Thursday, March 25, 2010

Drafting Analysis



The above picture shows the more useful set of draft information, as has 3 measures of the expected value of each teams roster (six if you count the sums separately from the averages, even though they tell you the same thing) (12 if you count the sums, averages, actual numbers and rankings all as separate, but that might be dumb).

The 2009 player rater shows what ESPN says a player did last year. Obviously players that didn't play last year got zero's, so all the prospects really bring down the average. Even worse, several players who came in, did poorly, then went to the DL all year ended up with negative numbers. The former blithe tots maxed this out, while moon shine has the lowest values (probably because of Liriano's way negative).

Second column shows the ESPN dollar value. Last year's champ leads the dollar values and again Moonshine is at the bottom (a running theme for this table). I doubted whether to even put positional rankings in, since the rankings for pitchers completely skew the other numbers. I figured the rest of the table was not particularly useful, so might as well add quantity.




Part 2 is mostly draft analysis. The big issue with all of this is the keepers. Didn't really matter which 4 keepers you had in what order, they were randomly assigned. It might even out a bit since it happened to everyone, but without putting too much thought into it, it seems like it flaws the rest. I also had to assign values to non-drafted players (since it was rosters as of this week, after people made changes already). I gave all non-drafted players a value of 201 (ESPN gives undrafted players 260 or 261 since their standard leagues pick that many players).

Anyway, here's the rest of the flawed rest. The first two columns show the number of players that were taken for each roster either before or after their average draft position (ADP). The 2nd two columns show how many players on each roster were taken more than 10 spots before or after their ADP. The Bulls again had the least reached for players (thanks auto-draft bot) and the most late round picks (again, thanks draft bot). Those unicorns had 10 players (of a roster of 20), that were reaches (thanks homer calls!). The limber lefties evidently couldn't wait for players, as they had only 3 later bloomers.

Almost all of this means nothing, but here it is for you anyway.

Post Draft Roto Rankings


I'm hoping to get something or things up about post-draft rosters and such before the season starts. It may well happen.
update: It did.
This is less than analysis but it took some doing.
[image][/image][link]http://lh4.ggpht.com/_yL5evi7KW4k/S6kigkCMlpI/AAAAAAAAEww/73VJBaRG4Ao/BP%20Preseason%20Roto.JPG[/link]

I took the preseason ESPN 2010 predictions and took the average per player on your roster. Note this is highly inaccurate because it assumes that the bench spots play all the time, which cannot happen. Also, all the rate categories are averages of averages, so it doesn't take into account expected AB or innings pitched.

It should give you a decent idea of what categories you have the players to expect to win versus which opponents.