Jump to content
Twins Daily
  • Create Account
  • Getting Defensive


    Seth Stohs

    Remember when Doug Mientkiewicz was really good at first base for the Twins one year, but Rafael Palmiero won the Gold Glove despite barely playing the position that year? Defensive analysis and Gold Glove voting have come a long way since then. Whether we agree with each of them, there are now at least several statistical data to measure defense.

    SABR Defensive Index is one of those data points and it is used to help select the Rawlings Gold Glove Award ®. According to their site, “The SABR Defensive Index accounts for approximately 25 percent of the Rawlings Gold Glove Award selection process that was added to the votes from the managers and coaches.”

    A perfect system? No. But a better system? Probably.

    Image courtesy of Wendell Cruz, USA Today

    Twins Video

    The Twins defense has been a topic in the second half, particularly in the infield. So I thought it might be fun to take a look at how the Twins stack up by this metric in its most recent analysis (through games of August 18th).

    Catcher

    Unfortunately for today’s article, there is a game- or innings-played minimum in this analysis, so the Twins don’t have a player at these positions that qualify. Regarding that Rafael Palmiero Gold Glove, this would have been nice back then!

    I think that the Twins feel good about their defense behind the plate. Jason Castro may not be quite what he was in previous years, but he’s still strong. And by all accounts, Mitch Garver has vastly improved his defense, specifically his ability to present pitches. While we can probably agree that their time split has probably helped keep them both fresh throughout the season, it means that neither qualifies for this.

    Pitchers

    I have always thought that Jose Berrios and Kyle Gibson are terrific defensive players. I think both show good range and athleticism. I still believe that, even if this data shows both in the bottom four in the AL, separated only by Lance Lynn.

    However, it also shows that Martin Perez is second on the list, behind only Mike Leake of the Mariners.

    First Base

    CJ Cron comes in at 1.0. I don’t know and won’t take the time to figure out exactly what that means or how it’s calculated. However, it’s a positive number which means he could be considered average or a little better than average.

    I think that’s fair. He got a lot of credit for some scooping early in the season. He hasn’t picked as many of late, but I generally think he’s more than adequate at the position. And, I think we can all acknowledge that he’s not A’s first baseman Matt Olson with the glove.

    Second Base

    Jonathan Schoop comes in at a -2.1 SDI, so again, a little bit below average. That ranks ninth of qualifying second baseman. I think we can all agree that his arm is plus-plus. His range may not be real great as he’s easily the biggest player on this list. Of course, in the last month, Schoop has lost much of his playing time to Luis Arraez.

    Third Base

    And, I don’t think any of us have any thought that Miguel Sano would rank terribly high among defensive third baseman. It would be nice to see what his numbers would look like though.

    Meanwhile, Marwin Gonzalez does show up on the list at 5.1 SDI. Sano has played 66 games at third base. Marwin Gonzalez has played 40 games at third base. In other words, Gonzalez’s numbers must include his time all over the diamond. But, frankly, that makes his 5.1, a decent amount to the positive, even more impressive.

    Shortstop

    Jorge Polanco has certainly struggled in the field the last couple of weeks, but until that point, his defense went generally unnoticed, which is a good thing. His SDI is -0.3, which would say that he’s been about average in 2019. I fact, he fits in at seventh out of 12 AL shortstops.

    Old Friend Niko Goodrum of the Detroit Tigers ranks third at 4.0 SDI.

    Right Field

    Max Kepler is tremendous in right field, and good in center field. He comes in at 6.1 SDI, second among AL right fielders behind only Red Sox Mookie Betts (8.9).

    While Betts will likely win another Gold Glove, Kepler absolutely should be a finalist and get serious consideration. I’m sure having 34 homers already and playing on a winning team help his case.

    Center Field

    Despite missing a decent amount of time this season, Byron Buxton’s 8.9 SDI ranks atop the list of 14 American League center fielders. In fact, Rays OF Kevin Keirmaier ranks second at 6.3 SDI and recently-released Billy Hamilton is third at 5.7.

    In fact, Buxton’s 8.9 SDI ranks fourth among all AL players, regardless of position. He ranks behind only A’s 3B Matt Chapman (13.0), Cleveland catcher Roberto Perez (11.9) and Rays SS Willy Adames (9.9).

    Left Field

    Saving the best for last? No, but saving the most interesting - at least to me - for last.

    Eddie Rosario’s defense has been a bit of a roller coaster throughout his career, and even in the 2019 season there have been plenty of ups and downs.

    Rosario posts a -2.4 SDI, so again, below zero. But what does that mean? He ranks fifth of ten qualifying left fielders. Only two AL left fielders have a positive SDI, and Royals Alex Gordon has a 0.2 SDI. So, I feel fairly comfortable saying that Rosario’s defense this year hasn’t been good, but it isn’t terrible.

    But to the big story… The top defensive left fielder according to the SABR Defensive Index is Oakland A’s Robbie Grossman. And not just by a little bit. Grossman has a 5.8 SDI, well ahead of Gordon’s 0.2.

    So what do we make of that? Does it completely ruin any value that SDI has in your mind? Does it mean that Grossman is actually playing really good defense in 2019? Truly, I don’t know how to read that, and don’t want to read into it. But it is statistical, analytical, based on the same information that every other left fielder is evaluated by.

    So, what do you think?

    It feels like - aside from the Grossman thing - this data kind of verifies what we probably thought of the Twins defensive play in 2019. It verifies that Byron Buxton is amazing and that Max Kepler is really good. Marwin Gonzalez has provided defensive value wherever he’s played. Beyond that, there are several very average defenders, and despite his recent struggles, Polanco has been OK for most of the year.

    What else do you see from this data?

    MORE FROM TWINS DAILY
    — Latest Twins coverage from our writers
    — Recent Twins discussion in our forums
    — Follow Twins Daily via Twitter, Facebook or email
    — Become a Twins Daily Caretaker

     Share


    User Feedback

    Recommended Comments

    Featured Comments

    Robbie Grossman - Oakland A's - LF

     

    Fangraphs:   Ultimate Zone Rating = +5.9/150 games

    Baseball Reference:   Rtot/yr = +19, Rdr/yr = +6.0

    Baseball Gauge:   Defensive Regression = +6.3 Runs above average

     

    By far the best defensive numbers he's had in his career. Beats me. Did he get new eyes and is getting a better jump? Better routes? Easy to play LF in Oakland? It's really someone else in a Robbie Grossman suit?

    Link to comment
    Share on other sites

    Robbie Grossman - Oakland A's - LF

     

    Fangraphs: Ultimate Zone Rating = +5.9/150 games

    Baseball Reference: Rtot/yr = +19, Rdr/yr = +6.0

    Baseball Gauge: Defensive Regression = +6.3 Runs above average

     

    By far the best defensive numbers he's had in his career. Beats me. Did he get new eyes and is getting a better jump? Better routes? Easy to play LF in Oakland? It's really someone else in a Robbie Grossman suit?

    or possibly publicly available defensive metrics aren't worth the electrons they're written with?
    Link to comment
    Share on other sites

    It demonstrates that we have no real accurate way of judging defense and there for the defense portion of war should be removed since it has no consistency or real basis in reality

    This. Outfield especially...aren’t even directionally correct/consistent. Makes sense when who makes the putout (center fielder or corner fielder) is often arbitrary...and each player defends a different area and geometry based on ballpark (among many other problems with current data). Meanwhile, infield metrics will get less reliable before they get more, given all the variances in how/when teams shift. Some day we will laugh at the era where we had defensive metrics that only considered results, without accounting for time, distance, two and three dimensional coordinates, etc...and without accounting for impacts of alignments and adjacent defenders. Meanwhile, I immediately back out the defensive component to a player’s WAR when considering any argument for/against that player based on WAR. Or simply go to his wRC+, and let my eyeballs and common sense add/subtract if there is anything on the defensive side to warrant it.

    Link to comment
    Share on other sites

     

    This. Outfield especially...aren’t even directionally correct/consistent. Makes sense when who makes the putout (center fielder or corner fielder) is often arbitrary...and each player defends a different area and geometry based on ballpark (among many other problems with current data). Meanwhile, infield metrics will get less reliable before they get more, given all the variances in how/when teams shift. Some day we will laugh at the era where we had defensive metrics that only considered results, without accounting for time, distance, two and three dimensional coordinates, etc...and without accounting for impacts of alignments and adjacent defenders. Meanwhile, I immediately back out the defensive component to a player’s WAR when considering any argument for/against that player based on WAR. Or simply go to his wRC+, and let my eyeballs and common sense add/subtract if there is anything on the defensive side to warrant it.

    Well said.  

    Link to comment
    Share on other sites

    Any stat that tells me RG is the best LF in baseball is puzzling, or worse. But regardless of how accurate defensive stats are, or aren't, you don't need analytics to see that this Twins team has sunken into a defensive morass......... There are days it's borderline embarrassing.

    Link to comment
    Share on other sites

    Defensive stats are half-baked. As long as defensive stats are included in WAR... War is probably three quarters-baked. 

     

    Grossman on top of any stat will tell you that. 

     

    He's standing in the right spot often enough and that's all it takes. Grossman could be last next year because he wasn't standing in the right spot often enough. 

     

     

    Link to comment
    Share on other sites

    or possibly publicly available defensive metrics aren't worth the electrons they're written with?

    while I agree defensive metrics have their shortcomings, benchmarking is a very strong validation tool. If even one of our metrics said different things about Grossman, it’d be fair to throw out the whole sample as hogwash. All 4 were directionally aligned.

     

    I leave “how good” absolutely up for debate. I don’t think defensive metrics have gotten the shades of the color palette, but give you a decent stoplight. I still think Robbie “the butcher” Grossman is bound for going back to his old ways.

    Link to comment
    Share on other sites

    Looking at the varied measures I think it is fair to say that Robbie has had a good season defensively thus far. Like many pitching and hitting stats those numbers tell a reasonably good story of a time frame but are not a good tool to predict future defensive performance.

     

    I wouldn’t want the Twins to make decisions based on a single season’s slash stats or ERA. I wouldn’t want them to make decisions a single seasons defensive metric either. I wouldn’t imagine that the A’s are using Grossman as a defensive replacement. His metrics and play for the Twins his first year were awful. He is not that bad. He isn’t this good either. Below average but not bad outfielder with skill to get on base.

    Edited by jorgenswest
    Link to comment
    Share on other sites

    I wouldn’t want the Twins to make decisions based on a single season’s slash stats or ERA. I wouldn’t want them to make decisions a single seasons defensive metric either.

    A single season's hitting or pitching stats contain a vastly greater number of tests of the player's skill in those areas, than fielding.

     

    An outfielder has had a busy season if he collects 300+ total chances. Many of those chances are absolute cans o' corn. Conversely, and unfortunately, many other opportunities to excel (or to fail) are never recorded as total chances, and are recorded simply as "hits" for the opposing batter, even though the ball came into an area of the field where under other circumstances a play might have been successfully made. Other "hits" are truly unreachable by even the fastest player - the distinction is not noted in the traditional fielding stats.

     

    Analogously, a batter faces a certain percentage of pitches that essentially every batter would do the same on - 55-foot breaking pitches, fastballs that sail three feet outside, strikes taken on 3-0, and so forth. Those are like the cans o' corn, and/or the balls hit totally out of reach of a fielder.

     

    But a busy season at bat consists of nearly 3000 individual pitches; these chances to excel or fail get combined into hits and outs and eventually the slash-line. So when you slice and dice, "cans o' corn" versus more challenging plays (fielding and batting analogues to one another), there is 10 times the quantity of data to judge the hitter (or pitcher) on, than for fielding. (Shortstop is slightly less troublesome, with about 2X the total chances of center field.)

     

    We are accustomed to looking skeptically at Small Sample Sizes of batting stats - 100 PA may not tell you much about the player. More than 400 pitches might be involved in that.

     

    A season of defensive data is inherently SSS.

     

    Robbie Grossman may have suddenly gotten better in the outfield. Or, more likely, his 2019 has been the fielding equivalent of Hurricane Hazle (or our Danny Santana), hitting amazingly for a quarter of a season, until suddenly the pixie dust disappears.

    Link to comment
    Share on other sites

    DRS correlates similarly season to season to slash stats and better than ERA. In a season sample I would trust them similarly and not be surprised by outliers like Danny Santana or Allan Anderson.

     

    My trust in all three as a predictor of future performance is low for a single season where the player is performing very differently from his norm.

    Link to comment
    Share on other sites

    The gold glove is still tied to how well you hit at position and not how well you play that position and then you add in player recognition and who you play for are all determining things in gold gloves. That is truth if it was all about defensive stats we would have quite a few different gold gloves over the years.

    Link to comment
    Share on other sites

     

    DRS correlates similarly season to season to slash stats and better than ERA. In a season sample I would trust them similarly and not be surprised by outliers like Danny Santana or Allan Anderson.

    My trust in all three as a predictor of future performance is low for a single season where the player is performing very differently from his norm.

    IMO one huge difference between current defensive stats, and things like slash stats is, while slash stats might be unsustainable, or lucky, or whatever, there is zero doubt they measure what actually happened. Player X got on base 175 times in 500 PAs, his OBP is .350. Maybe that's not a good way to estimate next year's OBP for that player, or maybe it is, but at least we're dealing in fact, about things that actually happened, and are accurately measured.

     

    The same can't be said about any of the defensive stats. They're simply numerical representations of what somebody, somewhere, decided they would attempt measure as a way of estimating how effective a given player's defense is.

     

    We don't even know if what they measure is important, nor if they measure it properly if it IS important.

     

     

    Link to comment
    Share on other sites



    Join the conversation

    You can post now and register later. If you have an account, sign in now to post with your account.
    Note: Your post will require moderator approval before it will be visible.

    Guest
    Add a comment...

    ×   Pasted as rich text.   Paste as plain text instead

      Only 75 emoji are allowed.

    ×   Your link has been automatically embedded.   Display as a link instead

    ×   Your previous content has been restored.   Clear editor

    ×   You cannot paste images directly. Upload or insert images from URL.

    Loading...

×
×
  • Create New...