Is there any argument to be made for calling here?

[eastbay]

$55 PP SnG. 3 left. I'm BB, blinds 250/500. Short button goes all-in for his 1300, it's 800 to call. I've got 1600 after posting. I've got J9s.

eastbay

[Jman28]

I'd plug it in the ICM calc against a range of hands if I wasn't playing right now.

Obviously this is somewhat player dependent a little bit.

I'd bet that in most cases it's -$EV, but not by that much.


-Jman28

[stripsqueez]

i dont think its much about EV

if you fold then the other guy becomes a favourite for second - your chances of winning start to become remote

if you call and win you come at least second and you are only a small underdog for first - if you call and lose your desperate but a decent distance from dead in the fight for second

these sorts of decisions remind me of classical positions in games such as bridge and chess - assets have a shelf life and recognising the optimal time to spend an asset is the skill

i think its close - at the table i think i would fold if i pegged the other guy as being too tight which is a common state in this game at this time (this also works because i might be given the chance to steal the BB next hand) - against many players your some chance to be in front pre-flop and J9s is not going to a big dog to many hands

stripsqueez - chickenhawk

[texasrattlers]

Wow. This is a really close one. I like to have chips to play with, so I probably fold and then push against shorty next hand if you get the chance. Calling may be the better play but I don't have the guts nor the smarts to make the call.

[texasrattlers]

[ QUOTE ]
these sorts of decisions remind me of classical positions in games such as bridge and chess - assets have a shelf life and recognising the optimal time to spend an asset is the skill

[/ QUOTE ]

I've played chess but not bridge. What the heck does this mean?

[stripsqueez]

[ QUOTE ]
I've played chess but not bridge. What the heck does this mean?

[/ QUOTE ]

i dont play chess (i do take an interest in and have some understanding of chess theory) so i couldnt give you a chess example

as for bridge the idea is best demonstrated by declarer play problems - in a broad sense a declarer starts with and accumulates assets - in bridge those assets are things like high card strength, entries, threats, and information - how and when to spend those assets seperates good from bad - sometimes they must be spent as soon as they are obtained because they will otherwise lose all or some value - sometimes they will gain value over time and are spent optimally by waiting - if you played bridge at a good standard i could give you examples

perhaps an easier way to make the point is by a straight forward examination of your chances of finishing first second or third in the tournament given the various scenarios - that is a more mainstream way of considering the problem

i cant be arsed going through all the scenarios and in practice you cant do that at the table - if you call the push and win the result is huge - if you call the push and lose your chances are severly damaged - contrast those scenarios with the position your in if you fold to the push - in which scenario do you make more money ?

the chips are your assets - are those assets optimally spent now or is their value higher by waiting ? - same question expressed differently

stripsqueez - chickenhawk

[ChrisV]

This talk of chess and bridge is a distraction. I know the bridge analogies stripsqueez is trying to make but there is no good analogy I can think of in chess.

I think stripsqueez is talking a bit abstractly about a point that can be made more simply. There are situations where you should make a play that is negative chip EV because it's positive $EV. These situations occur when you would rather have a shot at a big stack than be assured of a small stack, even if your average chip amount goes down by trying for the big stack. Commonly this occurs when the person stealing from you has a similar stack size and is your major rival. The steal therefore hurts you doubly, making a successful steal very damaging to your $EV. For example, if you have a stack of 700, post up 200 in blind, and are put all in. Even though you may not have pot odds to call, you might choose to call anyway if you think that your remaining stack of 500, posting 100 in SB, has almost no chance of getting you extra cash versus your rejuvenated opponent, whereas your winning stack of 1400 will crush your crippled opponent.

Another way of expressing it, and the way stripsqueez chose, is to say that your 500 remaining chips have higher value this hand than they will in subsequent hands. I think that's a bit of an abstract way of putting it /images/graemlins/smile.gif

[ChrisV]

I should add though, I think this is a super, super easy call even without invoking the above principle. SB rates to have a completely random hand - I know I'll be pushing with any two as long as there's guys considering folding J9s. J9s vs a random hand is 55%, which hands you a whopping +630 in chip EV.

[stripsqueez]

[ QUOTE ]
There are situations where you should make a play that is negative chip EV because it's positive $EV

[/ QUOTE ]

you mean like all the time ? - i make negative chip EV decisions early in a tourney by refusing to take marginally positive chip positions - i make negative chip EV decisions late by taking marginally negative chip positions

your right that my explanation is a bit abstract (and i think your explanation is as usual accurate and well expressed) but i have issues with tournament theory - given that we agree that chip EV isnt a good judge of what to do why is it common for analysis to even refer to chip EV ? - could it be that this is simply a comfort zone of certainty that flows from accepted ring game theory ?

whilst i'm venting abstractly - what does the term "$EV" mean ? - i mean i know what it means - it means is it a good decision or is it a bad decision - i reckon the whole poker world has gone EV crazy

stripsqueez - chickenhawk

[ChrisV]

Well sure, FOLDS which are -CEV/+$EV come up a lot. Less common are situations where you CALL at -CEV.

Chip EV is used as a tool because its an easily calculated, standard starting point for discussion of the hand. You then apply whatever modifications are necessary. Much like point count at bridge /images/graemlins/wink.gif. In my post above where I said that call is +630 chips, for example, I didn't think it necessary to go any further because the size of the chip EV I think precludes any more discussion. It's pretty rare that you decline a betting opportunity when the expected return is almost double what you're putting in.

$EV is just shorthand for "play which wins you money" or "play which loses you money", distinguishing that from chip EV - what terminology would you prefer us to use? /images/graemlins/wink.gif

[etgryphon]

What I understand the whole +chipEV and the -$EV is when you up against the decision where the pot is laying you the correct odds for most if not all your chips, but you are a dog to win it.

But it doesn't seem to make much sense that you can have a +chipEV but a -$EV I can only see it the other way around (i.e. -chipEV with a +$EV) It is an abstract concept over the short term because you can't reload in a tournament.

I haven't seen where we have really discussed whether making these moves over the duration of tournaments will be better than making the correct "tournament strategy" play.

-Gryph

[ChrisV]

As stripsqueez pointed out, half of SNG strategy rests on the idea of +CEV/-$EV plays. To take the simplest, standard example, suppose you have 5 chips on the button, the SB and BB are both allin for 250, and UTG calls. You have AA. Calling is clearly the chip EV play, but is insane in terms of $EV. It doesn't matter how many tourneys you run it over, calling will still be crazy.

Similarly, it doesn't matter how many tourneys you run it over, going allin on the first hand with a 50.1% chance of winning will still not be correct. Doubling your stack doesn't double your $EV. The Independent Chip Model (http://www.bol.ucla.edu/~sharnett/ICM/ICM.html) for example, gives your $EV as percentage of the prize pool as 18.44% with 2000 chips (and, obviously, 10% with 1000 chips).

[etgryphon]

I'm sorry Chris...I'm a little slow...

Is it a poor $EV play because in the AA example we are not winning enough chips to make the chips get us a better chunk of the prize pool (i.e. the chips lose value as you get more of them...)? But, it is the correct move because we will amass more chips because on average we'll be ~60% favorite to win more chips which is the goal? If I'm being clear...

Also, do we consider folding to be a +chipEV move?

Thanks for explaining this...its a new concept.

-Gryph

[tallstack]

IMO, you are getting way too much of an overlay to fold here. Your cards are likely live and suited - they are just not that much of a dog to anything but monster hands. You may even be able to discount monster hands if you believe that your opponent would slow play them to increase his chance of doubling up. You cannot be busted if you lose, and if you bust the button you have a stack that you can work with HU. I also did some ICM and pot odds calcs and they are both in the 2.4:1 to 2.6:1 range. It is hard to imagine that any positional or skill effects can work enough against these odds to make this a fold.

Dave S

[rachelwxm]

From ICM
Folding 28
losing Allin 24.1
Win all in 37.3

There fore you need to only win 30% of time to justify the call.

The only thing hold back the call is that you might think you can steal back from the other short stack next hand. But now that you are equal stacked, I am not surprised that big stack is pushing with any two next hand. /images/graemlins/smile.gif

[eastbay]

[ QUOTE ]
I should add though, I think this is a super, super easy call even without invoking the above principle. SB rates to have a completely random hand


[/ QUOTE ]

Well, except that he doesn't. You are giving too much credit. He rates to have at least Q high.

eastbay

[eastbay]

[ QUOTE ]
$55 PP SnG. 3 left. I'm BB, blinds 250/500. Short button goes all-in for his 1300, it's 800 to call. I've got 1600 after posting. I've got J9s.

eastbay

[/ QUOTE ]

A little more context: Button was way too tight, so I was almost certain I was facing Q high or better, and probably a lot better. Second, this tightness meant that I was taking his blinds very easily. Because of this, ICM and similar analyses are overvaluing a call and undervaluing a fold. This is because ICM has no concept of positional advantage and folding equity. It overvalues a small stack with no FE left, and undervalues my skill advantage in blind stealing.

These two factors made it close, but I called anyway. He showed KQ and I got no help, and went out on the next hand.

In retrospect, I think I fold and continue to take his blinds. Could go either way, though.

eastbay

[The Yugoslavian]

$EV is basically the value of the prize pool you can expect to win/lose due to your decision. CEV is the amount of chips you expect to win/lose due to your decision. There are many decisions in a 50/30/20 payout structure where a decision in +CEV but -$EV. Frankly, on the bubble, many decisions can be +CEV but -$EV. For instance, let's say the small stack has only 1BB worth of chips left. Going allin against a stack that has you covered is a horrible idea most of the time because you will stand to make 3rd place most of the time just by folding.

ICM tries to provide an objective comparison between CEV and $EV by estimating the value of your chips (many assumptions are made for the algorithm though) at any given point.

I know Eastbay was thinking about running short handed numbers to see if they jive with ICM (since many people doubt how useful it may or may not be). How is that project going Eastbay?

Anyway, you always want to make the best +$EV decision while you only sometimes want to make the best +CEV decision. The tricky part is that calculating CEV is relatively easy and accurate while calculating $EV is much harder.

[etgryphon]

Ok...

Is this the whole strategy of forgoing a good hand to put the small stack out of the SnG because you can be stealing chips and thus be making a +$EV moves on the other players?

or

Is the this the strategy of not going in three handed with a SS and the guy who has you covered and possibly losing and sending the SS ITM?

-Gryph

[Sidekick]

etgryphon I haven't seen a response to your question yet, so I will try and answer it as best I can.

In the example that ChrisV made, the reason that you want to fold in a +chip EV situation (even with a hand such as AA) is that by folding you are almost guaranteed to finish in the money by folding. The only situation where you don't is if both the other two short stacks manage to split the pot.

You are down to four players and the two other short stacks are all in. One of them is going to bust out in 95-99% of the hands. If you play the hand you can quadruple up your chips should you win, but you will still finish out of the money most times due to having such a short chip stack.

Also, in this extreme example you could wind up in second place easily by folding and sitting the hand out. That means that on one hand you go from most likely losing your buy-in to quite possibly taking 2nd place when you have 5 chips in front of you. This is a huge +$EV play.

Hope this helps.

[Marcotte]

[ QUOTE ]
This talk of chess and bridge is a distraction. I know the bridge analogies stripsqueez is trying to make but there is no good analogy I can think of in chess.


[/ QUOTE ]

If I understand this discussion (which isn't necessarily a given), then a chess example would be some sort of sacrifice or gambit. It could be as little as trading your bishop for his knight (on c6 or f6) or as big as sacrificing your queen to destroy his castled position.

[texasrattlers]

Sorry to pick up on this "distraction", but it really interests me.

It seems the key question in eastbay's situation is the CEV vs $EV question, as elucidated by several posters. However, I do not think the CEV vs. $EV has any application to chess. So, I do not think you can say sacrificing your Queen to destroy your opponent's castle position is negative CEV or negative anything because you are doing this because you have a well thought out plan of future moves that will ensure victory. So, just because you lose a "10 point piece" doesn't make it negative anything because all that counts is winning and the loss of your queen fits into your plan for winning. In a SNG your total points (chips) have value in and of themselves -- if you have a lot you can ease your way into the money or bully your way to 1st. A similar concept must apply in bridge. But in chess your points are meaninless -- all that matters is how your "assets" (whatever their point value) are strategically positioned on the board.

[stripsqueez]

[ QUOTE ]
Chip EV is used as a tool because its an easily calculated, standard starting point for discussion of the hand. You then apply whatever modifications are necessary. Much like point count at bridge

[/ QUOTE ]

my current bridge partner includes in the pre-alerts to our opponents "he uses the point count ranges on the card as a guide to his usual bid" - point count is a crappy tool because it doesnt describe accurately enough the value of high cards - i think the history of point count reveals that it exists in bridge theory because it makes the game easier to teach

someone has posted after running a computer simulation that its notionally right to call in this problem if your a 30% chance to win - my gut tells me its closer to 40% - chip EV is based on anything above 50% - seems to me that chip EV is a crappy tool too

[ QUOTE ]
$EV is just shorthand for "play which wins you money" or "play which loses you money", distinguishing that from chip EV - what terminology would you prefer us to use?

[/ QUOTE ]

how about right or wrong

stripsqueez - chickenhawk

[ChrisV]

In the end chess is a deterministic game of complete information, so any sound sacrifice must necessarily be based on a winning series of moves.

However, while the top players can analyse variations many moves deep, even they often make decisions based on abstract considerations, patterns in the game. In a chess book I read once, Spassky commented on a knight sacrifice he made, saying it was "one of those sacrifices that need not be analysed thoroughly, since it is absolutely clear that Black is not in a position to defend his king". I think that if you want an analogy with chess, a sacrifice for an abstract positional advantage is a good one. In both cases the player is recognising that the material, mathematical advantage he is giving the other player is illusory and that actually he is the one getting the advantage.

It's a half truth to say that in SNGs your chips have value of themselves, just as it's a half truth in chess to say that pieces have value in and of themselves. Chips always have a value of some kind, but the value changes dramatically over the course of the SNG and not always in linear ways. For example, in the AA hand I gave earlier in the thread, your 5 chips have immense value because they're keeping you alive. However additional chips (the 15 you stand to win if you call and your AA holds up) have virtually zero value.

By the way, if anyone is interested in abstract strategic thinking in chess, I recommend the audio commented games of Josh Waitzkins in Chessmaster 8000 & 9000. He does the variation analyses so you dont have to, but more often he comments on the abstract ideas behind moves. I personally found it a lot more interesting than actually playing chess.

[Marcotte]

Admittedly, any comparison b/w chess and poker (or bridge) will be strained at best, simply because chess is a game of perfect information and card games are not.

Perhaps a better analogy would be where you sacrifice a piece to gain a small (i.e. not game winning) positional advantage - something like giving your opponent doubled pawns or a weak color complex in one sector (e.g. all queen side pawns on black squares).

So your ChipEV (# of pieces, their value, and their positional value) might go down, but your $EV (ie chance of winning the game) might go up.

Conversely, in poker you may fold J9s in your SB or BB to a Button or CO all-in because you will have a positional advantage in the next few hands.

Apologies for fueling the 'distraction', but I love chess. /images/graemlins/grin.gif

[ChrisV]

[ QUOTE ]
my current bridge partner includes in the pre-alerts to our opponents "he uses the point count ranges on the card as a guide to his usual bid" - point count is a crappy tool because it doesnt describe accurately enough the value of high cards - i think the history of point count reveals that it exists in bridge theory because it makes the game easier to teach

someone has posted after running a computer simulation that its notionally right to call in this problem if your a 30% chance to win - my gut tells me its closer to 40% - chip EV is based on anything above 50% - seems to me that chip EV is a crappy tool too

[/ QUOTE ]

Absolutely - which is why I made the analogy, I think it's a good one. Even though point count is a pretty awful way to rate bridge hands, people still define their bids in terms of point count because it's a standard way to talk about hand strength. Chip EV is a bad way to talk about SNG decisions - the independent chip model isn't all that great either - but both of them are approximate measurements, data points that can be incorporated into an analysis.

[texasrattlers]

[ QUOTE ]
Chip EV is a bad way to talk about SNG decisions - the independent chip model isn't all that great either - but both of them are approximate measurements, data points that can be incorporated into an analysis.

[/ QUOTE ]

So, if it's a bad way to talk about right vs. wrong decisions, why do it? If you start w/ questionable assumptions I don't have a lot of confidence in conclusions drawn from those assumptions. I have never much appreciated "It's a bad way to look at things, but it's all we got" arguments.