The range of error is 5 runs either direction. Meaning if a player has 5 runs saved in my system his real value could be anywhere between 0 and 10. The same is also true of the popular fielding stat UZR.
My system doesn't look at bins or how many fielders would have caught that type system. It uses a few statistics that you probably have seen before. These are putouts, assists, and errors. I know these stats are what made fielding as messed up as what it had been for years. But in UZR the system is spitting out numbers at you. In this you can calculate it yourself because I'm displaying the formula.
The formula for SS is the following
Putouts+Assists divided by errors, divided by 4, subtract 5 from result.
Here are the results for Shortstops and how they compare to other fielding stats like UZR and DRS.
Name
|
FLW
|
UZR
|
DRS
|
Elvis
Andrus
|
5
|
5
|
11
|
Pedro
Flormien
|
4
|
4
|
12
|
Stephen
Drew
|
12
|
5
|
-2
|
Starlin
Castro
|
2
|
-8
|
-3
|
JJ
Hardy
|
8
|
6
|
8
|
Brandon
Crawford
|
5
|
4
|
2
|
Ian
Desmond
|
4
|
4
|
-3
|
Zach
Cozart
|
6
|
6
|
4
|
Alexi
Ramirez
|
3
|
6
|
1
|
Pete
Kozma
|
10
|
7
|
8
|
Jean
Segura
|
6
|
-1
|
3
|
Adeiny
Hechavarira
|
5
|
-9
|
-3
|
Alcides
Escobar
|
7
|
11
|
4
|
Troy
Tulowitzki
|
13
|
7
|
6
|
Jimmy
Rollins
|
9
|
-2
|
-15
|
Asdrubal
Cabrera
|
9
|
-13
|
-16
|
Jed
Lowrie
|
1
|
-7
|
-18
|
Johnny
Peralta
|
22
|
4
|
0
|
So you might notice that for the most part my system make sense and shows how well a fielder fields his position 85% of the time. My line of thinking is that if the system is right most of the time then it does the job. If I tried to fix some of the hits and misses on this, then I believe I could mess up some of the other numbers that were right in the first place. I'm happy with where it is at now.
I'll start at the bottom because that's the most visible miss on this list. Johnny Peralta is not the best fielder on this list. All though him being 10 runs ahead of the second place finisher is impressive really it just means he had very few errors which he did. Errors are not a very good way to measure defensive mistakes. I know this, and hopefully you know this. However, errors tell you how many mistakes they made based on a score keeper's opinion. In my mind an error is called as it should about 80% of the time. And as we know I, nor can you can measure fielding exactly right. We can try to do our best and that's what I attempted with this system.
Despite the popular opinion that Jimmy Rollins is declining on offense and defense from UZR and DRS my system says he is still a pretty good fielder. Let's dig deeper on this.
2011: 3.5 UZR 2012: 7.9 UZR 2013:-2.7
It's been said you need the player's last three years to make a good defensive projection or guess on his true defensive ability lately. His average is 2.9 which rounds up to 3 UZR over the last three years. Defensive unlike offensive doesn't seem to correlate well from year to year. His UZR score in 2012 correlates with his Fielding Linear Weights score in 2013. I believe Jimmy Rollins is a good fielding SS, so you might think it's a miss or a mistake but I think it's a good rating.
Asdrubal Cabrera is a miss, I can't defend this one.
My opinion on players "losing" range with aging is simple. It's not as drastic as some of the advanced defensive metrics rate it as. It seems like all old middle infielders lose range. This is not true. If you have lots of assists, putouts, and few errors the system and I assume you're a decent fielder. My system is forced to make assumptions and like I said 85% of the time there right. I'm not trying to say Johnny Peralta is the best fielding shortstop in the majors, I'm rather saying that the result for him makes a statement about his fielding. Range is mostly measured in assists in the system. I know you should account for other factors like groundball/flyball staffs and other things. Maybe soon I'll adjust for that. But for some reasons advanced defensive metrics in my view of things are forced to make the assumption that players lose range as they age.
Thanks for reading. The 2013 Second Base edition should be here soon. Follow me on Twitter
No comments:
Post a Comment