Should the draft be changed?

[37thchamber]

But at the same time, is it fair to the "experienced" rookie to be on the same regression as a S1 when they don't have access to the same amount of TPE that a S1 had? (Later time, training camps, etc.)

Maybe implement something that says if you had access to x% of available TPE for a season, it counts as your rookie year? Really it would depend on how much of a disparity we're willing to consider "fair" since there will ALWAYS be one. It's just not possible to avoid, as far as I can tell.

I reckon it would be fairer to say any player who has access to ... let's say 60% or more of the available TPE in their "first" season should be considered a rookie of that season when it comes to regression. Considering a little more than half of all season one players (88/161 by my count) had over 100 TPE at the end of season 1, while a quarter never updated (39/161 still have 50 TPE) and about a quarter (43/161) had 160 TPE or more, I think that might be a better way to address that particular problem.

If you plot the TPE distribution of season one players (using bands of ten points, i.e. 51-60, 61-70 etc etc) according to my records last week on a graph, there are two "clusters": one where you have about 40 players or so in the 80-120 range, and another of about 50 players in the 150-190 range. The mean TPE (ignoring any player with 50 TPE as a player who never updated) is about 135, so 85 TPE earned.

However, assuming the second cluster is people who remained active (because some whose activity dropped off would fall into the first cluster) I would estimate the "true mean" of TPE earned to be about 100, while the highest had about 160. If we take this as an indicator, it looks like the average active player picks up about half of available TPE*. So a player who has access to say... 60-70% of all TPE for a season likely isn't really at a disadvantage later on down the line compared to the others in their draft class. Assuming they get the average yield of TPE, they're maybe 25-30 points behind (2 activity checks * 2pts, 4 predictions * 3pts average inc participation pts, and 3 PTs * 4pts = 28 pts)... Which isn't far off the potential gap between players who sign up on the last day of season one and those who sign up in the week of the draft.

So maybe the trade deadline is a good cutoff for deciding which season is a player's first for regression purposes, actually, and you don't have to worry about calculating % available after all.

I dunno.

* Disclaimer: I didn't go and check how much TPE was actually available in season one, but I'm making educated guesses here. I figure there was about 200, based on the average amount earned and typical success rates on prediction threads etc.