Saturday, May 03, 2008

SNG Strategy And The "Kill Phil" Premise

Off to the in-laws this morning for a weekend of getting over-fed, offered too many drinks and generally not understanding a word anyone is saying. Thinking of using the break from routines to have a longer-term / bigger-picture business planning session.... we will see.

Here is something to think about.

The book 'Kill Phil' is all about leveling the playing field against better players by making them either fold or lay down all of their chips to your all-in bets. It does not give you an inherent advantage in a poker tournament to use this system, more negates the skill factor that very good opponents have in their favour over time.

One of the early ideas in this book explained why a 'pro' might decline a coin-flip with the odds in their favour early in a tournament.

The idea itself is simple.

If your average expectation for the tournament as a whole is 35% (made up numbers to illustrate the point) and a situation early comes up with a clear margin in your favour of 20% - that would eliminate you from the tournament if you lose... then the pro would actually be taking 15% the worst of it - when compared to their expecation of playing many small pots instead of gambling on one big one.

Of course - you'd need to be pretty confident in your ability to grind your stack up to turn down a nice edge like that... but the logic of the argument is clear - in Sklansky terms this is turning down a bet today in order to have more profitable bets tomorrow.

So, all very nice - but how does this relate to SNGs?

In several ways actually. Firstly let us think about those very early game situations where you are forced to make a decision for all of your chips. Say, for example, that you consider yourself exactly 50/50 vs your opponents range - yet there is 200 in 'dead money' in the pot.

Calling would be acceptable here right (with 1000 stacks, on average you will win 1200 for every time you lose 1000 here - making your average profit 100 chips... amost 10%).

But wait - what if your average expectation in this level of SNG was 20%? Would it still be'correct' to take a 10% (approx) 'gamble' early on???

For me there are two lines of thinking here... one that such bets are already included in our ROI, secondly that the majority of profits are derived from the bubble (or more accurately your opponents errors at the bubble) which makes it even more important not to gamble early.

Something to think about.

Will relate this same concept to later game play in a future post.

Cheers, Mark

No comments: