Harrington on "bubble" EV

[SeriousStudent]

I played in the Shooting Stars limit tourney today, and had a chance to congratulate Action Dan on his book. Then I asked him about a specific hand played in the PP Million Guaranteed NL tourney on Saturday ( see PPM AK bubble hand thread for the details).

To figure out the right play with that hand, one needed to be able to determine what the EV of all places in the tourney would be for a given stack size. Dan said that some people have attempted to develop an algorithm for such a calculation, but there is no reliable one. Has anyone here attempted to do so?

Summarizing the hand, I had AK in MP3 with 1.8 BBs, and was absolutely certain to make the money if I folded ( there were 223 left and the tourney paid $900 at 220). Should one fold, call or go all in?

Your EV for folding was $900 plus the EV of placing higher given a stack of 1.8x the then current BB, with 5 hands until the BB arrived. The EV if you won the all in, assuming the BB called with his 4:1 pot odds, would be $900 plus the EV of placing higher given a stack size of t6000 or 4x the current BB.

If you assume the AK had a 60/40 chance of winning the all in, then the
EV of winning would have to be 1.5x the EV of folding, or ($900 + EV t6000 above $900) greater than or equal to ($900 plus EV of t2700 above $900) times 1.5. One would have to believe that the EVs were not linear with stack size ( I do think this is a correct assumption as t6000 in this case is worth much more than 2 times t2700).

Let's say that the EV of t6000 is 3x the EV of t2700. For it to be break even, the EV of all money spots above the $900 for t2700 would have to be $300, and for t6000 it needs to be $900. Is that at all reasonable? You have to place at 90 from the 220 still in the tourney to get $1800 ( $900 more than you already are guaranteed).

If you use the rule of thumb for final tables ( don't think it is accurate at these very small percentages of chips in play but just as a reference), t6000 is 0.2% of all chips outstanding times the 275k first prize - $504, and 2nd $205, etc. How would you calculate the chance of placing at least 90th which pays $1800? Better yet the sum of all place EVs?

By the way, it was reassuring to hear than Dan felt the EV was higher to go all in than to fold, and that he thought most pros would do so ( which is what I did and of course finished at 222 out of the money against the straight from the BB 32suited hole cards).

[pokerraja]

I'm sure calling for Dan is +EV, a $300 score is not worth his time, as most high level pros. But for many, and myself included this is an easy folding situation.

[UOPokerPlayer]

I don't think I can disagree with Dan. The bankroll is not a consideration for him though. Alas, the argument of sacrificing EV lives on.
Any other pearls of wisdom you picked up from him? What was he like? I read he is notorious for saying absolutely nothing at all.