Sunday, April 29, 2007

SNG Strategy - Playing for 1st??!?!?!?

So many posts on so many forums discuss SNG bubble play, justifying that horrible call by reasoning 'well you have to play for 1st right...' WRONG... Wrong wrong wrong wrong wrong... its so wrong it hurts to see it written.

And you can not argue with these folks either - oh no, playing for 1st makes such good sense to them that argument will be rebuffed not by logic or reason - but by personal insults!

Actually I do not argue with them, it is not only pointless but my self-esteem is not actually centred on whether or not I win a forum argument... flame me all you want, I'll just go and make a nice cup of tea!

The other line of argument is ' play to cash - then go for 1st'. I feel this argument is actually misunderstood, it happens to be right if you look at Bubble Dynamics in terms of $ Equity and ICM... but it is only right by default.

So, what do I mean by this... well, we will start with the 'correct' (IMHO) line of thinking for the Bubble of a SNG and work backwards. I do not play for 1st, or play to cash... no no, its actually much simpler than that:

Make Positive Expectation Decisions In Terms of $ Equity.

Over and over again, one +$ev decision after another... ICM is 80% of the battle here, the other 20% is knowing when to ignore ICM! For example if I have a +1.3% push but the next big blind will put shorty all-in then its an easy fold... comes down to the 'passing up a small edge today to have a bigger edge tomorrow' type of thinking.

Now let us look at the nature of these decisions in terms of the play for 1st and play for 3rd arguments:

How does 'playing for 1st' fit to making positive expectation decisions at the bubble? Well to be honest it does not. A call of an all-in from an opponent that will give you the majority of the chips is either +$ev or it is not. The key point: if you make -$ev bubble decisions you will lose money over time regardless of the outcome of the hand in question...

So playing for 1st can only really be achived by calling an all in from another big stack right? if you think about it raising all in can not be considered playing for 1st - since unless your bigger stacked opponents are of the same school of thought they will usually fold rather than take a negative $ev gamble.

I usually fold too... if you risk $30 in equity to win another $10 you are laying huge odds-against yourself... think it through... 'gambling' for that 1st place might be costing you $10 or $15 per try!! Whether you are fighting a big stack or a smaller one then a call / push either has a positive expectation or it does not, playing for 1st to justify taking a negative $ev chance will quite simply cost you money over time. If your call has a positive expectation (you hold Aces!) then you are not 'playing for 1st' any more - you are simply making yet another positive expectation bubble decision.

Now playing to cash, lets look at the old argument that the jump from 4th to 3rd is as big as the proze jump from 2nd to 1st in a standard 20% / 30% / 50% payout SNG. This actually has a big effect on the expectation of our decisions, the reason is that the next person out gets zero and so the baseline for winning a showdown is a minimum of 20% equity (the extreme case being that you have 1 chip left and 2 others fight without you in the hand).

The prize pool dynamic is factored into the maths already, when we are assessing whether your push or call is + or -$ev then your current equity and the 'zero if you lose' are on the two sides of the balance sheet. In this way you are forced to 'play for 3rd' before considering the higher prizes... but (and an important 'but') your decisions are not made by some conscious effort to get 3rd place - your decisions are made based purely and simply on whether your action has a positive or negative $ expectation.

Here is the summary - stop thinking about playing for 1st or playing for 3rd - just make positive expectation bubble decisions again and again and again and you will win $ over time... those +ev decisions are mostly based on ICM but also include situational factors such as very small stacks being involved and your assessment of the likelihood of the other people in your game knocking each other out.

if you are a new reader wanting to understand the logic of ICM and positive $ev decision making then I'll direct you to the 'jump off page' for ICM in the list of Plan3t Gong articles on the right hand side,..

GL at the tables,

Mark

No comments: