Sklansky-Chubukov Rankings

[Barry5]

Apologies if this is in the wrong place.

I've just come across these rankings in the new Sklnsky/Miller book and amintrigued by then and the potentiqal impact on MTT and SNG play. One of my weaknesses is being too timid late on and the implications of following the advice seems quite profound to me.

Anybody any thoughts?

[IowaSkinsFan]

what are the new rankings?

[Barry5]

what are the new rankings?

In a (large) nutshell hands are given a number (KK is $954 whilst 32 o is 1.8) and they are aimed at giving you a figure that helps you decide how good a "move-in" hand you have. The figures are based on what your opponent would do if they saw your hand face up.

There are a number of pages in the book which devlop the thought processes but, of course, it being Sklansky, they are based on quite complex mathmatical calculations.

They are also aimed at short stack play so if your stack is less than the number for the hand (adjusted for position and blinds value) it tells you when to push. I was going to try it last night but, fortunately, never got short stacked enough.

[jameseyb]

Any chance you can scan it up so we can have a look? Otherwise, is the book worth buying for that alone?

[slumpy]

Does anyone have or use these rankings?

[BankItDrew]

The figures are based on what your opponent would do if they saw your hand face up.

I believe you may be leaving some details out. I'm pretty sure 99% of hands have zero value when they are played face up.

[Ragnar4]

1: You are in the SB for in a 1-2 game, everyone folds to you and you accidentally show your hand: AKo. Unfortunately, your opponent is a computer who flawlessly determines the best play now that it knows your hand.

You decide that this is a position where you must either A) Move all in, or B) Fold, for what values X (Where X is the size of your chipstack) is it better to move all in, and for what values X is it better to fold.

The math goes even further, but suffice it to say that in a 1-2 game where A) a perfect oppoent simply folds and B) a perfect opponent calls with any sort of edge here, slight or huge, that X is $332 dollars. For all values less than 332 dollars (or roughly 166 BB's) you should shove your money in, for all values more than 332 dollars you should play the hand normally.

*THIS IS NOT THE OPTIMAL PLAY* Just a compelling argument for SnG's when the blinds are extremely high, to define a perfecly mathematical breakeven point, and create a slight edge against superior players

Some interesting hands in a 1-2 game
KK $954
88 $159
JTs $36
RIPPYTYDE! 74s $5.

The chart is presented rather poorly in a "1-2 NL game" paradigm, but would be very easy to break down in X-1 type of odds.

So yes these hands do have 0 value when played face up, except for the part where poker is nothing more than a strugle for the blinds.

[Stacks]

think it would have been a shitton better if they did it in bbs rather than stack size in porportion to a 200nl game.

[bjsaust]

I've never really understood how this is in any way helpful. I mean...maybe in a live game where you actually plan to flash your cards at your opponent?

[thizzSantaCruz]

because an open shove with the given hand and stack sizes will always be profitable.

I've never really understood how this is in any way helpful. I mean...maybe in a live game where you actually plan to flash your cards at your opponent?

Knowing these numbers you can see whether going all in with 92o from the button in an MTT is good with your stack size. Or some shit.

[Ragnar4]

This only works 100% for the SB, and then only partially for the BB.
The math is waaay different when you're trying to fold 2 people and not only just one.

What you could do is level: Say, I have K4o and then shove, and when he folds because he thinks you have AA, you show your hand. Get them lulled into the idea that A) you're telling the truth, and B) that you're a terribad fish, when actually you're playing the percentages.

[Stacks]

This works from the BU too I believe. In NLHET&P it talked about this and it went on to say that from the BU it's best to take the # and divide by two then use that as your number. Therefore, with AK if you stack is somewhere around $166 then you can effectively shove worry-free as it's unexploitable.

And that's really what this is getting at. Making an unexploitable shove. This goes right along with Spoons "SB vs BB" thread a little while back. If you are in SB and shove a hand against a player that will only play against your hand if he has an edge, then following these numbers you will always be making a +ev move because he won't have a hand often enough to call, and even when he does you generally still have some equity.

Let's say your hand is K8o. I'm not looking at the numbers right now so I'm not sure. But depending on your stack he is gonna need something with close to 50% equity. Say if you shove 20bb he will need 48.2% equity, and so on with other stack sizes. Against K8s his range would be something like 22+, K8s+, K8o+, A2s+, A2o+, and possible some suited connectors. Therefore, he will not have a profitable range to call you with often enough, and therefore you scoop the blinds enough and win when he does call often enough to make this shove +ev and unexploitable because even knowing your hand you win money in the long run.

Now it talks about it being unexploitable because even if he knows your hand you win. It's even more +ev when he doesn't know your hand. If you shove alot of players will fold hands like small pps, weak aces, K8, and some suited connectors that in reality he should call with. So when you add in the mistakes he makes, which effectively make his range tigher, you win even more money.

[Ragnar4]

It all depends on your opponents range. If he limps 100% of the time on the SB when it's been folded around, then the standard chart applies right?

But if your opponent limps 75% of the time here, you need to multiply the S-C numbers by .75 in order to find the correct stack.

It seems to me that the reason this is unexploitable, is because you and your opponent don't know your opponents hole cards at the point at which you shove. But once they've made the decision to look at their cards, the math is completely effed up.

If your opponent has made the decision to limp, at this point, his hand is already probably better than any 2 cards, "who limps 73o in the SB.. seriously?"

edit -- 1 BTW, if your opponent does know your cards, this just seems like the highest variance coinflip EVAR.

edit -- 2 Just how big would your bankroll have to be in order to make this a +ev play that you could stand the swings at a NLHE table?

[Stacks]

Here some math I've already done. The scenerio: It's folded to you in the SB with A5o. You accidently flip your hand over, and your opponent now sees your hand. You still wonder if you should shove, even knowing your opponent will only call if he is a favorite against your hand. Should you fold or should you shove?

Okay. So according to Sklansky - Chubukov Rankings, A5o has a number of 56.54. This is regarding a 1/2 game so effectively 28bb (56/2). This means if you have 28bb or less you can profitable shove even if he knows your hand. If you have 29+bb then you should fold (or at least don't push). Let's look at the math and see.

Okay assuming you have 28bb this is the math. If you shove, opponent is getting 1.05:1 pot odds (29.5/28), or 48.7%. Therefore, he needs at least 48.7% equity against your hand to be breakeven.

The only hand that has that much equity is [22+, A5s+, A5o+]. And this range is 16.7% of hands. Also, this ranges equity against our hand is 66.29%. With some math we get.

*we shove, they fold (profit of 1.5bb): [1 - .167] * 1.5 =1.2495
*we shove, they call, we win (profit of 29.5bb): [.167] * [1 - .6629] * [29.5] = 1.66
*we shove, they call, we lose (loss of -28bb): [.167] * [.6629] * [-20] =-2.21

*So: 1.2495 + 1.66 + -2.21 = 0.6995bb/hand

______________________________________________

Let's see what this is like with say 29bb, instead of the 28bb.

With 29bb, the BB is being offered 1.0517:1 pot odds, so he needs 48.74% equity. The range will still be the same [22+, A5o+, A5s+]. So we still have this as 16.7% of hands with an equity of 66.29% against A5o.

*we shove, they fold (profit of 1.5bb): [1 - .167] * 1.5 = 1.2495
*we shove, they call, we win (profit of 30.5bb): [.167] * [1 - .6629] * [30.5] = 1.7170
*we shove, they call, we lose(loss of 29bb): [.167] * [.6629] * [-29] = -3.2104

So: 1.2495 + 1.7170 - 3.2104 = -0.2439bb/hand

Assuming my math is correct there (which I hope it is) with even just 1bb over the number you will be making a -ev move. This is to say that if the blinds were say 50/100 and you have 2800 chips you could shove A5o into the bb, even if you showed him your hand. However, if you have 29bb if you do so you will be losing money over the long run.

[Ragnar4]

does this become much much much more +ev if your hand is significantly less than the "break even point" ? IE say you only have 15 bb here?

What's more. Why is A5 o in your opponents range? Because it's a freeroll?

[Stacks]

Let's look at it with 15bb really quick. Pot odds of 1.1:1 (16.5/15). So he could only call with a hand that has 47.62% equity. The range now is [22+, A50+, A5s+, T9s, JTs], which is 17.3% of hands. This range has 65.51% equity against A5o. So:

*we shove, they fold (profit of 1.5bb): [1 - .173] * [1.5] = 1.2405
*we shove, they call, we win (profit of 16.5bb): [.173] * [1 - .6551] * [16.5] = 0.9845
*we shove, they call, we lose (loss of -15bb): [.173] * [.6551] * [-15] = -1.699

So: 1.2405 + 0.9845 - 1.699 = 0.526bb/hand

So, looking at the math it wouldn't seem that's the case. My assumption here is givent he smaller stack, which leads to the smaller equity their hand needs, allows other hands such as JTs,T9s to call profitably. This means that more hands can call, but you still don't win often enough. I'm thinking that at around 8bb or so it would be more profitable as scooping the blinds become more important, but I'm not sure.

And regarding him calling with A5o its because he has more than the equity needed. I mean he wouldn't fold knowing he will tie because then he would lose .75bb/hand (small blind + big blind).

[spoonitnow]

does this become much much much more +ev if your hand is significantly less than the "break even point" ? IE say you only have 15 bb here?

What's more. Why is A5 o in your opponents range? Because it's a freeroll?

Villain's range is determined by perfect play based on him being able to see your hand. That is, if he has the correct pot odds to call, he does so. If not, he folds.

There are lots of uses for this model, but one of the most used is that it helps to provide starting points for short-stackers to form their pre-flop ranges.

[bjsaust]

No, it tells you whether its profitable to do so if your opponent knows your cards. It doesnt tell you whether its profitable to do so v's any kind of calling range which is much more useful.

This is just weird.

[Stacks]

As I said bj, if it's profitable if your opponents know your cards, it's even more profitable if they don't.

Example: I'm in SB with 15bb and have A5o and shove. You are in the BB with 22 and 15bb. Do you call? Probably not. However, that is a mistake as your hand is the favorite. Now if you saw my hand I would hope you would call. It's calling mistakes like this that makes it even more profitable when they don't know your hand.

This goes for alot of hands. Same scenario, but I shove K4o. Your gonna probably fold K5o, however, that is incorrect.

[spoonitnow]

No, it tells you whether its profitable to do so if your opponent knows your cards. It doesnt tell you whether its profitable to do so v's any kind of calling range which is much more useful.

This is just weird.

I've noticed that a lot of players have similar attitudes.

[bjsaust]

If I have JTs and flash you, its profitable for you to call with Q4o.

How many people are calling with Q4o?

If I push with A2o and flash its not profitable for you to call with KQs, but how many people are folding KQs for say 10bbs?

It just gives you some really strange results that are not much use in real games. I mean, if you'd prefer to use a chart rather than learn ICM and ranges its as good as anything, but it seems extremely artificial to me.

[Ragnar4]

Ok, now lets say that you shove with your JTs and DON'T show your opponent.

Your opponent tightens way up, waiting for a dominant hand to call you.

He's made a serious error because he folded when he should have called, and even if he does call with that dominant hand, you draw out on him enough to still make the shove worth your while

If your opponent played perfect poker, (which he doesn't) you would still always make money. It's unexploitable.

This chart only applies to the SB, and ICM is still applicable.

[bjsaust]

But I choose to push JTs because of the range of hands my opponent calls with (hence the % of times he folds) and how my hand fares v's that calling range. People SHOULD fold more when I push A2o if I flash them, but if I DONT flash then they're calling with the same frequency, but A2o is much more likely to be dominated by their calling range. So JTs gets just as many folds, but has more 'suckout' potential when called than A2o.

So ICM tells us to push JTs and fold A2o, but these rankings would say the opposite.

It cant possibly be more +EV to define your pushing ranges on an imaginary scenario of your opponent knowing your cards, than to define your pushing ranges on his calling ranges if he DOESNT know your cards.

But, as a starting point for people who either cant be bothered learning P/F ranges and ICM, or just havnt had time yet and need a starting point its better than nothing I guess. I just cant work out why its held up as something extremely useful so much.

[spoonitnow]

But I choose to push JTs because of the range of hands my opponent calls with (hence the % of times he folds) and how my hand fares v's that calling range. People SHOULD fold more when I push A2o if I flash them, but if I DONT flash then they're calling with the same frequency, but A2o is much more likely to be dominated by their calling range. So JTs gets just as many folds, but has more 'suckout' potential when called than A2o.

So ICM tells us to push JTs and fold A2o, but these rankings would say the opposite.

It cant possibly be more +EV to define your pushing ranges on an imaginary scenario of your opponent knowing your cards, than to define your pushing ranges on his calling ranges if he DOESNT know your cards.

But, as a starting point for people who either cant be bothered learning P/F ranges and ICM, or just havnt had time yet and need a starting point its better than nothing I guess. I just cant work out why its held up as something extremely useful so much.

You don't necessarily know his calling ranges either. When you show him your cards, it doesn't matter what his calling range is, which is the point and why it's so useful.