I love position battles.
I love the pure competition between the players involved. It doesn’t happen much these days, but whenever there’s an honest-to-goodness competition between professional athletes for a particular role on the team, I eat it up.
I love hearing the players involved issue cliché after cliché in the process: “Well, gee wiz Mr. Interviewer, I just think So-And-So is one heck of a player and I’d be honored to share this position with him.” Or: “So-And-So is such a great guy. I’m just rooting for us both to do well.” Or maybe the classics: “I’m just going to let the chips fall where they may,” “that’s out of my hands,” and “whatever the coach/manager decides is best for the team.”
And I love the debates amongst fans. This part I love perhaps most of all, and that’s where I’ll venture today with respect to a position battle that’s been simmering for more than a year now and that might be coming to another crossroads soon. I’m talking about DJ LeMahieu, Josh Rutledge, and the only conduit either of them has to regular playing time in the forseable future: being the starting second baseman for the Colorado Rockies
This position battle is particularly interesting to me both because, while each player is of similar age, draft/prospect pedigree, contract status, and overall current ability, the way each player offers baseball value is so very different from the other. LeMahieu’s clear comparative advantage on defense is mirrored by Rutledge’s on offense. If these two players had more similar profiles, determining who should play would be easier. However, as it is, we instead need to decide between two very different sets of trade-offs.
Luckily, we do have some statistical tools are our disposal. (Of course we do. This is baseball!). In this piece, I’ll attempt to lay out the comparison/competition between the two and, in so doing, point out all the ways in which the numbers may not tell the whole story. Because they almost never do.
So Far, So Good (And Less Good)
First, how has each player done so far in the big leagues? Well, here’s the simple tale of the tape on offense (all stats through 6/17):
|LeMahieu – Career||295||101||6||73||24||14||0.277||0.317||0.361||0.677|
|Rutledge – Career||186||95||16||64||19||1||0.264||0.311||0.415||0.726|
|LeMahieu – 2014||68||33||2||19||5||5||0.257||0.325||0.327||0.652|
|Rutledge – 2014||25||13||1||8||0||1||0.350||0.418||0.533||0.951|
Chalk one up for Rutledge – especially this year – right? More power, better outcomes on the basepaths, much better average and on-base percentage lately… Well, not so fast. LeMahieu’s value comes on defense, remember? Let’s look at each player’s Wins Above Replacement, which captures defense. Each the three main providers of WAR calculations – Baseball Prospectus, Baseball Reference, and FanGraphs – uses a different formula, so I’ll reference all three of them. I’ve also normalized each one to WAR per 600 plate appearances. Converting WAR into a rate statistic accounts for the fact that LeMahieu has had much more time than Rutledge to compile bulk statistics on both offense and defense. WAR/600 PAs is an approximation of what value a player would offer over a full season at his current rate of production (Note: For those of you wise in the ways of WAR, I should point out that while not all of each player’s playing time has come at second base – LeMahieu’s played some third, Rutledge some short stop – they’ve each played the majority of thier innings at second base, especially this year.)
|LeMahieu – career||984||1.4||0.9||3.1||1.9||1.9||1.2|
|Rutledge – career||673||1.5||1.3||-0.5||-0.4||0.3||0.3|
|LeMahieu – 2014||241||0.2||0.5||0.7||1.7||0.6||1.5|
|Rutledge – 2014||68||0.6||5.3||0.5||4.4||0.4||3.5|
For this year only, Rutledge still comes out on top no matter which all-knowing baseball outfit you trust most. But the longer story is more complicated. Two out of the three calculations like LeMahieu’s overall career production better; and Baseball Reference’s metric downright hates what Rutledge has done thus far. But these are still relatively small sample sizes in the grand scheme of things, especially for Rutledge and super-duper especially for Rutledge in 2014. Besides, for purposes of this conversation, we’re more interested in what these guys will do than what they have done. Luckily, we have statistical tools for that, too…
So Now What?
There are several smart folks out there who have built models to predict player performance. Each projection system works a little differently, but they all operate under the same basic premise: by looking at what has happened with respect to a certain player (and in some cases, other players like him) – including, for young players, what they’ve done in the minor leauges (very relevent in this case) – and placing the right amount of importance on the right individual factors in the right proportion, we can predict what is most likely to happen with that player in the future. Note the emphasis on “most likely.” These aren’t guarantees. They’re simply educated guesses based on a lot of historical truth and good old math. Sometimes they’re wrong. The point is: they’re wrong less often than us inherently flawed human beings who are prone to get overly excited about an embarrassing error, a dramatically timed home run, a couple web gems, or that time that Worthless Player X struck out with the bases loaded that one time.
Here are what three of those projection systems – Baseball Prospectus’ PECOTA, along with Zips and Steamer (both available on FanGraphs) believe will happen the rest of this year. Since each projection system has made different assumptions about how many PAs each player will receive from here on out, I’ve again noted each WAR value per 600 PAs.
As you can see, each projection system expects Rutledge to perform better on a rate basis from here on out. Remember, these projections have much more to do with the player Rutledge has shown himself to be over the last several years, than they do with the unsustainable offensive force he’s been lately. The underlying factors driving these projections are exactly what we’d expect: better defense from LeMahieu, more offensive power and speed-based impact from Rutledge. The gap between them narrows from what it’s been to this point in the season – nobody, man or machine, believes Rutledge is going to keep OPSing .951 this year – but the gap remains, and it remains on a broad, unbiased basis amongst several different projection models.
We are fast approaching the point where this will cease to be merely an intellectual exercise. Nolan Arenado returns sometime in July. Given LeMahieu’s versatility and incumbency, and Rutledge’s offensive spark-plugginess, I would be surprised if either of them was optioned to AAA. However, barring additional injury, playing time will almost assuredly be significantly reduced for one or both. So how should the Rockies allocate that reduction?
Personally, I’m inclined to trust the statistical projections on this one and give most of the playing time to Rutledge. Having said that, I don’t think it’s quite the slam-dunk decision the numbers presented above make it out to be. I continue to be quite skeptical of the defensive metrics that feed into those WAR calculations. I find it eminently possible that WAR has undervalued LeMahieu’s defensive value to the team, and, on an even larger scale, that WAR undervalues and/or miscalculates defensive impact in general. I also believe there is something to the theory that there are intangible benefits to pitchers that accrue by nature of having superior defenders playing behind them.
However, I also believe some of those same intangible benefits accrue by nature of being given a lead to pitch with, and Rutledge is more likely to produce that. And given that even a cooled-off Rutledge offers so much more offensive firepower, I think he ought to be our man at second base when Arenado comes back. LeMahieu also profiles better in a utility role than does Rutledge, and could even serve as a late-game defensive replacement for Rutledge and/or prime double-switch candidate when opportunities arise.
We’re not talking about superstars or scrubs here, and the net difference between the two may be not be a lot, but margins tend to be very small in baseball in general. This is shaping up to be a very tight race for the playoffs, and despite what’s happened with the Rockies over the last month, they’re still in this thing and will likely get a lot better around the All-Star break simply by clearing out their disabled list. I want every incremental improvement I can get my hands on, and making the right call in LeMahieu v. Rutledge may just an increment that ends up mattering.