Introducing UZR TargetView - Graph of the Day
by Justin Bopp on Oct 6, 2009 7:00 AM EDT
in Graph of the Day
16 comments
Who's on first? What's on second? ...Is that a nuclear bomb in center field?
Some of you may have realized that I like colors.
Bright, ridiculous, garish, jewel-tone colors. I can't avoid them. Like Dayton Moore loves on base guys with a lot of range that can neither get on base nor actually field, I like to sprinkle little Skittles-inspired rainbows into my charts and graphs where ever and whenever possible. With that in mind, I ask you to forgive me for assaulting your eyes with this attempt to visualize Ultimate Zone Rating.
Full assault after the jump.
Development
The goal, as always, is to present known data in a form more easily understood when viewed and combined with other data. Some blessed individuals have the ability to look at a set of numbers and get an idea of the scope of what he or she is dealing with. I do not. Sad, I know.
One sabermetric stat least well presented, at least visually, is Ultimate Zone Rating. Most often we are given a small list of players and their positional UZR. For those with the ability to instantly know what a 1.2 UZR/150 really means in relation to a fielder's team, that's probably good enough. For visual creatures such as myself, though, I need a little help.
Questions I want to answer are, "how does one UZR compare with another?" and "just adding up a team's UZR can't possibly be the best way to see their defensive capabilities, can it?"
Processing
Pretty simple process actually. What I wanted to do is visually associate UZR with some portion of coverage of the field. Obviously, UZR is more than actual fielder range, but short of adding unicorns to represent arm/error/double play runs (which could actually be pretty sweet), we'll stick with "the bigger the circle, the better the zone rating for the fielder."
The Fun Part
I started with an individual look at first base in both leagues, and then expanded the idea to see how it might be applied for an entire team. Please note that in the latter, this merely measures the given starter's ability. The result is something that looks like Rainbow Brite meets the old DOS game, Scorched Earth. I present to you UZR TargetView:
Kevin Youkilis had himself a better-than-decent year at first. Butler "vastly improved."
Pujols is actually down from last year. Dunn's issues have already been noted in this column.
Pretty well-rounded defense for St. Louis, actually.
Opportunities for improvement:
This view may give the false impression of representing the physical range of a fielder, which may be unavoidable for this visual method. However, I think the visual is effective in showing that players with a lower rating are literally less capable of preventing runs than those with a higher rating. Other potential concerns are the color scale and the presentation, including whether or not the inside circles should be included for each player or not. On the team graph, note that every fielder with a rating above 7.0 (dark blue) would visually cover everything outside the baseline.
Other than those, and the issues some might associate with the stat itself, I think this might work.
0 recs |
16 comments
|
Email
Print
Comments
ROYGBIV
Pretty cool, though I share your concern that the circles don’t actually represent field coverage. For the team graph, why not just make a simple circle without all the inner circles?
Also, why do you follow this color pattern? Wouldn’t a ROYGBIV (most to least) pattern make more sense?
by studes on Oct 6, 2009 7:39 AM EDT reply actions 0 recs
Thanks, studes.
1. Team Graph’s inner circles (for each player): good question, one I debated myself both in the creation phase and in the article itself. I think the reason I chose to keep the inner circles, is that visually, the different shades of orange are harder to differentiate on different sides of the image. It also helps the eye recognize that they are also different sizes. In another excercise of this sort, I’ll produce one without the inner circles to show you what I mean.
2. They Roy G Biv thing is a stylistic choice. Sometimes red means hot, sometimes red means stop, sometimes red means best, sometimes red means negative. In this case I am expressing that blue-to-green-to-yellow-to-orange-to-red means the transition from “go fast” to “go slow.” It would probably be as effective to do it the other way around.
Thank you for the feedback. This is the kinda stuff that can turn a decent idea into a great idea.
"What we do in life, echoes in eternity!"
by Justin Bopp on Oct 6, 2009 1:49 PM EDT up reply actions 0 recs
I didn't have a problem with the circles...
As long as it’s explicitly stated that the circles only represent ability and not range, you should be fine.
I’ve been wondering for a while about visualizing UZR myself, but in a different way. I was wondering if someone could color code each fielder’s ability on a team in each particular zone so that certain patterns could be found, i.e. this team is extremely susceptible to hits up the middle, or pulled into the gap between 1st and 2nd, etc. I’d do it myself, but I don’t have any idea where to find the individual zone ratings, instead of the entire UZR.
by dcfish on Oct 6, 2009 1:26 PM EDT reply actions 0 recs
Awesome idea, dcfish.
1. On the point of expressing ability and not only range, do you have any ideas how each component of UZR might be expressed in a chart like this? My unicorn idea wasn’t that ridiculous; perhaps there could be some mark or shading that represents a player with a wide range that can’t catch, or a player with a strong arm but a short range? Which brings me to your next point.
2. This is a fantastic idea, but one entirely dependent on un-processed data. Specifically, un-processed data that we either don’t have access to or are not at liberty to share (which is why you can’t find it). I think there’s a real possibility with your idea here. Let me sit on it for a little bit and see what I come up with. Maybe there’s another way.
"What we do in life, echoes in eternity!"
by Justin Bopp on Oct 6, 2009 1:54 PM EDT up reply actions 0 recs
I figured as much...
1. I agree with your idea of using symbols though I also agree that how to scale just how good a thrower or flyball catcher they are is a problem. I don’t think colors would work, since with all the colors on the actual UZR circles, it would get very busy and confusing very quickly. Perhaps change the size, the bigger the better?
2. I knew that such info would be difficult to obtain, and the more that I think about it, the more I realize how much it would be possible for teams to abuse such data. This may not be such a problem once the postseason ends (one would think that individual teams most certainly would keep some track of where their defensive weak spots are, and would be working hard to correct these holes in the spring, so I doubt they would carry over to the same extent from one year to another), but the practical reasons may outweigh the academic ones in this case. Perhaps, though, it might be interesting to obtain aggregate zone ratings over the entire major leagues which would not compromise any one team, but would just give us a sense of where the general holes are, just to confirm or deny certain preconceptions?
by dcfish on Oct 6, 2009 2:47 PM EDT up reply actions 0 recs
I could be wrong, but it's not about the sanctity of team secrets,
but more about individuals rightly protecting the data that they collect.
"What we do in life, echoes in eternity!"
by Justin Bopp on Oct 6, 2009 2:54 PM EDT up reply actions 0 recs
Justin -- How about this. Pick a standard size for all players at 0 uzr and then make it smaller or - UZR and bigger for + UZR
You could go with or with out circles. Just an idea.
Jeff Zimmerman - Protecting the world from RBI's and Wins from my mom's guest house.
by Jeff Zimmerman (TucsonRoyal) on Oct 6, 2009 2:05 PM EDT reply actions 1 recs
isnt there..
isn’t there something like this but without circles? its like the whole field is colored blue the brighter spots are the UZR/150 ratings. sounds like a stupid description but i’ve only see it a few times, once on BP
by nmigliore on Oct 6, 2009 2:33 PM EDT reply actions 0 recs
Great job. As it stands now, the colors are actually unnecessary, just a function of the size of the circle, am I understanding that right? I mention that because if it is true, you are using both color and circle size to signify one piece of data. If you divorce the color and circle size, you could use the color to represent a different piece of data. For example,maybe the circle size represents the UZR, while the color of the circle is a shade of gray to show arm vs speed with white being an all speed, no arm guy, and black being an all arm, no speed guy, with gray being a balanced player. Also, making it a hexagon (or any other shape) would solve the “range” problem.
Just spitballing…
It’s a great tool! You really do get an idea at a glance of the strengths and weaknesses of a team’s defense.
by Don_W on Oct 6, 2009 2:35 PM EDT reply actions 0 recs
Holy crap. This is brilliant.
Not sure about shape changes, but divorcing color and size could be extremely effective. I like your idea about having colors represent the actual type of fielder. Let me think on that for a bit.
Btw, you guys make the best saber-community. Together we can find more and better ways to express data.
"What we do in life, echoes in eternity!"
by Justin Bopp on Oct 6, 2009 2:44 PM EDT up reply actions 0 recs
Taking Don's idea a bit differently...
I would think that perhaps you could use the color of the circle to differenciate between the type of value (Blue = Total UZR, red = Arm UZR, yellow = Range UZR, etc) and that you then have 4/5 rings for each fielder based upon each of these metrics? Any overlapping rings might create a problem.
Then you could see quite simply how the total UZR is accounted for — so say Player A has a +5 arm and a -3 range, you could see that his +2 UZR is made up of a big ARM circle and a small RANGE circle.
Just my contribution to Don’s thoughts.
by Trickman on Oct 7, 2009 2:52 PM EDT up reply actions 0 recs
I LOVE IT!! It may lead you to think the circle is the area covered defensively at first but you explained that- it certainly gives you an idea of how different fielders range compares. WELL DONE!
by adenzeno on Oct 6, 2009 9:21 PM EDT reply actions 0 recs
I like it
Not sure that it’s a huge improvement over Jeff Zimmerman’s UZR graphs, but like his, I like how you can just glance at a chart and immediately get an idea of where th better fielders are.
I will say that one thing I like about Jeff’s graphs is that it’s easier to identify who is average—yours are very intuitive for identifying the awesome fielders and horrible fielders, but I have a harder time judging players that are somewhere in between. Also, if there aren’t any outstanding fielders on a team, you might be prone to overrate the average fielders using your figure.
That all said, though, I like ‘em. They’re pretty.
Idea: I’d love to see this kind of thing employed using fan scouting report data. One year of UZR has a lot of noise, but I tend to think that the scouting data are more reliable, even if they are more prone to bias.
-j
by JinAZ on Oct 7, 2009 8:29 AM EDT reply actions 0 recs
Holliday
Not sure why Holliday appears with a big negative number in your Cardinals graphic. All the underlying data I see shows Holliday with substantial positive UZRs year after year, including 2009. Mayber if you’re using just the small sample of his Cardinal games, you might get a very small negative number, nothing close to -5.3 that I can find. Something looks wrong there.
by birtelcom on Oct 27, 2009 11:04 PM EDT reply actions 0 recs
Range vs Making outs
I’m not particularly savvy when it comes to statistics, so please educate me if I’m completely off base here. But what about using something like Range Runs (RngR — fangraphs) as the basis for the size of the circle, then use UZR to determine its color.
This would make the visualization more intuitive, ie coverage on the chart in some way equals actual ground covered while also offering information on the quality of the defense the player plays within that zone (accounting for arm, errors, dp, etc. — as UZR does).
-- Aidan Sonoda
R.I.P. Nick Adenhart - 4/9/09
In necessariis unitas, in dubiis libertas, in omnibus caritas.
by Aidan Sonoda on Oct 31, 2009 4:14 PM EDT reply actions 0 recs
Comments For This Post Are Closed
Secondary Sidebar
User Tools
FanPosts
Community blog posts and discussion.
BtB on Facebook
Certain photos copyright © 2009 by Associated Press or Getty Images. Any commercial use or distribution without the express written consent of Associated Press and Getty Images is strictly prohibited.
