(Another) Case Against Johnny Damon

Zach’s entry yesterday on the myriad reasons why the Yankees won’t be getting what they may think they’ll be getting – that is, a productive CF/leadoff hitter – hits all the right points; Damon’s value is declining, and Yankee Stadium won’t stem that decline much, either offensively or defensively. But one thing I’ve noticed in the entire mess surrounding Damon until now is that everyone is accepting, a priori, the idea that Damon’s value would have to decline in order to make signing him a bad move; that is, were Damon to miraculously maintain his 2005 numbers throughout the length of this contract, that the Yankees would have signed a good, if not great, contract. This inherently accepts the idea that Johnny Damon is or has been a significantly above average Major League Baseball player.

What I’d like to do, then, is take a look back at Damon, rather than forward. I entered this expecting to see that Damon wasn’t quite as valuable as he was perceived to be, and at least on a general level that assumption turned out to be correct. What I did not expect was to find that Damon turns out to be slightly less valuable, offensively, than average.

Seriously.

Before I go about showing you why, I should explain that I’m going to be using a stat – and then a derivative of that stat – that not everyone may be familiar with, though it’s been around for a while. I’d like to take a second, then, and define it. Those of you who know what I’m talking about should read this too, though, as it’s important in order to understand the rest of what I’m doing here.

Secondary Average, or SecA (a quick and decent definition can be found here), was, naturally, first derived by Bill James in one of his many attempts to come up with a metric that more accurately portrayed offensive ability than the more well-known and common ones (BA, OBP, SLG). What it is, essentially, is the other half of a batter that isn’t represented in Batting Average (hence the name, Secondary Average). It takes the number of extra bases – second base in a double, second and third in a triple, second, third and home in a homer – and adds to it the total number of walks. It then adds the number of stolen bases, minus the total number of times a player was caught stealing. It then divides the whole caboodle by AB’s, in the same manner that Batting Average is derived. Many statistical studies have shown SecA to be closely related to actual numbers of runs produced (though EqA, a more complicated and balanced stat, is somewhat better at this).

I have my problems with SecA – for example, I’m not sure how useful the inclusion of SB’s really is, and I’m still on the fence about using AB’s as a divisor rather than PA’s. It does, however, seem to have excellent correlative and predictive value as a stat. My biggest gripe with SecA is that it wholly discards singles, or first base in general; I find this a little bewildering, as – though I know sabermetricians love to look down on singles – they are fairly important. Granted, a batter that hits 50 singles is less valuable that a player who hits 50 doubles, etc., but we can’t all be Barry Bonds circa 2001, no matter what we have in our respective syringes. So, I’m staking out a position here as comparatively pro-single. With that in mind, and considering the symmetry between BA and SecA, I am going to use something I’ll call FA – Full Average – which is basically (SecA+BA)/2. A simpler way of calculating this would be Total Bases + Walks + Stolen Bases – Caught Stealing, with the final number being divided by AB’s. Formulaically, if you want to try this at home because you’re a huge dork like me, it would be (TB+BB+SB-CS)/AB. I am not the first person to have made this amazing connection, nor am I even remotely close, but at the same time you never really see it around, and I think it has a fair amount of actual value.

So, we’ve talked about BA, SecA, and FA. Now we can return to the issue at hand (which is, in case you forgot, Johnny Damon).

To my mind, Johnny Damon has three roles, two of which overlap. The first is as a hitter, purely and simply; as 1/9 of any given lineup. The second has to do with his specific role, leadoff hitter. The third, which Zach covered pretty thoroughly yesterday, is his defensive role as a center fielder. I’d like to deal specifically with the first two, and tangentially with the third (from an offensive perspective).

In 2005, Johnny Damon hit .316/.366/.439, for an OPS of .805. His BA was good for 9th overall in MLB, while his OBP ranked 43rd, and his SLG 82nd (This put his OPS rank at 64th overall). The BA was his highest since his excellent 2000 season with Kansas City, while his OBP and SLG were mediocre in comparison with his career numbers. By OPS alone, he ranked 5th among qualifying CF’s (of which there were 18 this year – qualifying meaning those players that had enough AB’s as a CF to qualify for the batting title). Good, if not spectacular; the names ahead of him on that list were Ken Griffey Jr., Andruw Jones, Jim Edmonds, and Grady Sizemore (just behind him, surprisingly, was David DeJesus of Kansas City). Among qualifying leadoff hitters, his OPS ranked 4th, behind Brian Roberts, Derek Jeter, and Grady Sizemore and just ahead of Milwaukee’s Brady Clark; good company in both lists.

So why isn’t that the end of this discussion? By those lists, Damon may not be the absolute best CF or leadoff hitter in the majors, but he still appears to be quite valuable. The reason, essentially, is that OPS is a misleading statistic.

Don’t get me wrong, I like OPS. It’s easy to calculate, there’s an understood benchmark, it’s gaining wide enough acceptance that I think I’ve even heard Joe Morgan put those three letters together before. But it is imperfect, and not just mildly. The reason it’s imperfect is the same reason why OPS dramatically overvalues Johnny Damon: singles.

OPS is calculated by adding the on-base percentage and slugging percentage of any given hitter. This is actually a fairly crude metric, in that the number itself is two steps removed from physical result; it’s the combination of two stats which rate those results. OBP measures times on base, while SLG measures times that a hitter collects multiple bases with one swing. Breaking down what each one values, we see that in OBP, walks, singles, doubles, triples, and homers are all equal; each counts as a time on base. In SLG, a homer is 4/3 as valuable as a triple which is 3/2 as valuable as a double which is twice as valuable as a single, and walks are not measured. Combining those two gives us the following value ratio for OPS: Homers are 5/4 as valuable as triples (one time on base, 4 bases) which are 4/3 as valuable as doubles which are 3/2 as valuable as singles which are twice as valuable as walks. Now, a walk may not be – to use little league parlance – ‘as good as a hit’, but they’re really quite close.

According to OPS, though, a hit – or to be more precise, a single – is TWICE as good as a walk, despite the fact that the two things result, at least for the batter, in the exact same situation. That may be muddled, so let’s just provide a quick example: Player A, in his first PA of the season, lines a pitch over short for a single. Player B takes 4 balls for a walk. Player A has an OBP of 1.000 and an SLG of 1.000, for an OPS of 2.000; Player B has an OBP of 1.000 but an SLG of null, and therefore an OPS of 1.000. This replicates itself over an entire season, so that in OPS, walks are significantly undervalued by the stat, while singles are significantly overvalued in relation to other types of hits.

That’s where SecA has come in, in that it removes singles – and, indeed, all of first base – from the equation. FA, on the other hand, includes them, but weights them properly, as exactly one base. Walks are also worth one base, doubles two, triples three, etc.

So, back to Damon. Damon’s rank in SecA in 2005 was, frankly, abysmal. At .235, he ranked 107th overall in Major League Baseball; among qualifying center fielders, he ranked 11th out of 18. Among qualifying leadoff hitters, he fared microscopically better, at 10th out of 17. By SecA, Johnny Damon was not in the top 50% of qualifying MLB hitters, leadoff hitters, or center fielders. He was worse than average.

But, as I said, I have one big problem with SecA, which I at least attempted to solve with FA. So let’s look at Damon’s ranks using FA as a benchmark. Compared to other MLB hitters, Damon ranked 76th, at .276. Among center fielders, he ranked 6th out of 18, and among leadoff hitters, 7th of 17. In other words, Damon was – at least according to this metric – only barely within the top third of MLB CF’s, and squarely in the middle of the pack among leadoff hitters. Johnny Damon was not any better than ‘pretty good’ in 2005.

I am actually struggling to write that, to a certain degree, because it seems so counterintuitive. How is it that Damon could rank well among his peers in OPS, or BA, but poorly in SecA and FA? What caused the shift? In essence, singles and walks caused the shift, and I want to take a second to explain how.

I broke down Damon’s 2005 season by type of times-on-base; that is, I took the total number of times on base via a type of event – a single, a homer, a walk – and divided that by the total number of times on base. In 2005, Johnny Damon got on base via the single a stunning 58% of the time; singles were also responsible for 53.3% of his total bases. Compare that to his better 2004, and you see a sharp change: in 2004, the first number was 48%, while the second was 43.2 – a 10% change in each case. More concerning than that was the alteration in his walks; in 2004, Damon walked once per every 9.2 trips to the plate, while in 2005 it was once every 13. Johnny Damon’s 2005 was more dominated by singles than any season he had in a Boston uniform, by quite a bit.

What’s more, in comparison to the rest of baseball, these number practically explode at you. Among all MLB hitters, only 13 hitters were more reliant on singles to get on base than Johnny Damon. Only 17 were more reliant on singles for their total bases. In neither case are we talking about hitters you particularly want on your team; the list is filled with hitters like Wily Tavares, Luis Castillo, Scott Podsednik, David Eckstein, Royce Clayton… hitters that are only fawned over by people who think OBP might be a tropical disease (or an early 90’s musical reference, though they’d be a letter off).

Are you saying ‘so what’ to that revelation? Singles are, after all, a part of the game; they amounted to nearly 65% of the total number of hits in the majors last year. They clearly matter. The question is, how much do they matter, and how meaningful are they? Looking at this anecdotally for a second, how many cheap doubles have you ever seen? The occasional bloop down the line, or an OF misplaying a ball. Homers, same deal; the occasional wind-aided pop-up, or Jason Giambi at Yankee Stadium, etc. But how many cheap singles have you seen? How many grounders with eyes, or dying quails, or little nubbers no one can glove, or most infield hits, or errors improperly ruled, or…? Cheap singles are all over the place; balls that, had fielders been positioned slightly differently, or certain little things hadn’t been quite the same, would be outs. Balls that ‘have no right’ being hits. Singles are the most unpredictable of hits, because it’s that much easier to get to 1st base than it is to get to second. Let’s look again at that list of players who have a high % of their total bases come from singles, because it’s an interesting one. In fact, I’m going to post the whole list, stopping with Damon.

Willy Tavaras. Jason Kendall. Luis Castillo. Scott Podsednik. Juan Pierre. Omar Vizquel. Scott Hatteberg. Royce Clayton. David Eckstein. Placido Polanco. Chone Figgins. Angel Berroa. Sean Casey. Darin Erstad. Brady Clark. Julio Lugo. Ichiro Suzuki. Johnny Damon.

Notice anything similar about most of those names? Almost all of them are speedy players. Kendall is actually quite fast, especially for a catcher; the only really odd names on there, in fact, are Sean Casey and Scott Hatteberg. Want a stat that’ll blow your mind? Of the 148 hitters in the majors that qualified for the batting title, those 18 players – 12% of the total – accounted for 26% of the total number of steals from the full list. These players, reliant on singles, also happen to be as a group incredibly speedy; how many extra singles do you think that nets them, on the whole, per year? What then, logically, happens to these players when they start losing a step? As reliant as they are on the single, all their offensive numbers would go down proportionately more than players less reliant on speed or singles. Speedy singles hitters, it can then be postulated, do not age well. Their value is too tied up in that extra push they can get out of their legs that nets them extra singles per year.

Take all of this together, and here’s what we can see about Johnny Damon; as a hitter, his 2005 season was overwhelmingly misleading as judged only by top-level stats. So much of his value was taken up by singles – in BA, in OBP, in OPS, even in SLG – that the true predictable skills he had as a hitter were masked. Johnny Damon was a below-average Major League hitter in 2005. Below Average. He fared somewhat better among center fielders and leadoff men, but not by enough to make his loss a real crime. For $13 million dollars a year, along with the promise of lost speed and plunging production, I want way more for my money than a singles hitter who failed, catastrophically, to do much else.