Sabermetrics: A Science
In this series I will describe the relatively new phenomena in baseball called sabermetrics. In the 2000's, sabermetrics have really started to gain popularity. Front office decision makers and hardcore, modern fans have really started to pay attention to these stats.
Why aren't sabermetrics well known by casual fans? Well, old habits die hard. Everyone loves batting average and home runs and ERA. However, these simple stats can be influenced by teammates, the ballpark, and worst of all, luck. I am taking on the responsibility of teaching you about the improved statistical analysis called sabermetrics
Installment IV: UZR
So far in this series I have written about a general statistic (win shares) that can be used for both hitters and pitchers, I have written about a metric used specifically for hitters (OPS and OPS+) and I have written about one that is for pitchers only (FIP and xFIP). One thing is missing: fielding. UZR, or Ultimate Zone Rating, is a type of metric used to measure a player's ability defensively.
The genius who is attributed the development of this extraordinary is known by his intials, MGL, or Michael Lichtman. He based it on ZR, or Zone Rating. UZR was first published in the late ???90s and it was the first of the new sabermetric statistics that evaluate the player's ability to field. Unlike many other metrics, UZR has quickly become quite popular among baseball fans and it is widely used as the #1 indicator of a player's defensive ability.
However, the accessibility of this stat has declined. MGL is currently working for the Saint Louis Cardinals. Because of this, the Cardinals organization has the easiest and most accurate access to the metric. However, MGL has recently decided to give out UZR and FanGraphs is one of the few sites that have it available. But keep in mind that the sites that have UZR typically have a simplified version that is slightly less accurate.
The most surprising thing about the popularity of UZR is the difficulty of calculation. The calculation of this metric is anything but easy. I will separate into steps in an attempt to make it easier to understand.
1.) The field is separated into 78 different sections (only 64 of them are used in the actual calculation.
2.) The out-rate and the percentage of balls hit into each respective zone is calculated for each and every one of those 64 zones. Very tedious work.
3.) The league average out-rate is subtracted from the player's out-rate (therefore, a negative number means below average and a positive number means above average).
4.) The difference from is multiplied by the number of balls hit into the zone (this is the Zone Rating).
5.) This is multiplied by the Zone Ratings of each zone that the fielder covers.
6.) They are then added (resulting in an unadjusted Ultimate Zone Rating).
This is a continuation from the previous process of calculating Ultimate Zone Rating. The process above is simplified and unadjusted. Here are the adjustments made to make UZR more accurate:
I.) Park Factors - This is pretty self-explanatory. Fields have different measurements. Fenway Park is a very good example of this. Centerfield at Fenway is extremely difficult to defend because there is a lot of space to cover and the walls are not perfectly straight. Leftfield at Fenway is much smaller than most in the MLB. This greatly affects the results of a player's unadjusted UZR.
II.) Batted Ball Speed - Again, self-explanatory. The speed of a hit greatly influences a fielder's ability to catch it. You are more likely to catch up to a high pop-up than a screaming line drive, no?
III.) Batter Handedness - The influence of which side a batter hits from is not as great as the previous too, but it does make a difference. It is obviously easier for third basemen to defend a ball hit to him from a left handed hitter rather than a right handed hitter. The opposite can be said for first basemen. This is because the ball travels much faster when it is pulled and the visibility is not affected by someone hitting from the opposite side of the plate.
IV.) Groundball-to-Flyball Ratio - The pitcher's tendency to give up a certain number of groundballs to flyballs also changes a fielder's approach. This changes how a fielder lines up before the ball is thrown and, therefore, how much room he has to cover.
V.) Number of Runners On Base - A fielder's approach to a hit changes depending on how many runners are on base. The fielder will go after a ball differently when there is a man on third than when no one is on. This is similar to adjustment factor VI.
VI.) Number of Outs - Like the previous factor, a fielder will approach the ball differently when there are two outs than when there are none. Pretty straightforward.
Overall, it is very hard to accurately calculate adjusted UZR. Even unadjusted UZR is complicated. First of all, you have to know what the 64 sections of the baseball field are. You also have to know how to do all the other stuff. I would be completely lost.
What it tells us:
Adjusted UZR is the most accurate, most popular metric used when determining a fielder's ability. To put it simply, UZR tells us how good or bad a specific fielder is. In a slightly less simple way of putting it, UZR tells us how effective a fielder is at turning hits that go into one of their respective zones into outs. So, basically, if Hitter A hits the ball into Zone B, what are the chances that Fielder C will be able to get Hitter A out. Understand? Good.
It is generally accepted as the best way to judge a fielder's ability. Of course, it does have its faults. However, it is not as misleading as stats like fielding percentage, which doesn't account for how much ground a player can cover. There is also a variation called UZR/150 which just averages the UZR out over a 150 game total.
Now, we all have heard about how the Red Sox have improved their defense, right? Here is how much their defense has changed (using 2009 UZR/150 statistics):
Important Note: UZR is not calculated for catchers and pitchers!
2009 Defense (using the player who played the most games at the listed position):
1B: Kevin Youkilis (15.2 UZR/150)
2B: Dustin Pedroia (10.6 UZR/150)
3B: Mike Lowell (-14.4 UZR/150)
SS: Nick Green (8.3 UZR/150)
RF: J.D. Drew (12.9 UZR/150)
CF: Jacoby Ellsbury (-18.3 UZR/150)
LF: Jason Bay (-11.2 UZR/150
1B: Kevin Youkilis
2B: Dustin Pedroia
3B: Adrian Beltre (21.0 UZR/150)
SS: Marco Scutaro (1.0 UZR/150)
RF: J.D. Drew
CF: Mike Cameron (10.3 UZR/150)
LF: Jacoby Ellsbury (21.8 UZR/150)*
*career numbers used at LF due to limited games spend there. His UZR/150 last season at LF was 30.3.
So, the Red Sox' defense should improve by an impressive 89.7 UZR/150. Keep in mind that this is skewed due to Ellsbury's lack of games at LF in his career. The Ranger's defense was much improved last year and it greatly helped their pitching. So here is a comparison between the Ranger's defense from 2008 to '09.
1B: Chris Davis (2.2 UZR/150)
2B: Ian Kinsler (-8.1 UZR/150)
3B: Ramon Vazquez (-18.3 UZR/150)
SS: Michael Young (-5.4 UZR/150)
RF: David Murphy (12.7 UZR/150)
CF: Josh Hamilton (-17.7 UZR/150)
LF: Brandon Boggs (8.8 UZR/150)
1B: Chris Davis (-5.1 UZR/150)
2B: Ian Kinsler (9.6 UZR/150)
3B: Michael Young (-10.7 UZR/150)
SS: Elvis Andrus (11.7 UZR/150)
RF: Nelson Cruz (13.4 UZR/150)
CF: Marlon Byrd (-9.5 UZR/150)
LF: David Murphy (0.6 UZR/150)
The Rangers' defense improved by a total of 35.8 UZR/150, which is noticeably less than the Red Sox. However, it is fair to say that the Red Sox won't have that much of an improvement since players will be adjusting to playing with each other.
It is safe to say that the Red Sox will have a much improved defense this season, and therefore much improved pitching.
Here are the top five in the MLB last season in UZR:
1.) Franklin Gutierrez, CF (29.1)
2.) Evan Longoria, 3B (18.5)
3.) Ryan Zimmerman, 3B (18.1)
4.) Carl Crawford, LF (17.6)
5.) Chone Figgins, 3B (16.7)
And the bottom five:
5.) Vernon Wells, CF (-18.2)
4.) Jacoby Ellsbury, LF (-18.6)
3.) Jermaine Dye, RF (-20.0)
2.) Yuniesky Betancourt, SS (-20.5)
1.) Brad Hawpe, RF (-21.3)
Here are the top five in the MLB last season in UZR/150:
1.) Franklin Gutierrez, CF (27.1)
2.) Adrian Beltre, 3B (21.0)
3.) Jack Wilson, SS (20.4)
4.) Ryan Zimmerman, 3B (20.1)
5.) Evan Longoria, 3B (19.2)
And the bottom five:
5.) Dexter Fowler, CF (-20.3)
4.) Michael Cuddyer, RF (-22.1)
3.) Yuniesky Betancourt, SS (-23.9)
2.) Jermaine Dye, RF (-24.5)
1.) Brad Hawpe, RF (-25.9)
- According to UZR and UZR/150, Brad Hawpe is the worst fielder in baseball.
- Franklin Gutierrez, as expected, is the best by far.
- A surprise appearance by Jacoby Ellsbury on the bottom five. Looks like diving catches and speed doesn't always mean a good fielder.
- There are very large fluctuations year-to-year in terms of UZR and UZR/150. For example, Mike Lowell had an 11.1 UZR in 2008 and then a -10.4 UZR in '09.
- It does not tell the number of errors a player had in one season.
- They are not recorded for catchers and pitchers.
- There is limited availability and most of the available ones do not have the formula down perfectly because it is so difficult to calculate.
UZR and UZR/150 are very useful metrics. I personally think using them is very effective. However, the fluctuations from year-to-year change the way that they should be used. When evaluating players, I would advise averaging the total from over a few years rather than just picking out one year.
There are a lot of ways to improve this, which I'm sure someone will manage to do in the upcoming years. There are a lot of extra factors that influence this metric, but they can be worked out. However, as of now, this is the best fielding statistic there is.
- David Gassko of The Hardball Times (which is a good source for information on sabermetrics) wrote a good article on a lot of defensive metrics, including UZR, here.
- Yet another article on The Hardball Times. This one questions the reliability of UZR.
- This is another article in Alex Remington's "Everything you wanted to know about" series.