How Bad is the Pirates' Defense?

On April 21, the Pirates were leading the Cubs 8-6 when closer Mark Melancon entered the game to start the ninth inning. A single, double, walk, single, strikeout, and ground out later, the Cubs had a 9-8 lead that they'd hold in the last of the ninth. Melancon had a blown save. 

Melancon ran off a streak of 35 straight successful save opportunities until August 18 against Arizona. As in the Cubs game in April, the Pirates had an 8-6 lead starting the eighth. As in the Cubs game, Melancon blew the save. Unlike in April, though, it's hard to pin the blame on Melancon. Pinch hitter Ender Inciarte led off the inning with a grounder that went under the glove of third baseman Jung Ho Kang for an error. Pinch hitter Jake Lamb struck out and A.J. Pollock popped out. The game should have been over. But because of Kang's error, there were only two outs. Second baseman Aaron Hill singled, putting runners on first and second. Paul Goldschmidt hit a grounder to first, which should have been, again, the third out of the inning, but it bounced off first baseman Sean Rodriguez into right field, where Gregory Polanco had a hard time picking it up. Inciarte and Hill scored, game tied, save blown. The Pirates went on the win in 14 innings, but Melancon blew his first save after 35 straight on two unearned runs caused by two errors.

This has been an issue for the Pirates this year. They've scored runs: 4.2 per game, fifth most in the league. The pitchers' 3.15 ERA is the second-lowest. But they've given up 0.4 unearned runs per game, the second most in the league. The Pirates play smart baseball, shifting their infielders to put them in position to make outs. But you can't make outs if you can't catch the ball. 

How bad has the Pirates defense been? Let's look at it in three ways.

Traditional stats. The Pirates have made 96 errors so far this season, the most in the National League. Of them, 57 have been fielding errors, also most in the league, and they've made 39 throwing errors, tied for the sixth most. The team fielding percentage of .982 is the third worst in the league, ahead of only the Brewers and Phillies.

A lot of the focus on defense has been on first baseman Pedro Alvarez, whose 18 errors leads major league first baseman by eleven and puts him on a pace to challenge former Pirates Donn Clendenon (24 errors in 1966) and Kevin Young (23 errors in 1999) for the most errors by a National League first baseman in the past 50 years. But there have been problems all over the diamond. Pirates first basemen and right fielders lead the league in errors. Pirates third basemen are third in errors, and Pirates catchers are tied for third. Pirates center fielders are tied for sixth, and the team's second basemen are tied for eighth. The only positions at which the Pirates' error totals have been below average are shortstop (tied for tenth), left field (tied for second to last), and pitcher (second to last). So yeah, they've made a lot of errors. That's not good.

Less Traditional Stats. Defensive Efficiency Rating, or DER, measures the percentage of balls in play that are turned into outs. It's not completely precise, because it includes an estimate of total plays made, but it's pretty close. The best-fielding teams in the National League, as measured by DER, are the Giants, Mets, and Reds. Each has turned 70% or more of all balls in play into outs. The Pirates, at 68.1%, are tenth. The league average is 68.7%. So they're below-average, but not at the bottom of the list. Atlanta, Washington, Milwaukee, Colorado, and Philadelphia all rate lower. 

The company Inside Edge provides another way of looking at fielding. It looks at all balls in play and calculates the likelihood that an out will be made. Inside Edge divides batted balls into six categories: 0% likelihood of becoming an out, 1%-10%, 10%-40%, 40%-60%, 60%-90%, and 90%-100%. The Pirates have been OK in the middle of the range. They've converted 29% of batted balls in the 10%-40% range into outs, the eighth highest rate in the league. They've turned 49% of balls in the 40%-60% into outs, seventh highest. And they've converted 80% in the 60%-90% range, the fifth most. All of those figures say average to me. But on the extremes...the Pirates have been successful at the second lowest rate in the league on balls that Inside Edge calculates as 1%-10% likely and 90%-100% likely. In other words, they don't make the hardest plays, and they mess up the easy plays. 

Advanced Stats. There are three measures that estimate runs saved by defense: Total Zone Fielding Runs (from BaseballProjection.com), Defensive Runs Saved (from Baseball Information Systems), and Ultimate Zone Rating runs (from FanGraphs.com, calculated by Mitchel Lichtman). Each methodology looks at every batted ball and attempts to determine the likelihood with which the ball would be fielded. Critics decry advanced fielding stats as numerical hocus-pocus, but they're based on careful observations of all plays. The Pirates are seventh in Defensive Runs Saved, eleventh in Ultimate Zone Rating, and tied for twelfth in Total Zone. The average of those three ratings in tenth.

Conclusion. The Pirates are below average, arguably bad, but not the worst. (That title would seem to be fought between the Phillies and Padres, with the Brewers weighing in.) They won't win any Gold Gloves around the infield, though their aggressive use of shifts minimizes the damage (except at first, where Alvarez really does play like a DH). The outfield has a good left fielder, a solid if overrated* center fielder, and a talented but erratic right fielder. The value the catchers bring is more measured by strikes gained and balls avoided than anything else. This isn't a team built to have an airtight defense. But an upgrade at first, and more focus on making the easier plays, would help a lot.

*Pirates fans: Don't jump on me. I'm not saying Cutch is a bad fielder. I'm saying--and the stats will back me up on this--that he's a below average center fielder. A below average center fielder in the majors is still a great fielder. It's just that when your peer group includes the likes of Cain, Hamilton, Kiermaier, Martin, Pillar, et al, it's tough to be above average. 

April Oddities

Here, in no particular order, are my view of the five strangest things in baseball's first month. (Full disclusure: I am not trying to imitate Jason Stark's Strange but True column, if for no other reason than that Stark is way more clever than I.)

• The First Place Houston Astros. Five teams finished April with 15 victories. Two, the Cardinals and Tigers, were expected in many circles to win their divisions. Two more, the Mets and Royals, were pegged as wild card contenders. Then there's the Astros. They're viewed as up-and-coming, with a young team (weighted average batter's age 27.3, youngest in the American League) and, according to ESPN Insider Keith Law, the third-best farm system in baseball. But the team was 70-92 last year. The Astros were 51-111 in 2013. They play in the same division as the Angels, who had the best record in baseball last year; the A's, who were a wild card team and had the best run differential in the league; and the Mariners, who were only a game behind the A's and picked by many to win the division this year. Now, granted, the Astros have played the easiest schedule in their division so far, but they have the second-best run differential (1.3 more runs scored than allowed per game) in the league so far. They had the best team ERA, third-best starter ERA, and third-best reliever ERA in the league in April, after finishing 12th, ninth, and fifteenth, respectively, last season. As expected, their batters are striking out a lot (24% of plate appearances, most in the league), and their team batting average of .238 was 12th in the league. But their .316 on base percentage is ninth and their .411 slugging percentage is sixth, so they've made up for the lack of batting average with walks and power. They've probably been a little lucky, and I don't expect them to remain in title contention--this year. But there may be enough here for the team to be a 2016 contender.
• Clayton Kershaw, 1-2, 3.73 ERA. There are 60 National League pitchers who qualify for the ERA title (one inning pitched per team game played). Kershaw is 32nd. I could list some of the unlikely pitchers who rate higher, but really, anybody ahead of Kershaw's a surprise. The defending NL MVP has led the league in ERA for each of the past four years. Now, granted, he had a 4.08 ERA in May 2014, but that included arguably the worst start of his career, when he gave up seven runs in 1.2 innings against Arizona on May 17. The last time before that he'd had an ERA as high as 3.73 was April 2009. So other than the one start last May, this is his worst month in five years. Here's the catch, though: His underlying statistics suggest he was a lot better than his record indicates. His strikeout/walk ratio of 6.1 is seventh in the league and he leads the league in strikeout rate. His fielding independent pitching, which is an ERA-like measure based on strikeouts, walks, and homers (and a better predictor of future performance than ERA), is 2.85, eleventh best in the league. If you normalize that for his unusually high (and possibly bad luck-driven) 22% ratio of home runs to fly balls allowed (fourth highest in the league), it drops to 1.95, which is best in the league. So expect a return to ridiculously good form from Kershaw.
• 10-13 Washington Nationals. They were a near-unanimous pick to win their division. Several saw them as the best team in baseball and the favorite to win the World Series. So seeing them half a game behind the Marlins and Braves, and fully five games behind the Mets, at the end of April is jarring. Like Kershaw's start, the Nationals' problems don't seem likely to persist. They ended April winning three straight, scoring 34 runs. The pitching's been OK but the team was 11th in the league in both batting average and on base percentage in April. Slow starters in April included shortstop Ian Desmond (.217/.287/.326 slash line; the league average including pitchers was .248/.310/.383), left fielder Jayson Werth (.175/.254/.211), and first baseman Ryan Zimmerman (.217/.277/.380). Starter Max Scherzer had a 1-2 record despite a 1.26 ERA; in two losses and a no-decision his teammates scored a total of six runs. The return of third baseman Anthony Rendon (currently rehabbing a sprained knee) who finished fifth in MVP voting last year, should help, and I'd expect improvement as players return to form and the amazing Bryce Harper (.286/.440/.545, still the second-youngest player in the league) continues his strong performance.
• Mike Moustakas, Batsman. I'm don't get too excited about a month's worth of batting statistics. Anybody can get lucky (or unlucky) for a month. Light-hitting Marlins second baseman Dee Gordon and Tigers shortstop Jose Iglesias batting .423 and .370, respectively? Not going to last. Neither will some of the poor performances (see next point). But Moustakas may be different. The third baseman was one of several highly-regarded Royals prospects a few years ago. In four years in the majors, he's been a disappointment. His batting averaged declined every year from 2011 to 2014 (falling from .263 to .212), as did his on base percentage (from .309 to .271). His OPS last year, .632, was the lowest of his career, third lowest in the league among players with 500 or more plate appearances, and 25% below the league average, adjusted for park. Last month, though, he was tenth in the league, at .942, 67% better than average. The reason it doesn't appear to be luck is that Moustakas has overhauled his approach to hitting--not easy to do at age 26. He's been an extreme pull hitter. Batting left, he only 20%-30% of his base hits went to the left side of the field in 2012, 2013, and 2014. As a result, he faced an extreme shift, with three infielders to the right of second base, in 71% of his plate appearances last year. He was easy to predict, easy to defense. But this year, he's got 32 base hits. Of them, thirteen singles, four doubles, and a homer--over 56% of his base hits--have gone to the opposite field. That's a real change, and the five extra-base hits indicate that he's not just popping bloop singles out to left. By using the whole field, he's become a different, and, finally, good hitter.
• Andrew McCutchen, Sub-.200 Hitter. I'm not going to get into this a lot, because I've written about it enough: My concerns about his health here, my analysis of what's wrong with him (mostly, bad luck) here. During April, among 86 National League batting title qualifiers, McCutchen's .194 batting average with 80th, his .302 on base percentage 59th, his .333 slugging percentage 73rd, and his .636 OPS 68th in the league. He had the worst batting average on the Pirates, and the second-worst slugging percentage. Now, hold that for a minute, because I want to talk about fielding. Yesterday, I heard two SiriusXM MLB Home Plate hosts go on and on about how one of the advanced fielding metrics rates Mets center fielder Juan Lagares as slightly below average so far this year, and how this proves that you can't trust some of these new statistics. Look, nobody's going to argue with the idea that Lagares is probably the best center fielder in baseball today. He won the Fielding Bible Award last year and was runner-up the year before. Without getting into the mechanics of how the advanced fielding metrics are calculated (I think they're sound), let's look at what the numbers for Lagares are saying: That he didn't have a fabulous April with the glove. Is that true? I don't know. I didn't watch every play he made, and neither did the folks on the radio. The people who calculate these advanced metrics did. But that's not the point. The point is that the stats are saying Lagares didn't have a great April in the field. So what? Getting back to McCutchen, the stats say that McCutchen didn't have a great April with the bat. Is anybody going to deny that? If a radio host said, "Andrew McCutchen hit .194 in April, therefore he's a bad hitter," we'd call him an idiot. We would also call him an idiot if he said "Andrew McCutchen hit .194 in April, so batting average says he's a bad hitter, which we know if false, so we can't trust batting average." No, no, no. But that's what the guys on satellite radio were saying: if the stats don't say that a great fielder is great every month, the stats are wrong. No, the stats are right, just like McCutchen's slugging percentage is right. But what they measure isn't a player's overall value, it's how they did over a 30 day period. Lagares's defensive numbers in April should be viewed the same way as McCutchen's batting numbers in April: An aberration, unusual but well within normal variation, for an excellent ballplayer. I feel pretty confident in saying that, by season's end, Juan Lagares will be rated as one of the top fielders in baseball, just as I feel pretty confident in saying that, by season's end, Andrew McCutchen will be rated as one of the top hitters in baseball. The fact that the numbers didn't reflect that in April about Lagares doesn't indict the fielding stats any more than McCutchen's .194 indicts batting average.