I’m a nut for fielding statistics, but until recently I could only get my hands on data back to 2000. Recently, however, Sean Smith devised an ingenious method to analyze fielding from as far back as retrosheet.org provides play-by-play data, 1956. Combined with STATS’ zone rating numbers from 1987 through 2007 as compiled by SG over at Replacement Level Yankees Weblog, we now have a way to grade every fielding season of the past 50 years. Only range (converting batted balls into outs) is considered—no outfield throwing arms, double-plays, or catcher data are included. Here are some things I learned by pivot-tabling my way through Mets’ history:

A Tradition of Ineptitude

From 1962 through 2007, the Mets as a team are 1013 runs worse than average. That’s about 22 runs per season, or two-and-a-half wins, on average. Sure, they’ve had good seasons—led by the 1969 Miracle Mets’ total of +66 runs (more on them later)—but they’ve had their clunkers, too. Here’s a pictorial history, smoothed by weighting the current season at 50% and the two bookending it at 25%:

Here are the Mets’ three worst fielding teams in their history:

• 1982 takes the bronze model, with a total of -66 runs compared to average. Dave Kingman (-15 runs) led the charge at first base, but Mookie Wilson (CF), George Foster (LF), Joel Youngblood (RF), Ellis Valentine (RF), Hubie Brooks (3B), and Tom Veryzer (SS) were all at least five runs worse than their positional average. That’s right—every position contributed negatively.
• As an expansion club, the New York Metropolitans came out of the gate appropriately, finishing the 1962 season at -79 runs. No single player was worse than -9 runs (Charlie Neal at second base), but this team put 37 players in the field, seemingly all of which were poor fielders. Casey Stengel ran out eight third basemen (-13 total), six right fielders (-16 total), six first basemen (”only” -3 total), five left fielders (-7 total), and five center fielders (-10 total). Second base and shortstop were more stable, but no more affective with a combined -30 runs.
• In 1988, the Mets reached their nadir of -111 runs thanks to such impressive performances as Mookie Wilson (-17) and Lenny Dykstra (-7) in center field, Kevin McReynolds in left field (-29), and Darryl Strawberry in right field (-30). Even the immortal Keith Hernandez was merely average over at first base and didn’t play enough to keep Dave Magadan’s -5 run glove on the bench. The best two fielders were Tim Teufel at second base (+4 runs) and Kevin Elster at shortstop (+7 runs).

What’s amazing about that -111 number is that the ‘88 Mets won 100 games and sported a 2.91 ERA, only allowing 532 runs in 160 games. If not the fielders, who gets the credit for that level of performance? Well, a rotation of Dwight Gooden, Ron Darling, Sid Fernandez, Bob Ojeda, and David Cone is a great start. Then add in Shea Stadium, which suppressed run scoring by 13%, and an NL run-scoring environment of 4.15 runs per game (as opposed to today’s 4.90). If the team had fielded just a league-average defense, the Mets’ team ERA could have been .70 runs lower, at 2.21—a disgustingly anemic number. Who needs defense when you’ve got an historically great rotation?

Silver Linings

The Mets have sported some very good defenses, too. As you probably remember, the 2006 team was quite good, coming in at +40 runs. Jose Valentin (2B), Jose Reyes (SS), Carlos Beltran (CF), and Endy Chavez (RF/CF) were all at least 10 runs better than average, with David Wright (-9) the only weak link over at third base.

But the best fielding team in franchise history were the Miracle Mets of 1969, coming in at +66 runs. Tommy Agee (CF), Cleon Jones (RF), and Bud Harrelson (SS) were all about a dozen runs above average, and the third base rotation of Bobby Pfeil, Ed Charles, and Kevin Collins combined to meet that level as well. This team won 100 games, finishing eight ahead of the Cubs. Their run differential, however, implied they were only a 92-win team, making the fielding contributions critical and a big reason the ‘69 team won the organization’s first World Series (the 1970 team was also impressive, as you might expect, coming in at +46 runs.)

All Time Leaders and Trailers

Here are the players who put together the best seasons at each position:

Pos     Player               Runs      Year
1B      Todd Zeile            +18      2000
2B      Jose Valentin         +13      2006
3B      Robin Ventura         +22      1999
SS      Rey Ordonez           +21      1999
LF      Cleon Jones           +15      1969
CF      Tommy Agee            +13      1970
RF      Joel Youngblood       +14      1980

Zeile might be number one at first base, but Keith Hernandez owns spots two through five with seasons of +15, +12, +10, +8. Two infielders from the 1999 team make the list—the right side was manned by John Olerud (+8) and Edgardo Alfonzo (+8) for a team total of +48. The next best infield was in 1969 with +32 runs.

How about the worst seasons at each position?

Pos     Player                Runs      Year
1B      Dave Kingman           -15      1982
2B      Gregg Jefferies        -21      1989
3B      Howard Johnson         -26      1989
SS      Frank Taveras          -15      1981
LF      Kevin McReynolds       -29      1988
CF      Mookie Wilson          -17      1988
RF      Darryl Strawberry      -30      1988

It’s no surprise why the late 90s were awful fielding seasons for the Mets given the table above. McReynolds’ and Strawberry’s seasons are especially impressive since they’re so much worse than the runners-up. The second worst left field season in Mets’ history belongs to Moises Alou’s 2007 campaign at -9 runs. Second worst among right fielders was Rusty Staub in 1974 at -14 runs.

How about over entire careers? Here’s who saved the Mets the most runs at each position during their tenure with the team:

Pos     Player                Runs     Seasons
1B      Keith Hernandez       +437       6
2B      Jose Valentin          +13       2
3B      Robin Ventura          +37       3
SS      Bud Harrelson          +45      13
LF      Cleon Jones            +36       9
CF      Tommy Agee             +27       5
RF      Joel Youngblood        +16       6

Seriously, Jose Valentin leads at second base with a mere +13 runs over basically one full season. No other Met has saved 10 runs more than positional average over his career.

And the fewest career runs at each position:

Pos     Player                Runs    Seasons
1B      Dave Kingman           -19      6
2B      Gregg Jefferies        -45      4
3B      Howard Johnson         -70      8
SS      Frank Taveras          -37      3
LF      Kevin McReynolds       -26      6
CF      Lee Mazzilli           -32      8
RF      Darryl Strawberry      -31      9

The short-lived Frank Taveras and Gregg Jefferies experiments should have ended even sooner. It’s great to see such big names leading in fielding awfulness. Seriously. I mean, this list just wouldn’t be as fun with players like DJ Dozier on it. HoJo at -70? I love it.

This is only the tip of the historical fielding iceberg, even just considering the Mets. Want to slice and dice the data yourself? Here’s a list of historical data sources. Have a question and want me to do the work for you? Leave a comment or email me.

---

Sky writes about baseball at his own blog, skyking162. He's 28, teaches high school math, and is a racket sport aficianado.

---

14 Responses to “Getting Defensive”

1. Comment posted by littlefallsmets on February 5, 2008 at 12:57 am (#605403)

   The interesting thing about defense in baseball is that… as much as it is somewhat under-rated by the casual fan?

   It is over-rated by the hardcore fan.

   Gimmie Straw-in-his-prime’s offensive explosiveness anyday. Gimmie Mookie’s solid hitting and his speed… okay, maybe at a corner instead of centre but… please, I will take that any day.

   Anyone actually taking Robin Ventura over Hojo or Wright? Anyone taking Brian Schneider over the Pizza Man in his prime? No, no, of course not.

   Below-average defense and a good bat? Not-completely-disasterous D and a great bat Defensive prowess is a great… add-on, it is a great tie-breaker with two otherwise-mostly-equal players.

   It isn’t spectacular-catch-ball, it’s baseball.

   If you can’t get on base, you lose.

2. Comment posted by JamesSC on February 5, 2008 at 9:10 am (#605414)

   Well, I agree with you LFM on one aspect, defensive stats to me are almost useless. To me defense is as much about the confidence your pitching staff has to pitch in the strike zone and be picked up by his teammates as it is about zone ratings and what not. To me defense is something that IS a product of what you see and not a bunch of stats and I will always feel that way no matter what stats they come up with.

   However, I would take Robin Ventura in his prime over HoJo in his prime and comparing Schneider to Piazza is silly because it is comparing apples to oranges, but you know there are a LOT of baseball people who would take Rodriguez over Piazza because of his defense.

   Obviously offense is a key to evaluating any player, but defense is important as well, to ignore either would be foolish.

3. Comment posted by andyglass1 on February 5, 2008 at 10:28 am (#605471)

   I seem to remember Kevin McReynolds being called the best defensive LF out there back in the late 80’s ….. So apparently this was just the NY press hyping up the local players ? He started out as a CF so how could he be so poor as a LF ? Robin Ventura was definitely an excellent all round player; too bad we only saw him for 3 seasons. We all knew Jefferies was challenged at 2b but these stats are amazing. Who knows how good he could have been without all the experimenting…. I’m surprised there’s no mention of Cameron, Olerud & Grote; they were 3 of our better defensive players.

4. Comment posted by Joe A. on February 5, 2008 at 10:33 am (#605476)

   If you had asked me who the best defensive LF ever to play for the Mets was, I would have said McReynolds. Not sure if that means these stats are bad or my memory is bad.

   I don’t find any of the other guys listed on the good or bad lists surprising at all.

5. Comment posted by Joe A. on February 5, 2008 at 10:37 am (#605479)

   I’m surprised there’s no mention of Cameron, Olerud & Grote; they were 3 of our better defensive players.

   Grote’s not included because catchers weren’t included. I didn’t think Cameron lived up to his reputation during his time with the Mets. And Olerud was very good, but I wouldn’t expect to see him ahead of Keith. What’s really surprising is that Zeile beat Keith’s best year.

6. Comment posted by davidg on February 5, 2008 at 11:05 am (#605507)

   Sky -

   Thanks for compiling the data. Defensive metrics always have issues, but looked at with a deep enough sample, some general trends come out - and I think they do here.

   One of the “mystiques” of the Met history that Omar (and others) like to talk about is “pitching and defense”. He believes that that is what brought the Mets their championships in years past. Pitching - yes, defense - not really. It’s completely overrated as a metric on which to build a team. That’s not to say a GM should find the worst fielders possible - it’s just that other criteria are far more important - namely pitching and hitting.

   When the Mets dominated the league the late 80s (86-88) winning on average 100 games each season, they were an average to below average defensive team (as your stats show). However they were filled with great pitchers and excellent hitters.

   Regarding today’s team, Omar has pitching at the top of his list and he has tried to improve that element of the team. But his judgment and moves have had mixed results - helping the team in 2006, hurting the team in 2007. Overall he’s had more success in his transactions for starters. Getting Santana, signing Pedro and trading for John Maine and Oliver Perez were all high caliber transactions. On the flip side,though, most of Omar’s bull pen decisions have been suspect including: letting Chad Bradford go, signing Mota and Schowenweis to large contracts and trading Heath Bell, Matt Lindstrom, Henry Owens and Brian Bannister. In general, Omar’s pattern is clear on pitching - he makes foundation moves for starting pitching and throws a lot of spaghetti on the wall to figure out his bull pen.

   Around defense, Omar may believe too much in that element - but so far it hasn’t hurt the team in a big way. Although the trade of Milledge for Church and Schneider, the signing of Castillo (who was once a gold glove, but now simply adequate) for an excessively long contract, and the drafting of many amateur players that don’t offer much more than a good glove may have an effect on the marginal performance of the team.

   Hopefully, defensive metrics continue to improve and Omar’s understanding about this element can be better understood in composing future Met teams.

7. Comment posted by John Peterson on February 5, 2008 at 1:30 pm (#605600)

   Historically, defense has been overrated, with offensive black holes playing as regulars because it was thought that their defense “saves 10 games a year,” and the like.

   Now, teams seem to understand that defense doesn’t make up for a poor offense, but they’ve taken this too far in some cases, not recognizing when a player is really bad, and costs his team a couple games a year, or really good, and saves his team a few runs.

   Two examples: Derek Jeter should have been playing third base for the Yankees, and A-Rod short. Derek Jeter is easily the worst defensive shortstop in the game today. A-Rod is not spectacular, but he would be average, as Jeter would be at third. The Yankees shouldn’t switch Jeter out for Rey Ordonez, but by switching A-Rod and Jeter they could save themselves some runs.

   More recently, the Astros acquired Miguel Tejada and just non-tendered Adam Everett. Tejada, an average shortstop in his prime, is declining defensively. Everett, on the other hand, is one of the absolute best defensive shortstops in the game. With a hole at third base, the Astros could have made an optimal move by putting Tejada over there, and leaving Everett at short. A very poor offensive player, Everett’s defensive is so good that it makes him well above average as a player. But the Astros failed to realize that, and wouldn’t even pay him one million dollars.

8. Comment posted by redstripe n chronic on February 5, 2008 at 6:55 pm (#605864)

   I think defense is a little more subjective than all this, but still interesting to think about.

   That 99 infield (less Piazza) was amazing defensively.

   What are Luis Castillo’s numbers like over the past few years? My guess is average and trending down, but at least he makes contact a lot. How about Schneider’s? (I am too lazy to look these up, but I’m sure some interested geeks out there have done it).

9. Comment posted by Sky Kalkman on February 5, 2008 at 8:17 pm (#605905)

   LFM — I’m not sure anyone would argue that offense plays a bigger role than fielding in determining the outcome of games and player value. The cliche is that offense is 50% of the game while fielding is 10-15%.

   But most players aren’t as different as Mike Piazza and Brian Schneider offensively. That’s why it’s important to put offense and defense on the same scale — runs. Treating fielding as a qualitative footnote to offense is less productive.

   And it’s not just how well you field your position, it’s what position you play. A guy who hits .280/.350/.450 is way more valuable in center field than at first base — about a win and a half.

10. Comment posted by Sky Kalkman on February 5, 2008 at 8:40 pm (#605919)

    AndyGlass1 — McReynold’s 1988 season was by far his worst as a major leaguer. Strawberry’s, too.

    McReynolds rates in CF as +18 in 1984, +10 in 1985, and -4 in 1986, all with San Diego. In left field with the Mets, he’s +4 in 1987, -29 in 1988, then basically anywhere between -5 and 0 the rest of his career. 1987 is the first season using STATS zone rating instead of Sean’s retrosheeet system. I doubt he was really that bad in ‘88, but I have no idea what might be up with that team. Any ideas from die-hard Mets’ fans?

    Mike Cameron rates as +4 in CF in 2004 and +4 overall splitting time between CF and RF in 2005. By STATS zone rating, he doesn’t come across as impressive as his reputation or something more advanced like UZR (more like +15 per season). STATS zone ratings do tend to have trouble with center fielders because they don’t account for batter handedness (center fielder will shift) and don’t include plays made outside of a fielder’s zone.

    John Olerud had seasons of +2, +8, and +8 from 1997 to 1999 with the Mets. Over his career he’s mostly in the +4 to +9 range

    Grote was a catcher, and catcher metrics are even fuzzier than range numbers. Sean’s system does have him at +2 runs per 1200 innings, which is about a full season.

11. Comment posted by Sky Kalkman on February 5, 2008 at 8:47 pm (#605921)

    DavidG — Very interesting about the Mets’ reputation as a fielding-first organization. I wasn’t aware of that.

    It’s really interesting how the “defense wins championships” mantra doesn’t pop up in baseball as much as other sports. Well, pitching’s overhyped for successful teams, I suppose. Fielding tends to be vastly underrated in terms of paying players money for it. For example, take Pedro Feliz’s contract, the Troy Tulo vs. Ryan Braun RoY debate, and Houston’s disrespect of Adam Everett. If you tell people that Khalil Green and Carlos Lee don’t have similar value, they’ll agree. But most people think Lee’s the better player, when it’s really Greene by a win or two.

12. Comment posted by Sky Kalkman on February 5, 2008 at 8:52 pm (#605927)

    Redstripe — Castillo’s on the way down, but did have a couple very good seasons.

    Team Year Pos Runs
    Fla 1996 2B 3
    Fla 1997 2B 1
    Fla 1998 2B 2
    Fla 1999 2B -1
    Fla 2000 2B 6
    Fla 2001 2B 3
    Fla 2002 2B 1
    Fla 2003 2B 11
    Fla 2004 2B 12
    Fla 2005 2B 5
    Min 2006 2B -2
    NYN 2007 2B -4
    Min 2007 2B -1

    Schneider’s at +8 per 1200 innings, not including 2007. That’s pretty good.

13. Comment posted by Alex Nelson on February 5, 2008 at 10:34 pm (#606006)

    McReynolds rates in CF as +18 in 1984, +10 in 1985, and -4 in 1986, all with San Diego. In left field with the Mets, he’s +4 in 1987, -29 in 1988, then basically anywhere between -5 and 0 the rest of his career. 1987 is the first season using STATS zone rating instead of Sean’s retrosheeet system. I doubt he was really that bad in ‘88, but I have no idea what might be up with that team. Any ideas from die-hard Mets’ fans?

    I think a lot of that might be Shea Stadium. I’ve often heard conjecture that Shea’s left field is very difficult to play thanks to the swirling winds that make Shea such a pitcher’s park. I know there were some who thought Todd Hundley should have been tried in right rather than left in 1998, because right field might actually be easier to play.

14. Comment posted by Sky Kalkman on February 6, 2008 at 8:47 pm (#606723)

    Any ideas why 1988 would be so different from other seasons, though?

Leave a Reply

You must be logged in to post a comment.