UZR on FanGraphs
I’m pleased to announce that UZR (ultimate zone ratings) stats are now available on FanGraphs!
This would of course not be possible without MGL (Mitchel Lichtman), and I’m extremely grateful that he allowed us to use his fielding model. He’s put in a tremendous amount of work getting the stats ready for use on FanGraphs and I really can’t thank him enough.
Unlike other versions of UZR, the ones that appear on FanGraphs use Baseball Info Solutions location data instead of STATS location data. Also, this UZR data does not currently include outfield arms and double plays are treated as regular outs. Here are the definitions for all the UZR stats available:
DG (defensive games): The number of outs made by an average fielder at his position given the exact distribution of balls in play for that player divided by the number of outs an average player at that position makes per game.
exO (expected outs): The number of outs plus reached base errors that would be made by an average fielder given the distribution of balls in play while that fielder was on the field.
RngR (range runs): The number of runs above or below average a fielder is, determined by how the fielder is able to get to balls hit in his vicinity.
ErrR (error runs): The number of runs above or below average a fielder is, determined by the number of errors he makes as compared to an average fielder at that position given the same distribution of balls in play.
UZR (ultimate zone rating): The number of runs above or below average a fielder is in both range runs and error runs combined.
UZR/150 (ultimate zone rate per 150 games): The number of runs above or below average a fielder is, per 150 defensive games.
The UZR stats can be found in the fielding section on the player pages or in the fielding section on the leaderboards. Later this week I’ll have them integrated into the team sections. Additionally, the plan is to update these stats at least weekly during the 2009 season.
Update: The team pages now have fielding data too!
David Appelman is the creator of FanGraphs.
Yes, yes, a thousand times yes! Thank you so much for having this awesome stat widely available on the site.
I came.
This site is the gift that keeps on giving.
Holy cow, this is very cool. FanGraphs is becoming more and more well-rounded. Thanks for the hard work.
Awesome, David. Thanks so much. And much thanks to MGL, too. Here’s hoping for a cross-comparison of STATS and BIS UZR at some point. Hint hint.
David, any chance of getting a stat report where each player has a different line for every position/team combination? For example, Mark Teixeira currently shows up team-less on the leader boards (unless you go NL or AL only) and so does Dunn for one of his positions. Thanks.
Outstanding. Thanks for providing this.
Christmas has come early.
…and so has kensai, apparently.
This site just keeps getting better and better. What will they think of next?
Everyone: Really happy everyone likes it so far!
Sky: I might be able to add a de-aggregate button for the leaderboard, where you’ll see how each player did on each team. Honestly, it’s not a high priority unless I find out there’s a big demand for it. In the meantime, you can always go into the individual player pages and do a “show partial” seasons to get the breakdown for Dunn or Teix. And the team pages which will be released later this week will also show the breakdown.
David, thanks. I’m not looking for that fix as much to find out one player’s split ratings, but to be able to have all the data in Excel so I can slice and dice team data as needed. Just an FYI. You prioritize as need be. (The same request goes for all leader boards, separating out reliever, starter, and hitter data between teams.)
I love you guys.
Congrats! Very nice addition. It’ll be particularly interesting to now be able to more fully compare how different defensive models compare as UZR can be contrasted with PMR more readily. Given the common source of data from both it’ll allow insight into the comparative methodologies.
you guys rock, as does MGL and everyone else involved
Amazingly awesome. This just made my day even better.
Thanks so much, this is incredible. FanGraphs has quickly turned into a site i visit several times a day now. Just a quick question, since the UZR numbers are above average, how many runs above replacement would say, a 0 UZR rating be? about 10?
Rahul, replacement-level players tend to be average fielders. So, in your model, average IS replacement-level. Other stats that use a significant replacement-level adjustment for both offense and defense (like BPro’s WARP) are putting the overall replacement level too low, about what a decent AAA player would provide.
Awesome work all over the site gentlemen, this addition is exactly what I’ve been looking for. Thanks for all the quality stuff!
I love you.
Teej,
We love you too. And everyone here commenting, this site wouldn’t be what it is without all of the thoughtful feedback provided. I know that Dave and I, Marc, Matt, and Rotographs guys (as well as our captain David) work very hard to produce quality content and to know that this site has gone from maybe site #9 on your priority list in April to #1 by a vast margin really, as cliche as it sounds, motivates me to get up in the morning and write.
Wow, this is awesome. Not only do you guys make these great metrics available to all of us, but the blogging content is also fantastic. Your hard work is very much appreciated.
This is great!
Not to be greedy, but would it be possible to get PZR for pitchers? This would cut through a lot of the DIPS-type debates.
Or — and this is a pipe dream — UZR for *hitters*. Say Pujols hits a ground ball that UZR says should be converted into an out 70% of the time. Whether it actually is a hit or not, give him credit for 0.3 hits. Then see how different his “UZR batting average” is from the .357 he actually hit.
I can already see problems with this idea. Certainly a fast runner will consistently beat his UZR batting average. Also, hitters who see an infield shift much of the time will fall short of their UZR BA. (Else, why would the infielders shift?)
It may also be that UZR can’t tell a well-hit ball from a really well-hit ball, so batters who hit the ball harder will beat their UZR BA. This would expose a weakness of UZR, or of the BIS dataset. If we have UZR figures for hitters, we can go beyond speculation and actually find out how big an issue this is.
Once again, thanks to David and MGL for all your hard work!
Let me join the chorus of praise. You guys have just done a terrific job over the past year turning this site into one of the absolute best sabermetric resources on the internet. A job very well done.
Eric, what you said is exactly true. It has rocketed up the list of everyday baseball site that I read.
Seriously, all the credit needs to go to David Appelman – his dedication to turning this place into a premier baseball site has obviously paid off. I loved FanGraphs before I started writing here, and even if he fired me tomorrow, I’d still make this one of my first stops every morning.
What Dave just said, times 1 million. David Appelman is such a tireless worker and is open to adding anything and everything that makes this website great. Programming this site and keeping the database processing times smooth is not an easy task, and his work inspires us all to keep on doing what we can to make this site fantastic.
It’s like Christmas in December!
When comparing players from various teams, should we be considering park factors?? Or are the Rockies really THAT much worse as a whole, and were the Rays really THAT much better??? 140(!) points between the two.
Broadcast James,
I was under the impression that UZR is indeed park-adjusted, but I could be completely wrong. As for the Rockies’ defensive decline, you can thank Brad Hawpe for being even more of a crappy fielder than he already was. A -37.2 UZR is just brutal.
The Rays were REALLY good at defense, and the Rockies were REALLY bad at defense, your eyes do not deceive you. Brad Hawpe was definitely the worst defender at his position in baseball, the +/- system agrees with this as well so I don’t think it’s an aberration with UZR. The Rays D was one of the things that helped the major turnaround from last year.
Great addition to what’s really become a first-rate site.
I agree about David. He is superfast to implement things and change them when needed.
Yes, this is great news and great data to have.
Will we ever see any split data in the stats section?
Much thanks to everyone involved… You’ve succeeded in making a great site even greater! As a Mariners fan, the first site I check is the USS Mariner…but the second is always FanGraphs. Keep up the excellent work, gentlemen.
So awesome thank you David, MGL, and the rest of the crew.
Quick question. Does UZR use a hypothetical standard for the average fielder at each position or is it an average of current players, somewhat like Dewan’s +/-?
And another quick question. In looking at the team defensive stats does UZR/150 represent the average defensive value of a single player on that team or that team’s aggregate defensive ability over that season?
Does anyone really believe Tulowitizki is th 18th best fielding SS in the game as UZR states he is? If so, I suggest whoever DOES believe this watch some Rockies games…He and Wilson are the best Fielding SS in the game.
Does anyone believe David Wright is the 8th best hitting third baseman in the game, as he ranks in wRAA? Or that Grady Sizemore is the 15th best hitting center fielder?
UZR does not say that Tulo is the 18th best fielding shortstop in the game. That is simply where it places the value of his fielding this season. Before you can know what it says about who is better than whom, you would need a projection for UZR, which would include multiple years of data and some regression.
Just an observation: So I am led to believe that Chase Utley, a guy with a rep as a poor 2nd baseman with no range, is actually the BEST 2nd baseman with great range? This is just one of many questionable things that jump out at me from these ratings. I’d say theres a flaw in the logic.
Who said Utley was a poor second baseman with no range?
Trying to figure out the UZR.
Lets say a club has a fly ball pitching staff leaving little for the infield to do. Using UZR, are all the infielders going to have a bad UZR because they do not see or get to the average number of balls?
No, quantity of opportunities are included in the calculation.
Thank you, good answer. Jacoby Ellsbury I am told is a bad fielder because he has terrible range and bad routes. The bad route, slow pick up I can see, but the range? The man, side to side can cover all but the left and right field lines. If he reaches it, it is caught.
Does a minus uzr actually mean he is bad? Is there some Bell curve to say that says if you reach this number you are bad.Adam Dunn -40, Ellsbury -18?