Just mulling over this at breakfast this morning. Some dumb HH to illustrate the point...

Here if we put button on shoving 35% of hands, ICM calcs say we are slightly +ev calling the push.

However, one thing not considered in this calcuation is that 6 players folded prior to button's push. Buttons cards are not chosen thus from a 50-card deck, as we've gained % information about 12 cards that were folded. It will then increasing the likelyhood that button has higher-ranked cards.

Could this consideration make any significant change to EV decisions? What situations would this effect be more powerful? Has this been discussed before? Am I just talking a load of crap?

Yeah it's been talked about before. I don't think it matters more than very marginally. Nobody, that I know, has got any sort of figures as for how much this offsets people's ranges, afaik.

Thanks, sorry for bringing up an old topic. I guess even in this situation - where there's a lot of dead cards, and yet player's being forced to open up fairly wide - the higher percentage of aces and kings in buttons hand won't differ greatly. The difference is marginal compared to your ability to put button a decent range.

Also just realised that this is not related to ICM, but EV calculations in general (a la pokerstove).

Meh I think this factor is pretty minor. Especially in this case where the table is described as tight so people are throwing away bad aces and kings and such. Not to mention you have an A in your hand.

In order for this to be matter I think you would have to know most of the players that folded will play any A, most Ks etc.