Every baseball fan has an opinion about baseball's advanced defensive statistics. It's understandable, really. Baseball defense is a tough nut. It is a much blurrier skill to measure than, say, hitting, for a pretty obvious reason. In hitting, you know EXACTLY what you're measuring. Every plate appearance is a starting point.
All you need to do is count how successful the hitter was in that plate appearance. We have come up with dozens of ways to measure success. Did he get a hit? An extra base hit? Did he walk? Did he strike out? Did he drive in a run? Did he move the runner over? And so on. But the plate appearance is your constant. It always starts there.
For many years, people tried to judge defense the same way. Most still do. That, I think, is where the "error" concept comes in.
See, fielding percentage and batting average are really measuring in the same way. Batting average measures how how many hits you get per at-bat. Fielding percentage measures how many plays you make based on total chances. They come from the same kind of mindset, the mindset of measuring how often you are successful based on opportunities.
But, it seems to me, the main problem is this: Fielding is not like hitting. Not at all. It isn't about how many plays you make based on opportunities. It is about how many opportunities you give yourself. Let's pretend there are two center fielders who face PRECISELY the same number and type of balls over a full season.
Center fielder A records 375 putouts without a single error.
Center fielder B records 410 putouts but makes 9 errors.
Who had the better year? Well, in this case it's easy … Center fielder B. Even with the errors, he made 35 more plays than the other guy. This isn't even close. But the problem with defense is, it isn't ever easy like that. No two center fielders face precisely the same opportunities over a season. There are fly-ball pitchers, ground-ball pitchers, hard-throwing pitchers, soft-throwing pitchers. And even if my some miracle two fielders DID, impossibly, face precisely the same barrage -- precisely the same fly balls, line drives, gappers, wall balls and so on -- they still would face different weather conditions, different sun angles, different ballpark dimensions and they would probably be lined up in different places depending on the manager's defensive strategies.
In this way, it makes little sense to judge baseball in a linear way. You need to think spatially. There are so many variables, they are so hard to measure, and this is why I think many are so skeptical about defensive stats. What statistic can really get at all these factors and unknowns? How can you compare a left fielder in Fenway with a left fielder in Colorado? How can you compare someone who plays on grass against someone who plays on turf? How can you judge whether a player SHOULD have reached that ball or COULD have made that play?
This, I think, is why so many people have through the years judged defense aesthetically, like art. We let our eyes tell us. We watch them play. We see them dive. We see them botch. We see them make smart plays. We see them overthrow the cutoff man. We see them pull back home runs. We see them allow grounders through their legs. And, over time, we come to conclusions about them.
This eye test does have a lot of value. It is the senses -- more than any statistic -- that can grasp and appreciate the magnificence of Willie Mays in center field or Ozzie Smith at shortstop or Johnny Bench behind the plate. Without the numbers, a .300 hitter will look a lot like a .270 hitter much of the time. But Brooks Robinson looked different from other third basemen, Keith Hernandez looked different at first base, Roberto Clemente was a marvel in right field. Who even knew their error totals? Who cared?
The trouble is -- and this is hard for many of us to admit -- the eyes fall hard for optical illusions. There was a time during this postseason when I saw an outfielder get a terrible jump on the ball, run it down, catch it, then I heard the announcer go on and on about what a nice play he made. An inning or two later, there was a ball hit into the gap, and the same outfielder got a fabulous jump on the ball, ran it down and made what the announcer referred to as a routine play. There is no doubt in my mind that the second play wasn't just harder but MUCH harder than the first. It just didn't look that way, and it wasn't described that way.
That stuff happens a lot. Take Dave Winfield. The basic numbers state unequivocally he didn't get to many more balls than average right fielders, and he didn't make fewer errors. Defensive WAR may not be your cup of tea, but it does calculate he was 23.7 wins WORSE than a replacement defender, which is staggeringly bad. But MAN did he have a great arm, and he had this beautiful way of throwing -- he would unfurl, like a big American flag, and he would uncork these majestic throws, and it was thrilling to watch. He won eight Gold Gloves.
So, finding a judging system that values the artistic beauty of great defense while also grading the players on their actual defensive production … well, I don't think we've found that system yet.
But, I must say … I do like the John Dewan plus-minus and runs saved statistics. I've explained these before: Dewan and his staff go back through video of every game in the major leagues, and they chart every ball hit in play. They then compare players to each other. A moderate-speed ground ball is hit two feet to the left of second base … how many shortstops get to that ball? A line drive hit to a certain spot in right field, how many right fielders catch it? And so on. Of course, there remain countless factors that defy fairness. But this is probably the most serious effort out there to calculate a player's defensive contribution.
The plus-minus part is simple. If a player is plus-10, John and his staff have determined that player made an aggregate 10 more plays than the average defender. Runs saved goes up another level and tries to quantify exactly how many runs the player saved (or cost) his team over an average defender. They are cool stats with a lot of work put into them. I'm not saying they are perfect. I don't think they are perfect. I'm sure numerous people could write me an essay tearing the statistic down to nothing. But I repeat: They are an orchestrated effort.
The runs saved statistic, more than any other, is a key component in the Fielding Bible Awards every year. A panel of 10 -- I'm fortunate enough to be on the panel along with Bill James, Rob Neyer, Strat-o-Matic Inventor Hal Richman, Peter Gammons, everyone who votes in the Tom Tango fan poll, ESPN's Mark Simon, the excellent Doug Glanville and others -- vote. Nobody is required or even encouraged to use runs saved. But we are given access to the numbers. And I have found that most of us DO use those numbers, at least as a part of the voting. I do.
And so: With that in mind, let's see how the Fielding Bible winners match up to this year's Gold Glove winners.
Catcher
AL: Matt Wieters, 5 runs saved (5th in baseball)
NL: Yadier Molina, 16 runs saved (1st)
Fielding Bible: Yadier Molina
There are no Dewan plus/minus numbers for catchers, but he does do runs saved. He has Wieters saving five runs this year, mostly because of his ability to shut down the running game, and he has Molina saving 16 runs, most in baseball. I would say these were excellent choices by the Gold Glove voters.
First base
AL: Mark Teixeira, 17 runs saved (1st)
NL: Adam LaRoche, 8 runs saved (5th)
Fielding Bible: Mark Teixeira
The runs saved numbers have been all over the place for Teixeira.
2007: 0 runs saved
2008: 21 runs saved
2009: 2 runs saved
2010: 6 runs saved
2011: 3 runs saved
2012: 17 runs saved
Yes, some years, have shown him to be an average first baseman. Other years, he emerges as the best first baseman in baseball. This was one of those years when he emerged.
People have said that the wide variations from one year to the next on certain players shows that defensive numbers are fundamentally flawed. Maybe they are fundamentally flawed, but I don't think that's the reason. Players' offensive numbers fluctuate pretty wildly too. I don't think the fact that one year Tex hit .308 and another year he hits .248 and another year he hits .306 and another year he hits .252 tells you that there's something fundamentally flawed about batting average.*
*There IS something fundamentally flawed about batting average, but that's not it.
Second base
AL: Robinson Cano, 15 runs saved (2nd)
NL: Darwin Barney, 28 runs saved (1st)
Fielding Bible: Darwin Barney
Again, there is strong, complete agreement.
Shortstop
AL: J.J. Hardy, 18 runs saved (3rd)
NL: Jimmy Rollins, minus-8 runs (30th)
Fielding Bible: Brendan Ryan, 27 runs saved (1st)
Well, the agreement had to end somewhere. Two things here, both real flaws with the Gold Glove system in my opinion.
One … Brendan Ryan is, I'm pretty sure, the best defensive player in baseball. Period. All positions. And he can't hit a lick. Both those things. In fact, those things are perfectly in sync … the only way someone could get 470 plate appearances in the big leagues with a .277 on-base percentage and a .278 slugging percentage is if:
1. He's a former superstar who had fallen off the cliff before managers could adjust and bench him (not Ryan).
2. He's a brilliant young prospect who the team is playing before he's ready to build for the future (not Ryan).
3. He can REALLY play some serious defense.
Yep, that's Ryan. He's an artist at shortstop. He's beautiful to watch play defense. There seems absolutely no way that anyone could miss this. J.J. Hardy is a good defensive shortstop too, but he's Salieri. Ryan is Mozart. Thing is, Ryan plays for a mediocre Seattle team that plays on West Coast time, and he didn't hit, and he didn't win. The Gold Gloves should never, ever miss a defensive player like Brendan Ryan.
The other problem: They will keep voting for a guy long after he lost all effectiveness. Jimmy Rollins was once a fabulous defensive shortstop. He won the Fielding Bible award in 2008 and, in my view, was absolutely the best defender in the game that year. But it's just not true anymore. The Dewan numbers indicate it -- minus-9 last year, minus-10 this year. Other defensive numbers like UZR indicate it too. But you can see it too. I don't know if Rollins is really a below-average shortstop now -- maybe the numbers overstate that -- but he's not the best in the league. I'm pretty sure Clint Barmes, who has many of the same perception problems as Ryan, was the best.
One other point: For years and years, the best defensive shortstops were all in the National League. This -- and I don't mean this in a snarky was -- was why Derek Jeter kept winning Gold Gloves. The Fielding Bible winner every year but one was a National League (the one AL winner Jack Wilson, had just come over from the NL). You had the fielding wizardry of Adam Everett, Troy Tulowitzki, Jimmy Rollins, J.J. Hardy, Brendan Ryan, Ian Desmond and so on. The American League was pretty light on legitimate Gold Glove candidates.
This year, though, I would say Ryan, Hardy, Yunel Escobar, Cliff Pennington, Alcides Escobar and Elvis Andrus might have been better than anyone in the National League.
Third base
American League: Adrian Beltre, 13 runs saved (4th)
National League: Chase Headley, minus-3 runs (24th)
Fielding Bible: Adrian Beltre
The great Beltre … he really does belong in the conversation with Brooks Robinson and Billy Cox and the Boyers and Buddy Bell and Graig Nettles and Pie Traynor as the greatest defensive third basemen in baseball history.
That said, though, Brett Lawrie had by the numbers an ASTONISHING defensive year at third base. He saved 20 runs this year. Other defensive numbers also picked up his excellence. At one point, his defensive numbers were so staggeringly good he was listed among the league leaders in Wins Above Replacement. I didn't see Lawrie that much this year, but some people I know in Toronto say he really was amazing night after night. He could have been rewarded for that.
Left field
American League: Alex Gordon, 24 runs saved
National League: Carlos Gonzalez, minus-13 runs saved
Fielding Bible: Alex Gordon
In this, I have sympathy with the voters. There really aren't many great defensive left fielders. There's Gordon … and there's Gordon. And there's Brett Gardner, when he's healthy. And there's Carl Crawford from three or four years ago. And there's Gordon.
The numbers suggest Cargo was pretty brutal out there and a pretty bad choice. Martin Prado or Ryan Braun would have been a much better choice. One thought: Gordon and Braun both started out as third basemen. Braun was terrible at third. Gordon was below average and sinking. They might be the two best defensive left fielders in the game now …
Center field
American League: Adam Jones, minus-16 runs (35th)
National League: Andrew McCutchen, minus-5 (25th)
Fielding Bible: Mike Trout, 23 runs saved (2nd)
OK, I know this will tick off some Baltimore people -- and I don't like doing that because I love Baltimore and love the Orioles and really enjoy watching Adam Jones play -- but I don't get the Adam Jones defense thing. Do … not … get … it. The numbers consistently and across the board suggest he's, at best -- AT BEST -- an average defensive center fielder. He's had a minus defensive WAR each of the last three seasons. He's had a substantially negative Ultimate Zone Rating (UZR) the last three seasons too. It's POSSIBLE these numbers are all wrong, but even if they are wrong, it's very hard to imagine them being SO wrong that Jones is not only better than the numbers but actually the best defensive center fielder around.
One quick story: I was watching an Orioles game last year with a huge Orioles fan, and he was ranting about how good a center fielder Adam Jones is, how stupid the defensive numbers are, how stupid I was for saying he wasn't that good and so on. And that game -- THAT VERY GAME -- a fly ball was hit to the gap, and Jones could not run it down. "Yeah, I'll bet you will use that to say he's not a good fielder," the guy grumbled.
YES! I will use that! Adam Jones does not get to fly balls that other center fielders reach. He just doesn't. It happened that day. It happened in the playoffs this year. It happens pretty often when I watch the Orioles play. He's a good power hitter. He will steal you a base. He will create runs. He will make nice plays in the outfield. But all in all, throughout the year, he's not a good center fielder. Giving him the Gold Glove over Mike Trout is mind-boggling. And even if you don't want to give it to a rookie, how about Denard Span (20 runs saved) or Colby Rasmus (7 runs saved) or Austin Jackson (5 runs saved)?
I love everything about Andrew McCutchen's game and since he won't win the MVP award despite an awesome season, I'm glad to see him get some hardware. That said, picking him over Michael Bourn (24 runs saved, 1st in baseball) is kind of ludicrous.
Right field
American League: Josh Reddick, 22 runs saved (1st)
National League: Jason Heyward, 20 runs saved (2nd)
Fielding Bible: Jason Heyward.
So, you go from center field -- where the Bible and Gold could not be further apart -- to right field where they are in perfect sync. It's hard to figure this game.
Pitcher
American League: Jeremy Hellickson, 2 runs saved (53rd), and Jake Peavy, 4 runs saved (22nd)
National League: Mark Buehrle, 12 runs saved (1st)
I'm not sure they should give a fielding award to a player who is forbidden from attempting to catch pop-ups. But Mark Buehrle really is a defense savant.
Dear trick or treat neighbors: If you put Bit-o-Honey in my kid's bags, I will strike down upon thee with great vengeance ...
Wednesday, October 31, 2012
