This is something else I haven't seen anyone try to do, try to somewhat seriously measure defense in college baseball statistically. So here, I picked 100 2012 college players and tried to see if I can sort them by a somewhat reliable defensive metric. Assists and Putouts are usually kept as official statistics. However, because the official NCAA baseball site's (perhaps ironically, the Junior College website is better when it comes to this, and that is why I included some Junior College shortstops) statistics page is set up so strange (and sites like College Splits simply don't even try to measure defense), I had to go to each team's site to calculate the range factor. Although range factor is crude and has it's own issues, this gives us a better statistic than fielding percentage when looking at college position players' defense. While RF/9 is better than RF/G, we just don't have innings data for really any college players. Of course, things like playing multiple positions (common in the college game, except, it seems, for shortstops. It appears that most coaches prefer to keep the same guy at shortstop all season) and DHing occasionally (which I accounted for when I could) affects RF/G if not accounted for. If a player had any real significant time at 2 different positions, I didn't include them.
So this is a mixture of players still in college and players that are now in the minors. This is why I put the current team that the player is with in the 4th column. Since range factor is not really a measure of catcher defense, the only catcher I put in there was by accident (and I just decided to leave him and I put 101 players instead of 100 players). I sorted it by position so they are grouped correctly and then by range factor. Below is the link to download the spreadsheet (or use this hyperlink):
https://docs.google.com/file/d/0B_DUd_c_mKWraTg5Zl9zNzZOS2c/edit
To see if there was anything to the predictability (as that is what you are trying to measure when you are using statistics in baseball/most other fields) of the college RF/G, for the players that also played in the minors that year (after the draft obviously), I compared their college RF/G to their minor RF/G and FRAA. Because there wasn't a ton of data, instead of using a spreadsheet (plus I didn't want to make readers download two spreadsheets), I just put them below (sorted like the spreadsheet, by position, then by college RF/G or CRF/G):
Jorge Flores SS 4.75 CRF/G, 3.83 MiLB RF/G, -1.9 FRAA
Anthony Melchionda SS 4.74 CRF/G, 4.32 MiLB RF/G, 0.0 FRAA
Kenny Diekroeger: SS 4.61 CF/G, 4.30 MiLB RF/G, .5 FRAA
Jason Stolz: SS 4.16 CF/G, 4.74 MiLB RF/G, 5.4 FRAA
Nolan Fontana: SS 4.03 CF/G, 4.73 MiLB RF/G, -1.8 FRAA
Chris Taylor: SS 3.81 CF/G, 4.28 MiLB RF/G, 0.0 FRAA
Austin Nola SS 3.77 CF/G, 4.59 RF/G, -.6 FRAA
Michael Faulkner OF 2.94 CF/G, 1.8 MiLB RF/G, .7 FRAA
Mitch Haniger OF 2.82 CF/G, 1.83 MiLB RF/G, -.7 FRAA
Travis Jankowski OF 2.73 CF/G, 2.34 MiLB RF/G, 1.1 FRAA
Kyle Johnson OF 2.52 CF/G, 1.39 MiLB RF/G, .2 FRAA
James Ramsey OF 2.37 CF/G, 2.00 MiLB RF/G, .3 FRAA
Derek Jones OF 2.15 CF/G, 1.95 MiLB RF/G, 2.7 FRAA
Jake Stewart OF 2.00 CF/G, 2.03 MiLB RF/G, -1.8 FRAA
Tyler Naquin OF 1.9 CF/G, 2.53 MiLB RF/G, 3.6 FRAA
Richard Palase OF 1.77 CF/G, 1.77 RF/G, .4 FRAA
Barrett Barnes OF 1.75 CF/G, 1.76 MiLB RF/G, 0.0 FRAA
Tyler Booth OF 1.49 CF/G, 1.91 MiLB RF/G, .2 FRAA
Jabari Henry OF 1.29 CF/G, 2.02 MiLB RF/G, -1.1 FRAA
Fernando Perez 3B 3.93 CF/G, 2.82 MiLB RF/G, 0.0 FRAA
Patrick Kivlehan 3B 2.8 CF/G, 2.36 MiLB RF/G, .2 FRAA
Richie Shaffer 3B 2.63 CF/G, 2.13 MiLB RF/G, -.9 FRAA
Patrick Claussen 3B 2.44 CF/G, 2.55 MiLB RF/G, .2 FRAA
William Carmona 3B 2.41 CF/G, 2.00 MiLB RF/G, .5 FRAA
Stephen Bruno 3B 2.28 CF/G, 2.42 MiLB RF/G, 1.5 FRAA
Tyler Hanover 3B 2.16 CF/G, 2.79 MiLB RF/G, 1.5 FRAA
Stephen Piscotty 3B 1.95 CF/G, 2.69 RF/G, .3 FRAA
Tommy Richards 2B 5.2 CF/G, 3.93 RF/G, -.3 FRAA
Jamodrick Mcgrunder 2B 4.06 CF/G, 5.42 RF/G, .3 FRAA
Taylor Ard 1B 10.33 CF/G, 8.91 RF/G, -5.2 FRAA
Jake Davies 1B 6.05 CF/G, 7.67 RF/G, .9 FRAA
BP's site has a lot of problems on the minor league statistics side and the 0.0 FRAAs for Melchionda, Barnes, and Taylor are not because they were neutral defensively (I believe Perez' is right though), but because the site didn't log it correctly.
There doesn't seem to be much correlation there, whether simply because the statistic is easily corrupted, sampling issues, or some other variable, it wasn't very predictive with the players we looked at. Of course, to really determine the predictability, we would need to look at a lot more players and use a much bigger sample size. With that said, as of now, we don't really have any reason to believe that college range factor has any predictability. A more advanced metric and better data would be very helpful in this arena. With the current data publicly available, we should be able to calculate "defensive efficiency", but this is done a team wide scale, which is not very helpful when trying to determine individual player's value for the draft.
No comments:
Post a Comment