There's something that I'd really like to find out ... and that's a way to quantify the value added to a team by a player's ability to field multiple positions. I don't think any major league manager or baseball fan would argue that having players with the ability to play multiple positions is a good thing for a team. But with all the statistical and objective analysis going on in MLB right now, why can't we find out a real, measurable value for the skill of positional utility?
There's a number of steps that would go into determining the value of a player (or value added to a team) through a player's utility, but today I'd like to take a first (and not final) step in that direction. To start, we need some sort of measure that can compare players and their positional utility, and to do so, I'm creating something that I call a McEwing Score (McE). McEwing Score is named after Joe McEwing, a utility player for the Cardinals, Mets, Royals and Astros who never played less than six different positions during a full season. As a Mets fan, I'll always have a soft spot in my heart for the scrappy McEwing, who could play anywhere and provides this new
junk stat with a great acronym.
In creating a methodology for determining positional utility, one should probably start with positional adjustments -- meaning that we should have a way to determine which position has the most defensive value, and use that as a baseline for describing how useful it is to have a player that plays each position. After all, I think we can all agree that, from an objective perspective, having a player who can play first base and left field at an average level isn't quite as valuable as a player who can play shortstop and center field at an average level. We need something to use to compare differing players.
The most commonly-used positional adjustments are the ones used in FanGraphs WAR and originally developed by Tom Tango. It uses about a 25-run swing to describe the difference in fielding between first basemen (the worst / easiest defensive position) and catcher (the best / hardest defensive position). It's very good, but I actually prefer the re-configured version that Tango posts here at The Book Blog. It offers the following scores for each position, using runs above or below average:
- +15 for C
- +9 for SS
- +3 for 2B / 3B / CF
- -9 for RF / LF
- -15 for 1B
These positional adjustments are useful in their raw form, but for my purposes, I'd really just like to find a way to maintain the comparative distance between the positions, but I don't really need the actual run adjustments. For my purposes, I'm going to start with a score of .5 for 1B, and then move along accordingly through the positions according to the same scale. That would mean that the following scores apply for positions when determining McE:
- +3 for C
- +2.5 for SS
- +2 for 2B, 3B and CF
- +1 for RF and LF
- +0.5 for 1B
So there we go. If we start from zero for every major-leaguer, then apply these adjustments for every position that they play during the season, we get a McEwing score. It's a relatively simple, descriptive stat, at least in this raw form.
I will make a first, small adjustment to these modified positional adjustments. Right field and left field have a very similar defensive profile, and being able to play either or both isn't really an indicator of positional utility. Many, many outfielders play both RF and LF, and the stats show that the RF-to-LF and LF-to-RF conversion doesn't result in any major change in defensive capability. Therefore, instead of giving a player +1 to their McE for playing both RF and LF, let's instead give a player +1 to their McE for playing RF and/or LF. If during the season a player plays either RF or LF, or both, the player gets +1 to their McE.
Now I'd like to make one further adjustment to the McE scoring process: I'd like to give no points for players who appear in a single game at a given position. This may seem arbitrary, but playing an inning or a game at a position can be a matter of emergency, not a matter of utility. If a player plays two games at a given position, it's safer to say that a manager is more comfortable utilizing a player in that role, rather than having to fill in on an emergency basis. I think this makes a difference, and weeds out things like Jorge Posada's single inning at 2B, which is not an indicator of his positional utility.
The last descriptor that I would like to identify is that, when forming a McEwing Score for a given player, we should not make assumptions about a player's ability to play additional positions. For example, let's talk for a moment about Evan Longoria. In 2011, Evan Longoria played only one position in the field: third base. Longoria's a phenomenal third baseman, and probably has the talent to play a number of positions on the diamond including 1B, LF and even SS. Heck, Longoria has experience filling in at SS in the majors and played the position well in high school and community college. But we shouldn't just make the assumption that Longoria can play SS (or any other position) and assign him an McE adjustment as a result. Let's try to keep this statistic as descriptive and assumption-free as possible.
With all of these things in mind, let's look at the 2011 season, and identify certain players (and types of players), and their associated McE values. All of these scores were hand-tabulated, so feel free to let me know if I made some kind of mistake during the process.
Ladies and gentlemen, I present to you your 2011 league leaders in McE. Bonifacio, McCoy, and Martinez all posted a McE of 9.5 in 2011, which comes from playing all of the most valuable defensive positions other than catcher. Each of these guys logged time at SS, 2B, 3B, CF and one of the two corner outfield spots. While each of these players had different value over the course of the season (Bonifacio had a fWAR of 3.3, while McCoy and Martinez managed just 0.3 and -0.4 fWAR respectively), each of them proved their versatility.
There's actually a relatively large sample of players who score an McE of 8, but the two players I've listed above are two of the most recognizable. In order to pick up an McE of 8 in 2011, a player had to log time at each infield position, as well as a corner outfield spot. I was honestly a little surprised that ten players logged time at five positions (no one fit my criteria in six positions for 2011), but this proves that there are plenty of versatile players on MLB rosters these days.
Craig is the only player to log time at five positions, but not spend time at shortstop or catcher. As such, Craig has an McE of 7.5, good for the third-highest score in the majors for 2011. Craig is a very different animal than many of the others near the top of the leaderboard, as he wasn't valued for his defense or flexibility, but rather moved around in an attempt by Tony LaRussa to get his potent bat into the lineup. I'll talk more about the Cardinals and positional utility in a future post, but Craig turned out to be a rather unique case in my research.
While not every utility infielder is as good as Maicer Izturis, anyone who racks up time at 2B, 3B, and SS earns an McE of 6.5, just like Izturis did in 2011. Almost every squad has a player who can fill in at short, second and third, and plenty of these players derive their roster space almost exclusively for their utility. In fact, only two ML teams didn't have a player in 2011 who scored an McE of 6.5 or higher. Feel like guessing who they were?
Fox and Gimenez are interesting cases, as they're players who can play catcher as well as two less-demanding positions: first base and left field. As such, they rack up an McE of 4.5, which is about the same as a player who can fill in at 2B and SS. To me, that sounds about right, as a player with catcher eligibility who can play multiple positions has valuable utility. Jordan Pacheco and Tony Cruz are like Fox and Gimenez, but
on steroids whoa that was a bad choice of words better, as they logged time at multiple infield positions as well as catcher.
If an outfielder plays all three (or CF and a corner OF position), that's worth an McE of 3, or about equivalent to the McE of a guy who only plays catcher. Given how common it is to find players who can play anywhere in the outfield (most major-league CFs can do it), this seems to pass the smell test for me. But there's a big defensive difference between players like Brett Gardner and players like Vernon Wells, and each manages an equal McE, so perhaps there's still tweaking to do in order to make the stat more effective.
So there you have it. McE may not be a value statistic yet, but it is a descriptive statistic that can be used to distill a player's positional utility for a given season. In the future, I'd like to find a way to modify McE to include a player's defensive ability at each position, as well as possibly including a weighting system for the number of innings or games logged at different positions. As always, any feedback, questions or criticisms of my methodology are welcome, and I'd love for you to leave your thoughts and discussion points in the comments below.