Building a way-too-early batting order for the 2016 Tampa Bay Rays

Pitchers and catchers have reported, and yet it is still too early to know what this Rays roster will look like.

Last season the Rays carried too many players into camp, but incurred so many injuries that they needed to trade for talent. This year the hope is for the opposite to be true but the lesson remains: you can't predict baseball, let alone Spring Training.

As the Rays arrive with way too many outfielders and  multiple first basemen, the hope is for another team's need to arise and for the Rays to make lemonade out of all these damned lemons.

We do not know who will make the 25-man roster, but we can certainly speculate, and my favorite way to do this is to build ourselves a lineup... but first we need some projections.

Projections!

Steamer and ZiPS projections have been out for a while, and the PECOTA spreadsheet is presumably incoming. Here at DRaysBay, we have a fancy way of tailoring that to handedness. The numbers are generated using a tool developed by Ian Malinowski and Jason Hanselman (The Process Report) that incorporates Bojan Kopravica's splits regression research into projections.

The basic idea, for those not familiar, is that instead of just looking at career performance for batters against righties and lefties, we need to regress their past performance against what we know about left-handed and right-handed batters generally.

So for a player with only 100 or so plate appearances but a wide split, we should actually assume that his split going forward will be only slightly higher than average. For a player with a long track record of a wide split, we can go ahead and expect that trend to continue.

After taking a long hard look at the Steamer projections, our normal go-to, these weren't enough for a clear picture. For example, Steamer is bullish on James Loney and Nick Franklin, but bearish on Logan Forsythe. Who can say that is a certainty? They project a solid performance in Triple-A to carry over for Franklin, but don't imagine Loney or Forsythe to be their 2015 selves, and you could take the opposite bet within reason for each of those ideas.

Our normal go-to for wOBA projections, Steamer isn't passing the sniff test. Let's take a closer look.

Steamer

Player	Batter Hand	Steamer Proj. overall wOBA	Steamer Proj. wOBA vs. RHP	Steamer Proj. wOBA vs. LHP
Brad Miller	L	.312	.323	.282
Brandon Guyer	R	.314	.301	.325
Corey Dickerson	L	.319	.327	.288
Curt Casali	R	.289	.284	.302
Desmond Jennings	R	.306	.299	.324
Evan Longoria	R	.325	.317	.345
Hank Conger	S	.287	.291	.271
James Loney	L	.306	.316	.277
Kevin Kiermaier	L	.307	.317	.278
Logan Forsythe	R	.314	.303	.335
Logan Morrison	L	.314	.322	.292
Luke Maile	R	.261	.251	.268
Mikie Mahtook	R	.285	.273	.291
Nick Franklin	S	.300	.306	.285
Rene Rivera	R	.263	.258	.275
Richie Shaffer	R	.294	.284	.298
Steve Pearce	R	.322	.313	.337
Steven Souza	R	.318	.312	.334
Tim Beckham	R	.270	.262	.278

Here's how the above shakes out positionally:

Pos.	Player	Batter Hand	Proj. wOBA vs. RHP		Pos.	Player	Batter Hand	Proj. wOBA vs. LHP
C	Hank Conger	S	.291		C	Curt Casali	R	.302
1B	James Loney	L	.316		1B	Steve Pearce	R	.337
2B	Logan Forsythe	R	.303		2B	Logan Forsythe	R	.335
SS	Brad Miller	L	.323		SS	Nick Franklin	S	.285
3B	Evan Longoria	R	.317		3B	Evan Longoria	R	.345
LF	Corey Dickerson	L	.327		LF	Brandon Guyer	R	.325
CF	Kevin Kiermaier	L	.317		CF	Desmond Jennings	R	.324
RF	Steven Souza	R	.312		RF	Steven Souza	R	.334
DH	Logan Morrison	L	.322		DH	Richie Shaffer	R	.298

Steamer narrows our roster selection by dropping Mikie Mahtook, Tim Beckham, Rene Rivera, and Luke Maile from the roster. Life is not that easy.

Beckam and Rivera are out of options, so to remove them from line up requires removing them from the Rays, or hoping they accept a minor league assignment, which Rivera is unlikely to do. The Rays have never been one to sacrifice catching ability for offense, so it's possible that Casali will be on his way to the minors unless an injury or trade occurs.

Another problem with the above is that it includes 15 players when only 13 spots are available on the roster, but we'll get into that shortly.

For comparison, here's the same treatment of breaking down the ZiPS wOBA projections into handedness, based on career split trends.

ZiPS

Player	Batter Hand	ZiPS Proj. overall wOBA	ZiPS Proj. wOBA vs. RHP	ZiPS Proj. wOBA vs. LHP
Brad Miller	L	.315	.326	.285
Brandon Guyer	R	.317	.304	.329
Corey Dickerson	L	.333	.342	.301
Curt Casali	R	.291	.286	.304
Desmond Jennings	R	.317	.309	.336
Evan Longoria	R	.321	.313	.341
Hank Conger	S	.291	.295	.275
James Loney	L	.299	.309	.271
Kevin Kiermaier	L	.320	.330	.290
Logan Forsythe	R	.321	.309	.343
Logan Morrison	L	.310	.318	.288
Luke Maile	R	.264	.253	.271
Mikie Mahtook	R	.291	.279	.297
Nick Franklin	S	.298	.304	.283
Rene Rivera	R	.257	.252	.269
Richie Shaffer	R	.290	.280	.294
Steve Pearce	R	.324	.315	.339
Steven Souza	R	.327	.321	.343
Tim Beckham	R	.282	.273	.290

The above gives us the following rosters against each handed pitcher:

Pos	Player	Batter Hand	Proj. wOBA vs. RHP		Pos	Player	Batter Hand	Proj. wOBA vs. LHP
C	Hank Conger	S	.295		C	Curt Casali	R	.304
1B	Steve Pearce	R	.315		1B	Steve Pearce	R	.339
2B	Logan Forsythe	R	.309		2B	Logan Forsythe	R	.343
3B	Evan Longoria	R	.313		3B	Evan Longoria	R	.341
SS	Brad Miller	L	.326		SS	Tim Beckham	R	.290
LF	Corey Dickerson	L	.342		LF	Brandon Guyer	R	.329
CF	Kevin Kiermaier	L	.330		CF	Desmond Jennings	R	.336
RF	Steven Souza	R	.321		RF	Steven Souza	R	.343
DH	Logan Morrison	L	.318		DH	Corey Dickerson	L	.301

This version of the Rays roster is at least 13 players, but the big change here from Steamer is a lack of James Loney in the mix. Unless he's traded, the Rays are likely to keep his glove in the line up, but the projection seems a touch unfair.

According to our ZiPS projection manipulation, Loney's wOBA vs. RHP is lower than Logan Morrison or Steve Pearce, who are each at .309 wOBA. That mirrors Loney's 2015 production, but he was suffering under multiple injuries in his first season on the Disabled List. Why would ZiPS expect Loney to be the same as his injured self?

By contrast, Loney had a .337 wOBA vs. RHP in 2014. If the truth is somewhere in between, he belongs in the lineup.

Drawing the battle lines

The battle between Nick Franklin, and Tim Beckham will be one of the Spring Training story lines to watch. One system votes in favor of the once-heralded Nick Franklin, the other votes in favor of the out-of-options Tim Beckham. His contract status gives an upper hand, but it may be defensive flexibility that decides who remains on the roster.

Unfortunately, we cannot platoon every position. With only four spots open on the bench, at least two of the names above likewise need removed. The last two names added above were prospect Richie Shaffer and OF Desmond Jennings. Shaffer is easily optionable, but Jennings becomes the new DeJesus.

The existence of Brandon Guyer and Desmond Jennings on this roster is redundant. Between the two, Guyer has one more year of team control but a similar projection. The real lineup will slot Kevin Kiermaier into the CF role, likely on an everyday basis.

Before injuries and before the promotion of Kiermaier, it was Jennings who was the Rays starting center fielder so he is not without value, and Guyer has already demonstrated an ability to be successful off the bench (not an easy thing to do), so I would presume Jennings would be the one to be dealt if his knees are up to snuff. Still, that's another Spring Training battle to isolate.

That leaves designated hitter open against left handed pitchers, and this is likely to be a revolving door given the Rays matchup-specific propensities. I would venture to say Corey Dickerson gets the lion's share to start the season, but it should be noted Logan Morrison does have a slightly stronger projection against south paws.

So how do we choose a projection system. ZiPS is generally bullish on the Rays than Steamer, but taking a simple average to temper expectations does not justify the use of one system or another.

Comparing the systems, ZIPS has more of a recency bias, which helps the projections hone in on improvements at the plate. This creates projections for players like Logan Forsythe or Kevin Kiermaier more in line with their recent major league performance.

On the other hand, if we're going to believe that James Loney is better than his 2015 self, then we should really lean toward Steamer's suggestion. Steamer gave a higher projection to only a handful of players: Loney, Longoria, Morrison, Franklin, Rivera and Shaffer.

Steamer remembers the good years for most of these players, and the system has a strength in projecting minor league performance, so there are merits to be discussed. To make things quick, I'm going to "agree" with Steamer on all of those players, except for Longoria. If you think I've chosen wrongly, take it up with me in the comment section.

With all of that in mind, here's our blend of splits projections:

Player	Batter Handedness	Proj. wOBA vs. RHP	Proj. wOBA vs. LHP	System
Brad Miller	L	.326	.285	ZiPS
Brandon Guyer	R	.304	.329	ZiPS
Corey Dickerson	L	.342	.301	ZiPS
Curt Casali	R	.286	.304	ZiPS
Desmond Jennings	R	.309	.336	ZiPS
Evan Longoria	R	.313	.341	ZiPS
Hank Conger	S	.295	.275	ZiPS
James Loney	L	.316	.277	Steamer
Kevin Kiermaier	L	.330	.290	ZiPS
Logan Forsythe	R	.309	.343	ZiPS
Logan Morrison	L	.322	.292	Steamer
Luke Maile	R	.253	.271	ZiPS
Mikie Mahtook	R	.279	.297	ZiPS
Nick Franklin	S	.306	.285	Steamer
Rene Rivera	R	.258	.275	Steamer
Richie Shaffer	R	.284	.298	Steamer
Steve Pearce	R	.315	.339	ZiPS
Steven Souza	R	.321	.343	ZiPS
Tim Beckham	R	.273	.290	ZiPS

Building a way-too-early batting order

Among the nineteen players under consideration, only four names are in the top eight against both right and left handed pitchers: Evan Longoria, Steven Souza Jr., Steve Pearce, and Corey Dickerson.

That's too much platooning. Not only do we have to limit ourselves to a 25-man roster, but several other players will be playing every day regardless of projections. And even with that consideration, Dickerson is still likely to be bumped to get all the infielders into the lineup (unless he plays first, and Steve Pearce plays second... but we've already agreed Forsythe's projection is an unknown, so let's not get ahead of ourselves).

For the time being, lets pair the list down and cut out Rene Rivera, Luke Maile, Mikie Mahtook, Richie Shaffer. That leaves the "25th man" roster spot up for debate between Tim Beckham, who had major league success last season, and his bench competition Nick Franklin in 2015, who had it a few years prior. Only one of those players has an option, though, so we roll with Tim Beckham.

As for the batting order itself, there are several considerations.

As mentioned above, some players need to be out there every day to be themselves (Kiermaier, Forsythe), and we generally know that player preferences will be honored by the Rays, which probably slots Longoria third and Forsythe fifth. Rumor has it that Dickerson is also being considered for the fourth spot in the line up, providing power and a level of protection to Longoria.

Additionally, Topkin notes that Kiermaier will be slotted ninth on a consistent basis as the everyday center fielder, removing him from lead off.

With those disclaimers out of the way, here are your possible batting orders for 2016.

Way-too-early Batting Order vs RHP

	Position	Player	Batter Handedness	Proj. wOBA vs. RHP
1	SS	Brad Miller	L	.326
2	RF	Steven Souza	R	.321
3	3B	Evan Longoria	R	.313
4	LF	Corey Dickerson	L	.342
5	2B	Logan Forsythe	R	.309
6	DH	Logan Morrison	L	.322
7	1B	James Loney	L	.316
8	C1	Hank Conger	S	.295
9	CF	Kevin Kiermaier	L	.330

Bench: Casali (.286 vs RHP/.304 vs LHP), Pearce (.315/.339), Guyer (.304/.329), Beckham (.273/.290)

It's really difficult to place Forsythe in this lineup. Last season he held a .324 wOBA vs RHP, which slots him in ahead of Morrison in the lineup. For now I'll keep him there and you can imagine Frosty will outperform that .309, but if it wasn't working out for whatever reason, you could bump the line up and bring Pearce in to spell.

The rest falls into place pretty nicely.

Way-too-early Batting Order vs LHP

	Position	Player	Batter Handedness	Proj. wOBA vs. LHP
1	LF	Desmond Jennings	R	.336
2	RF	Steven Souza	R	.343
3	3B	Evan Longoria	R	.341
4	2B	Logan Forsythe	R	.343
5	1B	Steve Pearce	R	.339
6	DH	Brandon Guyer	R	.329
7	C2	Curt Casali	R	.304
8	SS	Tim Beckham	R	.290
9	CF	Kevin Kiermaier	L	.290

Bench: Conger (.295 vs RHP/.275 vs LHP), Dickerson (.342/.310), Loney (.316/.277), Miller (.326/.285)

It doesn't get much more right handed than that.

Here we have Steve Pearce starting as the first baseman, although he's more destined for a utility role, so I would expect Loney is still in there getting reps, but this line up serves the projection purpose.

Beckham finds his way into the line up here in a platoon at short, but unlike the Loney situation, this is close enough that Miller could keep his everyday role.

Forsythe jumps into his comfort zone with this lineup, and out of necessity, Guyer is the DH. Positions are flexible, so he could flip with Jennings based on health.

If the Rays wanted to get Corey Dickerson in the mix, he could designated hit here with Kiermaier on the bench, but as mentioned above KK is likely an everyday man.

Concluding Thoughts

Everything above is wrong. Projections are not perfect, and they have no idea what players like James Loney and Logan Forsythe will do in real life, nor do they know exactly what the potential of Nick Franklin or Mikie Mahtook or Richie Shaffer will become. Nothing written here is gospel.

And we didn't even talk about defense! This has merely been interesting look at what could be on offense.

If you made it this far, thanks for reading 2,500 words about something that will not happen. I hope you're opinions are more informed than they were before.