can only increase your EV over time.

Pretty much standard aggressive poker theory.

Now onto the real 'theory'

From a game theory perspective, I can see where this comes from.

As the chip disparity grows, the leading player can afford to make riskier and riskier plays b/c losing a hand still leaves him with enough chips to win a reasonable amount of the time.

So, your premise is that when given a situation to which we are EV indifferent, we should choose the one which will make our opponent make riskier moves going forward. If I am wrong with my thoughts so far, skip the rest.

My problems:
-Your opponent gets off easy. He increases his equity without having to make a decision. When we put out opponent on a range and get an EV neutral spot, any hand he plays outside of that range is +EV. So you are putting a lot of faith in your reads and your opponent here.

-So for this to make any sense, we are assuming our opponents are playing near optimal poker. So his range will slide wider by a small amount given the new disparity in stack sizes. For us to take advantage of this. He needs to be dealt a hand in exactly that new portion of his range that was adde, coupled with us being dealt a hand that makes a call +EV.

-The problem with this is, his increased range is directly affected by the stack/blind ratio of both players. So, the higher the blinds the more likely we are to be able to take advantage of it. BUT, the higher the blinds the less equity we will have if we do win the favorable hand after folding the neutral hand.

**Basically, I think this strategy would work in a game where the cost associated with waiting (paying blinds) was lower, and/or the edge you expected to gain was larger.

[eastbay]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]

Pushing has EV difference of -.000001% of prize pool (edited this for clarity)

Jman's Theory says: PUSH
Matt's Theory says: FOLD

Am I making sense?

[/ QUOTE ]

Here's a novel idea: test your theory.

I think you will find it is false.

If you can't test it on full scale poker (although I don't see why it's not possible), come up with a simplified model that retains the properties that you think are essential to making your theory correct.

eastbay

[/ QUOTE ]

I would like to do this.

As for full scale poker, I think it would take many years to come up with a sample significant enough.

A full scale model, sounds good. Unfortunately I have no idea whatsoever how to do this. Wanna help?

[/ QUOTE ]

Not really, no.

I will just say that I have done investigations along these lines before, however, both with simpler model games as well as full preflop push/fold NL Hold'Em and have consistently found that there are no circumstances HU where taking -cEV situations is superior to some other strategy which never does. In fact, very strong strategies will collapse very quickly once you add -cEV moves to them.

eastbay

[Jman28]

[ QUOTE ]

So, your premise is that when given a situation to which we are EV indifferent, we should choose the one which will make our opponent make riskier moves going forward. If I am wrong with my thoughts so far, skip the rest.

[/ QUOTE ]

You are close, but this is a misunderstanding.

The reason we want to increase chips disparity is that it widens OUR pushing range, creating more +EV opportunities, whether we are the small stack or the big stack.

To reword, 'widening our own pushing range' is the same thing as 'finding ourselves in more +EV situations' since we should only widen our range if the extra hands in there are +EV.

[ QUOTE ]
So for this to make any sense, we are assuming our opponents are playing near optimal poker.

[/ QUOTE ]

I know you said to ignore the rest, but I think this is where your confusion might be though.

For my idea to work, we are assuming the opposite of this. We assume that WE play near optimally, and our opponent does not.

When chip disparity increases:

It becomes +EV to push more.
It becomes +EV to call more pushes.

These are both adjustments that we count on our opponent to NOT make.

For example:

Hero 4600
BB 5400

Blinds 300/600

Hero has 86s.
BB will call with 22+,A2+,KT+,QJ+
Push is +.9%

Now...take 1k and move it.

Hero 3600
BB 6400

Blinds 300/600

Hero has 86s.
BB will call with 22+,A2+,KT+,QJ+
Push is now +1.1%

This is the effect of our villain not adjusting his calling range. Pushes become better for us as the stack disparity increases.

I realize (now more than before) that the difference is minimal. I don't think this idea is a very big deal, as I said before, in that it will have much impact on your game. It will not.

[Jman28]

In addition to this last post, I want to add that if your opponent has different leaks (calls/pushes too frequently) you would want to employ a different strategy.

This time, leaning toward DECREASING chip disparity, since then he will often be making pushes and calls that are even more -EV.

The reason my basic idea is to increase the difference is because generally, our opponents leaks are not pushing/calling enough.

[microbet]

Ok, It looks like a marginal spot. Maybe a few hands could be added to the range, maybe not. Villian himself would be unlikeky to be able to answer the question.

As far as asking where you are, I was just wondering if you might be in SoCal (20 million people are).

I'm thinking about looking for a live tourney tonight. Anyone interested? Hopefully Yugo can come. He wouldn't have a date, would he?

[A_PLUS]

I think you are making the mistakes I mentioned, but I likely did a bad job explaining what I mean

I think you are confusing the EV difference (pushing vs folding) and overall equity. You mention how the same situation turns from +.9% to 1.1%. This is a direct result of the size of our stacks. We have less equity to begin with in the 2nd example, so increasing it by a larger % still makes us worse off.

We have 392$ in equity to start #1
We increase this by .9% with a push = $395.5

We have 372$ in equity to start #2
We increase this by 1.1% with a push = $376.1

So yeah, we have more higher % pushes in case #2, but I'd rather have an EV neutral push when I start with 392$ in equity than a 1% positive spot when I star t with 372$.

That is the basics as to why I think it is wrong. For this strategy to work, you would need the starting equity of #1 and #2 to be closer (392$ ~ 390), and/or the difference in EV % to be much higher.

[Jman28]

[ QUOTE ]
I think you are making the mistakes I mentioned, but I likely did a bad job explaining what I mean

I think you are confusing the EV difference (pushing vs folding) and overall equity. You mention how the same situation turns from +.9% to 1.1%. This is a direct result of the size of our stacks. We have less equity to begin with in the 2nd example, so increasing it by a larger % still makes us worse off.

We have 392$ in equity to start #1
We increase this by .9% with a push = $395.5

We have 372$ in equity to start #2
We increase this by 1.1% with a push = $376.1

So yeah, we have more higher % pushes in case #2, but I'd rather have an EV neutral push when I start with 392$ in equity than a 1% positive spot when I star t with 372$.



[/ QUOTE ]

I think you're still misunderstanding. I realized this would be a problem for some:

[ QUOTE ]

I'm a little uncomfortable with the 'sacrificing chips' wording, because it may lead some people to think that I'm saying you are better of with 30% of the chips than with 31%. You are not.

[/ QUOTE ]

I know that you are ALWAYS worse off with less chips. My point is, in a sense, that you aren't AS bad off as ICM leads you to believe. (because of the more +EV pushes you can make)

[ QUOTE ]
That is the basics as to why I think it is wrong. For this strategy to work, you would need the starting equity of #1 and #2 to be closer (392$ ~ 390), and/or the difference in EV % to be much higher.

[/ QUOTE ]

I'm quite sure that you will always be better off with more chips than with less, (all other factors the same) no matter how small the difference.

[Jman28]

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]

Pushing has EV difference of -.000001% of prize pool (edited this for clarity)

Jman's Theory says: PUSH
Matt's Theory says: FOLD

Am I making sense?

[/ QUOTE ]

Here's a novel idea: test your theory.

I think you will find it is false.

If you can't test it on full scale poker (although I don't see why it's not possible), come up with a simplified model that retains the properties that you think are essential to making your theory correct.

eastbay

[/ QUOTE ]

I would like to do this.

As for full scale poker, I think it would take many years to come up with a sample significant enough.

A full scale model, sounds good. Unfortunately I have no idea whatsoever how to do this. Wanna help?

[/ QUOTE ]

Not really, no.

I will just say that I have done investigations along these lines before, however, both with simpler model games as well as full preflop push/fold NL Hold'Em and have consistently found that there are no circumstances HU where taking -cEV situations is superior to some other strategy which never does. In fact, very strong strategies will collapse very quickly once you add -cEV moves to them.

eastbay

[/ QUOTE ]

Here's one for you.

You're playing against opponent X. You've been playing with him for 6 hands heads up.

Hero: 6000
X: 4000
Blinds 250/500

Opponent X is a tall slender man with a rugged handsome face. He has been folding every hand except for AA.

He's waiting for those aces baby! And he's gonna bust you so good when he gets em.

Now you are dealt 94s in the sb. He will only call with AA. However, this opponent X has decided that if you push 4 times in a row into him (you've already pushed 3) that he will adjust his range for the rest of the tournament to calling and even pushing himself with 22+, Ax, Kx, Qx.

Clearly, the optimal strategy is to fold this 94s, even though pushing the hand is +cEV. Then push the next three chances you get, then fold again.

Now, in real life, examples aren't this clear cut. They are more like the one's which I am trying to describe in this thread.

Would you mind opening your mind and thinking about them now that I have shown you how +cEV plays are not the optimal strategy 100% of the time? I would like to hear your thoughts on the idea based on it's own merit rather than based on the fact that you have tested different situations and come up with the conclusion that what was +cEV in those situations was always best.

I would be glad to help set up a simulation if anyone with the knowledge to do something like that would assist me.