How productive is a Jarrod Dyson platoon?

When the Royals re-signed Alex Gordon, they plugged the biggest hole in their 2016 outfield.  In doing so, they likely used up most of the money that was available for the plugging of outfield holes.  Now that Alex Rios has fulfilled his contractual obligations and is unlikely to return, the Royals find themselves in need of a new starting right fielder.

While the Royals likely do not have the resources necessary to acquire another outfielder of Alex Gordon's caliber, they may have a perfectly serviceable option in-house.  Jarrod Dyson has spent the last four seasons serving as the fourth outfielder every team would love to have.  He has occasionally filled in as a starter, but Ned Yost has preferred to use him a defensive replacement and pinch running extraordinaire.  There are many lives worse than that of a major league bench player, for example mine and probably also yours, but every bench player aspires to one day crack the opening day starting lineup.  After six seasons in the majors, Dyson finally may have earned his chance.

Jarrod Dyson is fast.  It is the one thing everyone knows about Jarrod Dyson.  It is indisputable.  It is one point on which mainstream fans and stat nerds agree.  Unfortunately, much of baseball is played while just standing around, and that has never been Dyson's strength.  In 1,202 major league plate appearances, Dyson has hit for a 84 wRC+, which is far from embarrassing, but farther from inspiring.  If you're seeking that inspiration, you may find it in the fact that Dyson's platoon splits have been exploitably dramatic.

When facing right-handed pitching, Dyson has been an adequate hitter.  Against left-handed pitching, Dyson has hit like a defense-first backup catcher.  This disparity has led many Royals fans to claim that Dyson should be part of a platoon.  Combining Dyson with an adequate right-handed platoon partner would seem to be better than starting Dyson against all pitchers, and the overall production of the pair may actually be pretty good.

If the Royals choose to seek out a platoon partner, there are several players who likely could be cheaply acquired, even this late in the offseason.  Chris Denorfia has become the archetypical right-handed platoon partner, and he is seemingly always available.  Marlon Byrd is a capable hitter, and is still legally allowed to play the outfield, probably. Drew Stubbs is either a rich man's or poor man's Bubba Starling, depending on the year.  Austin Jackson may fall into the Royals' price range, if the market continues to give him the cold shoulder.  Dayton Moore may even be able to find a serviceable option right under his nose, if his 2016 payroll fibs haven't completely obscured that area. Paulo Orlando already has a Royals uniform, and while he can't really hit much (his .269 on-base percentage was the second-lowest among all outfielders in baseball), the short end of a platoon won't require him to do too much of that.

These players all seem serviceable, but is one better than the others?  Some may seem more appealing than others at first glance, but baseball has given us the opportunity to try to figure this out.  If we come up with a best guess of how we would expect each player to perform against right and left handed pitching in 2016, we can combine those numbers to estimate the total production we could expect from each platoon pair.

Process

We've already looked at Dyson's career splits, but since those numbers are baseball statistics, they're fibbing a bit.  In baseball, our sample sizes are always smaller than we would like, and cutting things in half sure doesn't help much.  Most players have pretty average platoon splits, so it takes many seasons of plate appearances before we can say with confidence that a player's platoon splits clearly indicate his talent and aren't partly a fluke.  To determine Dyson's true talent level against right handed and left handed pitchers, we would need to regress his career splits toward the average splits for a left handed hitter and then apply those regressed splits to our best guess of his 2016 performance. Gross.

We could go through this process for Dyson and the potential platoon partners, or we can check and see if the nerds have already done it for us.  Thankfully, they have.  Last year, Fangraphs debuted a splits projection tool that made Steamer splits projections available for daily fantasy betters who want to make slightly better bad decisions.  We don't know the exact process used to generate these numbers, but they seem to be about what we would expect.  The tool only provides projections as of the end of the 2015 season, but we can roughly adjust those for 2016 if we assume that platoon splits degrade at the same rate as overall stats and adjust by the same ratio.  No, this won't perfectly account for everything, but it will be pretty close.

To determine the overall batting line of a platoon, we need to estimate how many plate appearances will go to each hitter and estimate the number of those plate appearances in which the batter will have the platoon advantage.  Let's assume that left handed pitchers start against the Royals 33% of the time, the starter pitches 67% of the plate appearances in each game, and 25% of bullpen plate appearances are pitched by lefties.  These approximations don't exactly match reality, but they should be in the right ballpark.  We can use those numbers to come up with this distribution of total plate appearances.

You could argue that Yost should pinch hit to eliminate some of those plate appearances where the batter doesn't have the platoon advantage, but hitters are worse when they are pinch hitting, and the decision to pinch hit there is probably another topic for another post on another day.  Now that we have this distribution of plate appearances, we can weigh the hitting splits we found earlier by this distribution to calculate the total batting projection for these potential platoons.

Predictably, when Dyson is platooned with a right handed platoon partner, the overall production is better than what Dyson would produce alone.  Denorfia, Jackson, and Byrd project to be better offensive platoon partners than Stubbs or Orlando, but they're all in the same ballpark.  In this study of platoon partners, Mike Trout represents the upper bound of what we could reasonably expect any human to achieve, as usual.

Now that we have projected platooned batting production, we are done!  Just kidding. While Dyson's platoon partner would be better at hitting right handed pitching, he would almost certainly be a worse defender and a worse baserunner.  We need to account for that.  We will do this by combining the player's baserunning and defensive projections and making a slight adjustment for the fact that we are selecting the plate appearances in which the player is most likely to generate baserunning opportunities. This allows a projected platoon WAR to be calculated.  It is worth noting that all these WAR projections were made for the 2015 run environment, so they can be compared to each other, but should not be compared to other 2016 projections.

Results

When we factor in baserunning and defense, Dyson makes up much of the ground his bat cost him and seems to be comparable to the platoon pairs.  A platoon with Austin Jackson appears to be the preferable option, but it is also likely the most expensive.  The other projections are so close together, it is hard to say that one is preferable to the others.

This does not necessarily mean that acquiring a player to platoon with Dyson is pointless.  These projections are based on Steamer projections and Steamer may have undervalued one or more of the platoon partners or regressed Dyson's platoon splits too aggressively.  On the other hand, many believe that Steamer undervalues Dyson's baserunning and defense.  Dyson has the kind of speed you really do have to see to believe, and Steamer has never had that pleasure.  If Dyson's defense and speed are even more valuable, that would elevate the full time Dyson scenario relative to the platoon scenarios.

The fact that these numbers are almost identical is actually evidence that the Royals should platoon Dyson.  If similar total production can be generated while Dyson sits on the bench for some games, Dyson is free to generate value from the bench in those games.  In games Dyson is not starting, he is available to pinch run, and he certainly provides value in that role.  Yost also may use Dyson as a defensive replacement in some games, allowing him to reclaim some of the defensive innings this projection allocated away from him.  Some also doubt that Dyson's little body can hold up under the grind of starting every day.  There is this idea that little things tend to be brittle.  Jose Altuve and David Eckstein disagree, but that's a rather, um, small sample size.

A few months ago, this entire discussion would have seemed pointless.  Ned Yost had never given any real indication that he was open to the idea of platooning.  Writing out new lineup cards wastes paper, leaving people without a tree to stand in.  Lately, in an interview with 610 sports, he seems to have changed his tune.

Well, right now you know it's probably going to be a little bit of a platoon between Dyson and Orlando... we feel like that's enough to cover it.... Dyson has done a great job and has never really gotten an opportunity to play a whole lot, and we feel that the time is right for him and he can handle it, and Paulo Orlando did a great job last year and between the two we feel like we have the position pretty well covered.

Is the old school still on winter break?  Really, this shouldn't be too big of a shock.  These projected numbers highlight an interesting possibility, and Moore and Yost are not immune to numbers.  They both passed high school math, and they're surrounded by a team of people who passed math at colleges that have lacrosse teams.  If Yost's comments demonstrate a true change of heart, some Royals fans may finally have the starting right fielder(s) they've been clamoring for.  The Royals have always been full of surprises.  In the past those surprises have been terrifying.  Lately, they have been rather fun.