[Blarg]

I think the figure is almost unknowable for ROI.

The finishes in or next to the money are huge spikes except for 3rd to second. 4th to 3rd and 2nd to 1st are huge differences monetarily, and any number of random developments on the flop, turn, or river can have a huge difference in pushing you from one of those positions to the other.

Two out of three changes between finishing positions being both so vulnerable to a random quirk of the deal and also potentially netting you such huge changes in your prize money means you can only come up with a representative indicator of your skill level over an unbelievable number of games.

How many hands do you really get to play per tournament with three players? Or heads up? Well, if you get eliminated before you get ITM or heads up, as even the best players do in most tournaments -- none. Well, that's certainly an auspicious start. Then, if you actually get heads up, how many face-offs are going to be notably absent of the influence of luck? Few. And how many hands are you really going to play in those before somebody gets busted?

So you take a small part of a small number before you even have anything to measure. You'll have to accumulate a ton of SNG's before that small slice of them amounts to much of anything numerically, in the middle of all the noise of good and bad luck. And an ROI figure refined from an incredible number of games will still have an absurdly outsized ability to be swayed by completely ordinary patches of unusual luck for an incredibly long time.

And that's if your skill level is remaining somewhat consistent throughout the measurement period, which is also one heck of a gigantic IF.

I was just noting in another thread yesterday that out of the 600 or so SNG's I've played, I don't think there's a single point along the line in them that is representative of my skill level, from the very best to the very worst of my results. 600 and it's basically worth zero data points as far as my ROI goes. How long will it take till I even get something worth averaging into something else?

I'll need one hell of a daunting number before I can even begin to speak of what my fair long-term ROI is. Who knows what the number could be, but I could see tremendous swings happening even within a cluster of 1000 SNGs, so 50 or 100k of them starts to seem like a lot smaller number than it would sound.

And that AT ONE LEVEL, if your skill level is remaining in a narrow range -- uh oh.