Aggregate Defensive Evaluations – 2009 LF
Update: Tangotiger pointed out to me that fans scouting report numbers for Holliday and Wilson looked off. Turns out I was calculating players who switched teams for the Fans Scouting Report incorrectly and as a result Matt Holliday and Jack Wilson are now adjusted accordingly and the tables below have been updated.
Earlier this week I ran the the Aggregate Defensive Evaluations (ADE) on 2009 shortstops. For those who missed it, this is an attempt to take 5 different fielding metrics (UZR, Fans Scouting Report, John Dewan’s DRS, Total Zone, and Total Zone with Location), put them on the same scale and then see which player’s defensive abilities we are fairly certain about and those which we are not.
In response to some comments, I’ve added a weighted average and standard deviation. This excludes standard Total Zone in favor of TZL. It also weights UZR, DRS, and TZL 3 times each and then the Fans Scouting Report only 1 time. (These are the last two columns)
It’s probably not much of a surprise that Carl Crawford sits atop the list. UZR has actually rated him as 56 runs above average the past three years, more than double the next closest player. Same goes for pretty much all the defensive metrics.
Matt Holliday I would say has the highest level of disagreement of any player. The Fans Scouting Report hates him, DRS loves him, UZR thinks he’s above average, and Total Zone thinks he’s just average. I’d consider the inconsistency with Holliday different than Juan Rivera’s situation where there’s also a high level of disagreement. In Rivera’s case at least all metrics agree he’s average or better.
Ryan Braun is also pretty interesting in that the Fans rated him as +15, while all the other defensive models thought he was well below average. This season Ryan Braun continues to be rated poorly by the defensive models. It will be interesting to see what the fans think of him next season.
I’d still consider these reports a bit of a work in progress, but for those interested, here’s the shortstops again, but this time with the weighted averages column:
David Appelman is the creator of FanGraphs.
The averages are really cool, but I’ve got to say I’m even more excited to have the FSR runs numbers on the site. Great addition (assuming it’s an addition).
Well… it’s not really an addition. I’m also not really sure it’s safe to say these are concrete runs either. It’s more of a scale from -20 to 20 ish to match up with UZR and the others. I guess it’s probably good enough.
You’re not going to list FSR runs on the site?
Have you seen this method for converting FSR rankings into runs? Check out the Fans Scouting Report section here: http://www.basement-dwellers.com/2007/10/player-value-part-3a-fielding.html
Basically, weight each trait depending on the position, compare to average, and convert from points to runs via standard deviations.
Also, any chance of re-zeroing the DSR numbers?
awesome.
are you going to incorporate these into WAR instead?
Thanks David, I used to do something like this on my site but this will get a much wider audience and it’s something that needs to be seen. I get frustrated when people go to the extreme of using UZR (or any other defensive measure in isolation) as a definitive measure of a player’s fielding contribution. At the same time, I dislike seeing people being dismissive of fielding stats. Analysis of fielding stats is an art as much as a science, but it still yields valuable information.
Can’t wait for this to be added to the Leaderboards!
Must…sort…by…wAVG…
i dont understand how Scutaro is a better defender than Alex Gonzalez
i’m a blue jay fan and i have seen both play and to tell you the truth Alex is a far superior defender with a slick glove and above average range.
AA did an amazing job at acquiring Yunel for Alex, the numbers ( and from what i have seen so far) back that up, amazing range+great accurate arm = successful shortstop defensely.
How can Juan Pierre be rated so highly? Little Leaguers have stronger arms, he breaks on balls like a little leaguer and 1/2 the time he takes the wrong angle. His speed makes up for a lot of mistakes, but everytime he takes a bad angle or the ball reaches the wall a runner from first plates. Anyone with decent speed can advance on all but the most routine flyballs. I’d take Braun or Young in LF for defensive purposes over Pierre any day of the week without factoring in their bats.
It’s great knowing the White Sox corner outfield is manned by objectively the worst outfielder (although those numbers are for LF, can’t imagine it’s [much] different in RF) and the worst outfielder subjectively (like it’s been said, Pierre may have good numbers, but he’s not a good LF guy).
Actually, defensively, Pierre is well suited to left field. Shortest throws and he has always been rated highly because of his range. He’s not suited offensively, because it is arguably the least important position on the field and one that generally needs to produce some power and at least a high OBP.
It could be that there are holes in the numerical defensive ratings. For example, the impact of allowing runners to advance could be underweighted… but it could also be that the underweighting is being done by the memories of individuals (who simply cannot, no matter what they claim, accurately remember each play and weight them equally). It’s not how you look getting to the ball, it’s whether you get there or not.
We will know that the defensive metrics are good when: a) different numerical philosophies yield consistent results and b) those results are reasonably predictive (because fielding skill just cannot be as volatile as today’s metrics are) and NOT when they agree with our personal impressions.
This is awesome.
Holliday has a couple of high-profile bad plays to his name (i.e. the brutal dropped fly ball in last year’s NLDS). It wouldn’t surprise me if he’s an overall average-ish fielder the way the other metrics suggest, but a few memorable bad plays that stick in the fans’ collective memories drive his FSR number into the ground.
Never mind, I see there was an update to his FSR.
Not at all surprised to see Braun with a high fan score. People in Milwaukee are still completely blinded by the athletic build and the ballet-esque crow hop he incorporates into every throw. He plays almost every type of chance wrong, and teams score from 2nd on singles to him like its Carlos Lee or Juan Pierre out there.
It’s uncanny, really. Every time I watch the Brewers play, I see teams just running wild every time the ball goes out to LF. I’m pretty shocked that anyone who watches him enough to submit a FSR could see otherwise. He is a brutal, brutal defender.
I’m actually quite surprised that he was rated this well. Among Brewers fans who actually watch Braun play, there’s pretty much a consensus that he is athletic and toolsy but cannot get a break on a ball for the life of him. If anybody watched his play in the 9th inning of Wednesday’s game against the Cardinals, it was pretty brutal. He has to be one of the worst 10 outfielders in baseball.
Is Brett Gardner so bad that he disappeared off the face of left field? He has been the primary LF’er for the NYYankees – at least, so I thought (and a pretty decent defensive player IMO).
Hate to answer for Dave but…these numbers I are from 2009, *not* 2010. See the title. Which is why Gardner isn’t on the list (he may show up for CF), why a player you observe as poor this year may have played better last year, etc.
Everybody who’s anybody knows Paul Janish is an extremely gifted defensive shortstop. It’s good to see all the different metrics back him up on that.
Great idea for this work. I particularly like the weighted version you’ve added to this posting. FSR tells us more about fan bias than defensive performance. I bet those FSR ratings would correlate way better with survey results on which players are believed to do things like “hustle”, etc. than they do with the numerical defensive ratings systems.
I’m surprised Juan Rivera scores so high, although he does have the highest standard deviation also.
Rivera always seems to get to balls and has a strong arm. His main weakness is an ineptness at caroms. Remember that this guy could play a passable CF before he broke his leg.