[LearnedfromTV]

You are misunderstanding me. I agree that expectation in a satellite should be measured and compared in $ terms to expectation in other games. Because clearly, if you can get the cash more efficiently in another game, you should. And whether entering a satellite to an MTT is a good idea is in large part dependent on whether you have the bankroll for the larger event, since the value of the tickets should be thought of as a portion of your cash bankroll and you shouldn't devote a large portion of your bankroll to a single event. I could go on but I think you did a fine job with this. Let's just say I agree.

What I'm arguing is something much simpler:

Take two tournaments, same buyin, same field size (for the sake of controls on the argument assume field size is constant), that pay ten percent of the field.

One is a standard $250 MTT. The other is a $250 supersatellite where 10% wins seats to a $2500. The second clearly has lower variance, assuming a player is equally skilled at both formats. If a player wants to play 1000 $250 tourneys and use the proceeds to play as many $2500 tourneys as he can afford, playing satellites will be a lower variance way of winning the same (expected) number of seats, and therefore a lower variance way of experiencing his overall expected return. To be clear I am talking about MTT-style supersatellites that pay 10% of the field, not something like a double shootout which pays 1 in 81 or a MTT supersat that pays 1 in 50 like a $300 sat to a WS package.