[A_PLUS]

I agree with a lot of this. I have been talking about this main idea a lot recetnly.

With a skill advantage, you would like to have as many chips as the rest of the table. I think that statement is pretty much agreed upon. In a cash game, when you have a skill advantage, being short hurts your win rate. So, in cash games there is some advantage to being as deep as the table.
Lets say you buy into a 2/5 NL game, where everyone has 2000$ behind, but the max buy in is 500$. Wouldn't you take a 49.9% gamble to double your stack on the first hand

In tournaments the statement from above still holds. (in terms of CEV). But often in tournaments, you reach a point, where you start pushing 1+ times an orbit trying to maintain a playable stack. So, although a correct preflop push fold strategy is CEV (and not too bad in terms of CEV BB/100), other factors come into play. There is a significant opportunity cost of folding.

The CEV of a preflop push, can be broken down into:
-hand value
-position
-stack size
-FE (related to position and stack size)

Lets say there is no antes, to make things easy.
-I am on the button with 11xBB, and an average image. Pushing 72o is pretty much CEV neutral. (.08BBs to be exact with a calling range of A8s+,ATo+,66+,KQs).
-Same table, I folded the previous hand, and pick up 72o in the CO. An open push now, is worth -.29BBs)
-Pushing AJo UG, is worse than pushing 72o from the button, when you have 11xBB.

This is nothing new, but just an illustrative way to show opportunity cost in action. Every hand we fold makes our position worse, which hurts. Every blind we take hurts our stack and our F.E. When we are far from the money, and very short, this effect is very dramatic.

The whole point of that ramble, was to say that due to the opportunity cost of folding, we are forced to gamble more when we get short, which hurts our CEV win rate more than we would see in cash games.

So, given that short stacks dont play as well in tournaments. We are correct to take slightly (I dont know how much) the worst of it, to avoid reaching that point (also assuming that the table has some larger stacks where we have an edge).

For example. Early in an MTT, I have 20xBB and think about calling a push for my stack, that I know to be -.25BBs in CEV. (everyone else has 50xBB, blinds double at the next level)

-If the current level is going to last another 100 hands, I will fold.
-If the current level is going to last another 5 hands, I will call.

I think that having 20xBB at a 25xBB deep table, is worth .25BBs in value more than having 10xBB at the same table. Since it is unlikely that I will double in the next 5 hands (more likely in 100 hands) I think that the structure does change your play.

I realize that this doesnt necessarily address the speed of the events, since we will reach the point in the 100 hand level where you have 5 hands left and the plays are identical. But, in fast structured events, you may never have more than 10 hands until an increase, so you are talking about the opportunity cost on basically every hand, whereas it may effect 1% of your hands in a slower event.

I see Mason's point, and agree with it. But he is just catching the author on a technicality more than anything. The author is explaining in a simple way how to account for an quickly upcoming change in blinds. Mason is saying that you just use the same strategy you do in a slow event the 1% of the time you are faced with the same situation.

I see the arguement like this.
Arnold: Batting in hardball is much different than batting in arch pitch softball.
Mason: No its not, they are identical. You are dumb
Arnold: No way, you do all of the following differently.....
Mason: You would do all of those things the same if the hardball was pitched 12 feet high, and 35 miles per hour, so you are wrong...and dumb

** Thinking a little more, I am starting to side with the author. The cost of folding is higher when levels are short, since you will be faced with the need to double in the next 10 hands after the first increase, whereas in slower events, having 100 hands to double allows more selective play.