How Fans and UZR Disagree: the Rockies

The last post introduced the topic at hand and looked at the San Francisco Giants as they were the team with the biggest overall spread between their team FSR and UZR ratings. They happened to have much higher UZRs than FSRs in both years.

FSR vs UZR in '09 and '10

The two other highly outlying dots on the same side as the Giants are the 2009 Mariners and 2010 Diamondbacks. On the other side of the disagreement line are the Colorado Rockies. While the fans were of a dimmer opinion of the Giants’ fielders than UZR was by a large margin, the fans had a much higher feeling on the Rockies than UZR claims is warranted.

The Rockies had a +39 FSR in 2009 and a +22 in 2010 but their UZR was a mediocre -20 in 2009 and a slightly worse -28 in last season. Just as the Giants’ positive gap between UZR and FSR was the biggest in the league, the Rockies negative gap between the two in both 2009 and 2010.

Coors Field could be having an effect here, especially since the biggest differences between FSR and UZR at the individual level are almost entirely due to the outfielders. In 2009, Dexter Fowler and Brad Hawpe had a combined FSR score of -2 runs, but according to UZR were worth an astoundingly awful -36.2. This last season, it was Ryan Spilboroghs and Carlos Gonzalez’s turn. The fans rated Spilboroghs at -1 FSR and Carlos Gonzalez at +10 FSR while UZR pegged them at -16.7 and -2.7 runs respectively. The outfield is not the entirety of the discrepancy between the two systems, but it is clearly the majority factor.

Print This Post

Matthew Carruth is a software engineer who has been fascinated with baseball statistics since age five. When not dissecting baseball, he is watching hockey or playing soccer.

18 Responses to “How Fans and UZR Disagree: the Rockies”

You can follow any responses to this entry through the RSS 2.0 feed.
  1. Fantastic stuff Matt. I have picked on the discrepencies between UZR and the eyeball test specifically with the Giants’ and Rockies’ outfields for over a year. It’s mind boggling that Pat Burrell and Aubrey Huff can have better UZR’s than Carlos Gonzalez. There definitely seems to be a park adjustment in UZR that seems amiss. Willy Taveras was a plus fielder essentially every season of his career, except Colorado. Meanwhile, just 2 of 17 outfielders for San Francisco since 2009 have a negative UZR.

    Vote -1 Vote +1

    • hjrrockies says:

      It’s probably better to discuss in the Rockpile sometime, but it’s worth noting that several Rockies outfielders had better UZR’s at home than on the road.

      But I agree that something must be amiss, somewhere.

      Vote -1 Vote +1

      • I’m aware of that, and to be honest, I have no explanation. Though it’s possible that it is due to players inexperienced with the different flight patterns of flyballs going from sea level to Coors.

        Vote -1 Vote +1

      • hjrrockies says:

        Very valid point.

        I just want my cold, black, SABR heart to come to terms with CarGo’s Gold Glove.

        Interesting that Seth Smith was, excluding Brad Hawpe, the worst rated Rockies outfielder on FSR, yet he was the only one with a positive UZR for the season. Strange…

        I guess the best way to be a great outfielder at Coors is to play terrible defense by the eye test…?

        Vote -1 Vote +1

      • CircleChange11 says:

        Carlos Gonzalaz UZR — Compare as an A versus as a Rockie.

        In OAK he was getting more PO/INN whether he played CF or a corner OF spot (it is a SSS).

        We know CarGo and DFowler have speed. I find it difficult to believe that even with poor judgement they wouldn’t just be league average (at worst).

        Vote -1 Vote +1

  2. Resolution says:

    I’ve slowly grown a real distrust over UZR’s numbers for Rockies outfielders. CarGo going from positive in CF in Oakland to a negative corner OF at Coors, Willy Taveras being strongly positive in CF before and after his stint on the Rockies (with the Astros and Reds), and then negative at Coors. Dexter Fowler having great defensive scouting reports and then terrible numbers. Something weird is going on…

    Vote -1 Vote +1

  3. Nivra says:

    Does outfield positioning play a role here? Do OF’ers play deeper in Coors? UZR doesn’t change zone based on park, so it might grade OF’ers worse due to all the shorter line drives and flies that drop.

    Another hypothesis is the thin air. Flies and line drives may just move through the air faster at Coors, giving OF’ers less time to field. If that’s the case, then the number of balls they get to within their zone will drop when moving to Coors. This should be reflected in Coors PF. And yep, Coors inflates triples by a lot, doubles quite a bit, and singles by a little ( If this theory is correct, then the majority of the singles PF increase may be due to line drive and flyballs, but the denominator for singles would include all GB singles as well, which would mean the actual increase in LD/FB singles is greater.

    Vote -1 Vote +1

    • Resolution says:

      Technically though, the outfielders can run a bit faster right? Not sure if it’s enough to compensate…

      Vote -1 Vote +1

  4. West says:

    Dexter Fowler is too athletic to be that bad in the outfield, once he learns how to play he’ll be one of the best.

    Vote -1 Vote +1

  5. UZR is a Joke says:

    There are some definite park effects that don’t get accounted for in UZR. Safeco/Petco/Oakland are pitching parks that inflate UZR, while Coors has a negative impact like you point out. I think if you looked at the BABIP by ballpark for the past decade, some fine tuning could be made.

    Vote -1 Vote +1

  6. Breadbaker says:

    Does a “pitcher’s park” inflate UZR, or does a park where more balls are in play increase UZR and thereby become known as a “pitcher’s park” because it is in fact a fielder’s park?

    Vote -1 Vote +1

  7. Mr Punch says:

    Seems to me that there’s something wrong with OF UZRs at the extremes – pitchers’ parks, hitters’ parks, odd cases like LF in Fenway. Infields are (basically) standard; outfields are not, especially LF/RF.

    I’m also somewhat suspicious of UZR at 1B, by the way, because this is a position (like C) where the most important function is handling thrown rather than batted balls.

    Vote -1 Vote +1

  8. CJ says:

    This is one of the teams that DRS (fielding bible) and UZR disagree. DRS has the Rockies at +30, which is similar to the FSR. I generally look at both DRS and UZR before I make conclusions about player defense. Generally the two systems are in agreement, but occasionally DRS and UZR are at polar extremes on a player. My personal preference is to place more weight on DRS, partly because I am skeptical of the ballpark adjustments which were added to UZR a couple of years ago. The adjustments made significant changes to the then existing ratings for a few players which just didn’t make sense.

    Dewan at the Fielding Bible claims that ballpark adjustments are unnecessary for his +/- metric because of the granularity of the grid used in his system (if I recall his explanation correctly). I can’t really judge the validity of his explanation, but “ballpark” is a potential difference between UZR and DRS.

    UZR and DRS are pretty much in agreement on the Giants, by the way.

    Another interesting comparison to examine is the team ranking on Out of Zone plays. A team (like the Mariners) which ranks high on both UZR and OOZ seems to make sense. The occasional team, like the Giants, which rank high on UZR but relatively low on OOZ raises a question mark in my mind. I’m not really sure what it means. I have always been intrigued by players who rank the same on UZR but have big differences on OOZ (after adjusting both to rate stat). My guess is that players with more OOZ plays have a bigger “wow!” factor which affects the fan rating disproportionately. Also, is it possible that such differences reflect a flaw in the way that UZR handles out of zone plays? (I would note that I was just looking at the OOZ rankings for teams without any adjustment for chances…so I don’t know how that would change things.)

    Vote -1 Vote +1

  9. Xeifrank says:

    I love the concept of the “fans scouting report”. I think many of the fans who place ballots for the FSR have a pre-conceived bias based on their likely knowledge of other metrics like UZR, +/- etc… But I think this is a good thing, because the FSR numbers are likely a mix of UZR and scouting. The final FSR numbers are a combination of the two, I am not sure what I’d put the weighting at, but they are in effect regressing each other. Plus FSR gives you catching data and can probably handle 1B defense better than UZR. Just my $0.02.
    vr, Xeifrank

    Vote -1 Vote +1

  10. Scott Gianelli says:

    Is it possible that the shortage of oxygen slows down outfielders enough that they fail to reach some fly balls that, trajectory and hang time otherwise being equal, they would catch at sea level?

    Vote -1 Vote +1

  11. Travis Bickle says:

    Maybe one day all you saber geeks will learn what every G.M. in baseball would tell you if they didn`t care about upsetting thier fanbase. That the uzr is total crap and no MLB exec or player pays any attention to it. Go play your fantasy leagues geeks.

    Vote -1 Vote +1

  12. Kyle McG says:

    You need error bars for the fan scouting report. Since fan scouting is based on a distribution of numbers, you can’t just rely on the mean to tell the story.

    Especially the Rockies. Look at the team table: See how few people are weighing in on the Rockies? Rox should have relatively big error bars.

    Vote -1 Vote +1