Orioles 2018 Blueprint Follow-Up: Why I Decided to Non-Tender Zach Britton

Zach Britton (photo via Keith Allison)

A couple of weeks ago, some of the writers here at Camden Depot went through our annual exercise of outlining our vision of how the Orioles should approach the offseason. My blueprint was the first to be published, and within it, I made the somewhat controversial decision to non-tender Baltimore closer Zach Britton. By doing so, the move saved me $12.2 million in payroll, but significantly weakened my proposed bullpen. For our fictional exercise, I thought the move was necessary to make the club as competitive as it could be for the 2018. And instead of getting into all of the reasons as to why I thought it was best to non-tender Britton within that piece (essentially inserting 1,000+ word tangent), I promised that I would post a follow up article detailing my reasons for doing so. Well, here it is.

With one more season of Manny Machado and Adam Jones, I wanted to put the best team possible on the field to compete in 2018, without much consideration as to what the team would look like in the following season. To do accomplish that, the two main things I was focused on in my blueprint was strengthening the bench with players who could play multiple positions and remaking the starting rotation. A more complete bench was the easy task, and in my view that goal was accomplished by signing Jon Jay, Howie Kendrick, and Cliff Pennington.

There is no getting around the fact that the 2017 Baltimore starting rotation was terrible. The only good thing about them was that the 3 worst performers were all going to be free agents.

2017 Baltimore Orioles SP Ranks

Like many of the other writers, I believed the lack of quality starting pitchers in the minor league system meant that it was essential to sign at least 3 starting pitchers. I believed that the only way to significantly improve the rotation was to get the best starting pitcher available, and in my mind that was Yu Darvish. With Darvish costing $17.3 million per year (and considering all the other areas this team needed to improve as well as a limited budget), my decision essentially boiled down to whether I would want Britton or Darvish. I chose Darvish.

The main reason I felt why I ad to go with Darvish was due to just how BAD the Orioles starting rotation was in 2017/could be in 2018. The Orioles have proven that they don’t need a great rotation to be successful and make the playoffs. However, the 2017 rotation was so bad (combined with the lack of internal options), I believed that anything less than Darvish would be applying a band-aid to a wound that required 40 stitches. And while he’s looked at as an injury risk, he’s coming off a sub-par season, where he was still worth 3.5 fWAR (3.9 bWAR). He’s averaged 4.1 fWAR (4.2 bWAR) per 180 IP over the course of his career. To put that into perspective, the Orioles haven’t had a starting pitcher eclipse the 3.0 fWAR mark since Erik Bedard did it in 2006 (4.6) and 2007 (5.0).

Comparing average Yu Darvish against Zach Britton at his absolute best (2.5 fWAR during the 2016 season), the Orioles get an additional 1.6 wins for the added cost of $5 million. I realize that Darvish would come with an additional 3 years of commitment according to our BORAS projection system, but as I mentioned, I was only concerned with 2018. Granted that average Darvish versus the best Britton argument doesn’t work when using bWAR bWAR (which based on run prevention instead of FIP), but what is the likelihood we see 2016 Britton again in 2018? It’s not impossible, but coming off an injury-plagued season with good (but not great) numbers, I think it’s highly unlikely.

Britton missed most of May, all of June, and the beginning of July due to a strained left forearm. He was then shut down for the rest of the season after his September 18th appearance due to an MCL sprain in his left knee, an injury that was serious enough for him to receive a stem cell injection. When he did pitch, he was effective, but certainly not his usual self, finishing with career worst strikeout and walk rates (since being transitioned to the bullpen full time).

Obviously I don’t know exactly how bad either of those injuries are, but that forearm injury (combined with his performance) was a big reason why teams weren’t offering a trade package the Orioles liked at the deadline. Along with the knee injury, he didn’t do anything the last two months of the season to bring his trade value back up. When you combine the injuries, the performance, and the assumption that he’ll make $12.2 million in his 4th and final year of arbitration, I don’t believe that he has much trade value. This kind of torpedoes the idea of tendering him a contract and trading him later in the offseason. If the Orioles couldn’t find an attractive offer at the 2017 trade deadline, they’re certainly not going to find a better one this offseason. Furthermore, I don’t think his value will be that high at the 2018 trade deadline if the Orioles are out of it. The only way the Orioles get a good package for him is if he pitches like he did in 2016 again and there’s a desperate contender who needs a closer. Again, that scenario is possible, just not probable.

The need to greatly improve the starting rotation, combined with the injury questions and decreased trade value of Zach Britton, led to my decision to non-tender him. I ultimately thought the $12.2 million saved could be better spent to make the team better in 2018. So if I were actually sitting in the GM chair in real life, would I still take the same course of action? Probably not (actually, almost definitely not).

For the purposes of our blueprint exercise, I stand by my decision. But I don’t think it’s a smart one to make in real life. And that’s not because the fans would probably be pissed about non-tendering a favorite (although that should be given consideration as well). The real reason you don’t non-tender Zach Britton is because it’s not guaranteed that you’ll be able to reinvest those savings into players that will improve the team. The readers who commented on my blueprint started to hit on this topic a bit. Yu Darvish, Alex Cobb, and all of the other players we proposed to sign are human beings. They have a choice of where they want to play. If they don’t want to play in Baltimore, they’re not going to sign with Baltimore. The blueprint exercise takes that choice away from them. Non-tendering Britton in real life could very possibly yield a team without Zach Britton and an extra $12.2 million. That may be ok for the owners of the team, but from the perspective of being competitive, the Orioles would be decidedly worse off.

A Tale of Two Chatwoods

Tyler Chatwood is an interesting player and one that has been contemplated by other sites, including here and Camden Chat.  However, I want to tease out some things.  First, I want to visit BORAS (pi) 1.3, our site's most recent version of contract prediction for free agent pitchers.  It pegged a new Chatwood deal as being in the neighborhood of 4/41.9.  This is a much higher value than what was put forth recently by MLB Trade Rumors: 3/20.  At that price, I think Chatwood would be a no brainer because of how interesting his performance has been these past two years at Coors (6.07 ERA, 148.1 IP) and away from Coors (2.57 ERA, 157.1 IP).

This presents a very drastic difference in production and almost as he was indeed two different pitchers (more on that later).  How would all three of these pitchers compare under the BORAS model if they spent a whole year at home or away?

4 years, 41.9 MM (actual)
Minor League Deal (home projected)
6 years, 146.6 MM (road projected)

It is rather a stunning difference and when we see something like this, we want to ask why?

	4S	2S	CH	Sl	CU
Home	40%	23%	4%	23%	10%
Road	28%	35%	5%	19%	13%

You can see his approach changes when he goes out on the road.  On the road, Chatwood relies more on his pitches with movement.  He leans on a sinker and curve a bit more and moves away from the four seamer and slider.  Why?  Because, depending on where he is, he has to use what works best.  At Coors, what works best is his fastball.  On the road, the fastball still works, but he gets more movement out of his other offerings.  He turns to his sinker, which sinks more. He goes with a slower breaking ball because it breaks more.

Horizontal Movement

	Home	Road	Diff
4S	-1.8	-1.5	0.3
2S	-5.7	-6.9	1.2
CH	-2.9	-4.7	1.8
SL	0.6	1.5	0.9
CU	3.0	4.3	1.3

Vertical Movement

	Home	Road	Diff
4S	8.7	9.6	1.1
2S	7.7	7.6	0.1
CH	5.2	4.4	0.8
SL	3.1	2.6	0.5
CU	-5	-9.2	4.2

Although Chatwood has been termed a sinker ball pitcher outside of Coors Field, that is a bit of a misreading of the numbers.  While his flat fastball gets flatter away from his current him and his sinker runs a bit more, what really changes is the run and depth of his breaking ball.  His breaking ball goes from being a show me pitch in Coors to something with pretty damaging potential away.

It would be interesting to see how Chatwood is able to adapt with having a fairly consistent collection of pitches instead of having to deal with the differences Coors offers on a regular basis.  Can he be that 24 MM arm?  Does a more stabilized pitch set also means a more readily prepared batter?  What he has accomplished on the road has been remarkable these past two years and there is good reason to believe that his pitches support that potential.  While much attention has been handed to Alex Cobb, Chatwood might well be the real star here.

What Is a Jimmy Yacabonis?

Joe Reisel's Archives

Jimmy Yacabonis was named to the International League team for the 2017 AAA All-Star game. Photo courtesy of Les Treagus / Norfolk Tides.

Recently, the authors of - and guest contributors to - Camden Depot have provided their opinions on what the Orioles should do this offseason. Every contributor acknowledged that the Orioles need pitching. While the need for starting pitchers is huge and obvious, many commentators (not necessarily on Camden Depot) have speculated that the Orioles could try to meet that need by trading from their bullpen. It is noteworthy to me that none of the commenters have suggested using pitchers from their AAA Norfolk team on the 2018 major-league team, except that Alec Asher and Gabriel Ynoa might be desperation options as long relievers/fifth starters if nothing else works.

Although it's noteworthy, it's really not surprising. The 2017 Norfolk Tides were not a good AAA team and their run prevention was also not good, finishing ninth in the fourteen-team International League despite playing in a pitcher's park. And many of the Norfolk pitchers were expressly acquired by the Orioles to serve as depth options, ready to be promoted if the Orioles needed a long man in the bullpen or if they were hoping to catch lightning in a bottle. While the Orioles did catch a bolt labeled "Richard Bleier" in a bottle, pitchers such as Damien Magnifico, Jordan Kipper, Paul Fry, and Andrew Faulkner were only of interest in our end-of-season game of "Name Those Tides."

But I do wonder if the in-organization options for the Orioles staff are being written off too quickly. There were four Tides pitchers who were considered at least moderate prospects before 2017 and/or pitched reasonably well for the 2017 Tides. It's at least conceivable that - well, I don't know if the fans I'm thinking of would be considered casual or sophisticated - some fans might wonder if relief pitchers Jimmy Yacabonis and Stefan Crichton, or starting pitchers Chris Lee and Jayson Aquino, could play for the 2017 Tides. Over the next couple of months, I'll be looking at these four pitchers based on my observations/data collection from my 2017 work. (Newcomers - I work for the Tides as a Major League Baseball Advanced Media (MLBAM) datacaster and for Baseball Information Solutions (BIS) as a minor league scorer.) First up - Jimmy Yacabonis.

Jimmy Yacabonis was the Orioles' 13th-round draft pick in 2013 out of St. Joseph's University. Although he had control problems, he pitched effectively while rising through the Orioles system - except in the Advanced-A Carolina League (Frederick); he spent parts of three seasons there and so only got to AAA in 2017, at age 25. His 2017 was, at least at the most elementary level, quite successful. With Norfolk, Yacabonis pitched 61 1/3 innings in 41 appearances, with a 1.32 ERA (and no unearned runs); a 0.95 WHIP; and no home runs allowed. In a couple of stints with the Orioles, Yacabonis pitched 20 2/3 innings in 14 appearances, with a 4.35 ERA (again with no unearned runs) and only two home runs allowed. Looking at those stats, he looks like a pitcher who could help in a bullpen role, and has even shown the ability to pitch more than one inning in an appearance.

Of course, those are his most favorable statistics, and other aspects of his performance raise doubts. With Norfolk, Yacabonis walked 28 batters and struck out only 48, unimpressive totals for a relief pitcher. He was even worse in his time with the Orioles, walking 14 while striking out only 8. His WHIP in his 20 2/3 innings with Baltimore was 1.55. It is possible - maybe even probable - that his overall effectiveness was a product of luck, rather than ability.

I saw 14 of Yacabonis' Norfolk appearances, and will look more closely at his performance to see a little more clearly what he is. First, his basic performance data:

Date	Opp	BFP	IP	H	R	ER	W	K
Apr 10	GWI	6	2	0	0	0	0	2
Apr 15	CHA	4	1	0	0	0	1	0
Apr 26 (1)	BUF	3	1	0	0	0	0	0
May 06	DUR	5	1	2	0	0	0	2
May 07	DUR	6	1 1/3	1	0	0	1	0
May 20	CHA	11	2 2/3	0	1	1	3	3
May 31	PAW	4	1	1	0	0	1	0
Jun 03	ROC	8	1 1/3	1	0	0	3	1
Jun 18	LOU	7	1 1/3	2	1	1	2	0
Jun 28	LOU	7	2 1/3	0	0	0	2	0
Jul 27	SWB	7	2	0	0	0	1	3
Aug 19	TOL	6	2	0	0	0	0	3
Aug 22	CHA	3	1	0	0	0	0	0

That's a pretty solid line with a couple of red flags. In 20 innings, he allowed 2 runs (0.90 RA) on only 7 base hits. He pitched at least one complete inning in all 14 outings and at least two in 6. On the negative side, he walked 12 in those 20 innings and struck out only 14. This does appear to be a representative sample of his work, as his rates are completely in line with his full-season rates.

Next, a look at the results of Yacabonis' pitches:

Ball	135
Callled Strike	63
Swinging Strike	20
Foul	48
In Play	54

Although the Ball total may appear high, it's really in line with the counts of most Norfolk pitchers (at least the ones I looked at previously.) What's more distinctive is that batters swung at fewer of Yacabonis' pitches than they did at the pitches of most other pitchers - it's not radically out of line, but the percentage of swings is at the low end. I see two possible reasons for this - either it's known that Yacabonis doesn't have good control, so batters don't swing; or Yacabonis has a deceptive delivery and batters don't see the ball well. (Or both.)

The most interesting data is below, which is the number of pitches Yacabonis threw at different ball-strike counts:

0-0	81
1-0	43
0-1	31
2-0	16
1-1	33
0-2	18
3-0	5
2-1	19
1-2	26
3-1	9
2-2	27
3-2	12

The interesting thing is that over half the time Yacabonis pitched a ball on the first pitch. It's been established that pitchers are more effective when they have an 0-1 count, as opposed to a 1-0 count, after the first pitch. It does seem that, if he does fall behind 1-0, he doesn't have much trouble throwing the second pitch for a strike - he only had 16 2-0 counts.

So, after all this, I still don't know what to make of Jimmy Yacabonis. That he was effective in AAA is undoubted. It's still not clear whether he'll be able to be effective as a major-league pitcher. In general, I believe that a pitcher who's been effective at a level deserves a real chance at the next level, and Yacabonis fits that criteria. But his lack of control, his lack of strikeouts, and his inability to throw strike one give me pause. I can understand why the Orioles, in 2018, would rather use a pitcher with better predictors of success even if that pitcher was less effective than Jimmy Yacabonis.

The Orioles Need Better Outfield Defense

The Orioles outfield defense was considered poor by defensive metrics. Per Fangraphs, the Orioles outfield had a UZR of -21.1 runs, ranked 29th out of 30 teams. I’ve been known to be skeptical of UZR in the past, but this year UZR has just three teams with an outfield defense worse than 10 runs (1 win) and six with an outfield better than 10 runs. This is a significantly different from other years, suggesting a drastic change in methodology. Statcast’s outs above average metric tells a similar story as it ranked the Orioles outfield defense as worth -15 outs above average, tied for second worst in the majors. According to its sprint speed metric, the Orioles outfield was very slow and this helps explain its ineffectiveness.

Adam Jones has been the foundation of the outfield defense for the past ten years. He’s been fast enough to play center field while having above average offensive production. But good things can’t last forever and Jones has finally faded defensively. Statcast ranked Adam Jones as being worth -7 outs above average, 201st out of 210 outfielders. The reason why he was so ineffective was because he’s now slow. Per Statcast, his sprint speed was only 27.1 feet per second ranking 53rd out of 58 center fielders. It isn’t clear whether his defensive positioning had a detrimental impact on his speed, but this suggests that he needs to be a corner outfielder going forward. This could be problematic because his bat isn’t good enough to make him any better than an average right fielder offensively. Depending on his 2018 performance, the Orioles probably shouldn’t offer Jones a qualifying offer, nor should they extend him unless he signs a short-term deal at a significant discount from his current contract. There’s only so much a team should pay for leadership.

Fangraphs ranked Trey Mancini’s outfield defense as being worth -7.9 runs per 150 innings while Statcast valued him at -5 outs above average. Trey Mancini ranked 39th out of 58 left fielders with a sprint speed of 26.9 feet per second. That’s probably fast enough to play left field for the next few years, but it does mean he’ll need defensive help. With Trumbo and Davis on the squad for the next two years, Mancini will be forced to play left field barring an injury or a platoon.

Fangraphs ranked Trumbo’s outfield defense as being worse than -10 runs per 150 innings, while Statcast valued his defense as -5 outs above average. Mark Trumbo was graded as a DH, but had a poor sprint speed of 26.3 feet per second. Trumbo clearly has no business playing in the field, but his inability to hit as a DH suggests that he needs to play in the outfield. A center fielder with good speed and range can make up for Trumbo’s inability to run, but that center fielder isn’t Adam Jones at this point in his career.

Fangraphs ranked Smith’s outfield defense as being worth -5 runs per 150 innings and he ranked 49th out of 51 right fielders with a sprint speed of 26.4 feet per second. Seth Smith is a free agent this offseason and will likely not be a factor in the Orioles’ 2018 plans. However, his limited speed certainly contributed to the Orioles outfield defensive woes.

There is limited help on the roster. Chris Davis, with a sprint speed of 25 feet per second, was the 45th slowest out of 48 first basemen and now is about as fast as Matt Wieters. Davis has played in the outfield before, but can’t do so in the future if he is so slow. It would make sense to use him as a DH at this point. Austin Hays has had limited playing time as a major league outfielder, but had poor defensive results according to both Fangraphs and Baseball Reference as a right fielder. Statcast doesn’t have enough information to have a useful opinion of his speed/defense. He appears to be a defensive upgrade over Mancini/Trumbo but still below average.

The sole bright spots were Rickard and Gentry. Joey Rickard ranked 28th out of 51 right fielders with a sprint speed of 27.5 feet per second and Craig Gentry ranked 5th out of 51 right fielders with a sprint speed of 28.5 feet per second, suggesting that Gentry would be a valuable pickup on a minor league contract. However, Rickard’s .241/.276/.345 line suggests that he’s most valuable as a defensive replacement/pinch runner. In addition, he’s merely bad against left-handed pitching, suggesting that he could even be a platoon outfielder in a pinch. 

Camden Yards isn’t a large outfield, but even the Orioles need some outfield speed in order to play adequate defense and it is unlikely that anyone on their roster can provide that speed with acceptable offense. This will have to be a factor when the Orioles look at free agent outfielders if they want to improve.

The Orioles have been linked to Carlos Gonzalez, and MLBTR predicts that the Orioles will sign him for 1 year and $12 million. This would be a poor decision as Gonzalez’s sprint speed ranks 45th out of 51 right fielders at 26.6 feet per second. Not only is Carlos Gonzalez extremely slow, but he was ineffective last year against left-handed pitching with an .206/.241/.321 line with a 3.6% walk rate and a 30.7% strikeout rate. He was decent against left-handed pitching in 2016, but was terrible against it in 2015. With his limited speed, I’d expect his BABIP to plummet in future years. At this point in his career, he’s a defensively challenged outfielder that can only play against right-handed pitching. This has value to some teams, but the Orioles should probably pass given that they already have speed challenged players like Mancini, Davis and Trumbo on the roster and are lacking a fast center fielder to make up for their inability to cover ground.

Curtis Granderson is another player linked to the Orioles, but his sprint speed ranks 57th out of 58 center fielders at 26.6 feet per second. In addition, his low BABIP of .228 in 2017 and .256 in 2016 suggests that his speed is perhaps slower than it seems or that he’s vulnerable to shifting. In any event, he’s not fast enough to help solidify the Orioles’ defense playing right field.

A number of other free agent outfielders have the same issue. Jay Bruce might be an interesting addition, but his sprint speed is 47th out of 51 right fielders at 26.5 feet per second. Jonathan Jay is an interesting option, but his sprint speed ranks 56th out of 58 center fielders at 26.6 feet per second. Cameron Maybin ranks 43rd out of 58 center fielders with a sprint speed of 27.8 feet per second. Austin Jackson ranks 46th out of 58 center fielders with a sprint speed of 27.6 feet per second.

Maybin and Jackson are potentially decent corner outfield options with the ability to help the Orioles defense improve its range slightly. Neither of the two are considered top 50 free agents by MLBTR, while BORAS thinks they’ll each receive roughly 2 years and $20 million. 

Carlos Gomez ranks a passable 36th out of 58 center fielders with a sprint speed of 28.1 feet per second, but that’s a significant drop from his 2016 speed of 28.7 feet per second and his 2015 speed of 28.5 feet per second. At 33, one needs to wonder how long his speed can hold up. MLBTR projects him to earn 2 years and $22 million while BORAS has him at 3 years and $31 million. Gomez has been better against right-handed pitching than left-handed pitching over the past three years and could be platooned with Rickard. He’s a definitely reasonable free agent option for the Orioles.

At 32, Lorenzo Cain is one of the fastest runners in the majors with a sprint speed of 29.1 feet per second and would almost definitely help the Orioles outfield defense. He’s been above average offensively in the past and would be a strong leadoff option. MLBTR and BORAS are nearly in agreement about what Cain will earn, roughly $65-70 million over 4 years and will likely receive a QO. The Orioles will need to decide whether they can afford to sign him and whether it’s worth giving up the draft pick necessary. They’ll also have to see how long they think Cain can retain his elite speed.

Jarrod Dyson is another fast runner with a sprint speed of 28.8 feet per second. Jon recommended signing him in his final blueprint, and I think that he’d make sense. BORAS projects Dyson to earn a prohibitive 3 years and $32.5 million, but MLBTR has him receiving 2 years and $12 million. Dyson is old, but is above average defensively in center field and would allow the Orioles to use Jones in right field. Dyson historically struggles against left handed pitching, making him a viable platoon option with Joey Rickard. The Orioles almost certainly won’t offer 3 and $30 million to Dyson, but offering 2 years and $15 million would be an offer I could see them make. Dyson should be the Orioles first choice given their payroll situation and the fact that they need three new starting pitchers even if Cain is the better player.

The Orioles outfield defense was poor last year because their outfielders were slow. Their outfielders are only getting older and slower and so their defense will only continue to degrade without adding talent from outside the organization. Signing a player like Jarrod Dyson, Carlos Gomez and Lorenzo Cain would be a definite upgrade to the Orioles’ outfield defense and help them compete in 2018.

Would You Be Happy With Carlos Gonzalez And Jason Vargas?

The baseball season is over, and fans are desperate for things to talk about. That leads us to one of the an annual exercises of MLB Trade Rumors, when Tim Dierkes and/or a combination of writers predict free agent landing spots. As you'd imagine, it's an incredibly difficult job to correctly forecast which free agents will land where, and more than anything, it's just for fun.

The 2017/2018 predictions were published yesterday. For the O's, they include two signings: Carlos Gonzalez (1 year, $12M) and Jason Vargas (1 year, $10M).

If you read any part of our blueprint series, or are just vaguely familiar with the current makeup of the O's roster, you know they desperately need starting pitching. An offseason in which the main starting pitching signing is Vargas would, to put it mildly, not inspire confidence!

Again, though, MLB Trade Rumors' post just features a bunch of guesses. Some of them are right; most of them are wrong. If you tried the exercise yourself, you'd probably do just as well, or close. Just as an example, I looked back at the last five years of predictions for the Orioles. Here they are (with top 50 rank in parentheses):

2013: Nick Swisher (6), Joe Saunders (25), Jeff Keppinger (41)
2014: Scott Feldman (27), Nate McLouth (41)
2015: Nelson Cruz (9), Nick Markakis (24), Jung-ho Kang (50)
2016: Scott Kazmir (18), John Jaso (45), Justin Morneau (48)
2017: Mark Trumbo (8), Michael Saunders (20)

Of the 13 predictions over the last five offseasons, the Orioles signed one player: Trumbo.

Dan Duquette has said the Orioles intend to reload this offseason. Signing Carlos Gonzalez and Jason Vargas would not qualify as reloading. The O's don't like to spend big in free agency, especially on starting pitching. So what are they going to do? Make a few trades? Sign a lefty like Vargas as the main rotation upgrade? Grab a couple of starters off the scrap heap? Let it ride with sort-of interesting but unproven starting options like Miguel Castro and Gabriel Ynoa?

If the Orioles are actually going to reload, they'll have to do a lot more than those two signings. But then again, the O's version of reloading a starting rotation or roster is much different than other teams'.

Sensible, yet Expensive: A Davis/Trumbo First Base Platoon

One of the major points I made in the 2017/18 Orioles Blueprint finale was that the club should begin the season with a first base platoon of Chris Davis and Mark Trumbo.  This would be a major shift in what was intended when either of these players were signed to their contracts.  Chris Davis still has five years left at 17 MM a season with 6 MM on top of that deferred every year.  From a management perspective, it is difficult to come to the conclusion that your big ticket item, your monetary face of the organization after just two years on his current deal needs to ride the pine against same sided pitching.

When Davis signed, there certainly was some fear here.  Davis has two major things working against him.  One, his hit tool is on the threshold of being unplayable.  His immense power compensates when he is able to put the ball in play, but that depends on him making enough contact and getting power in that contact.  In 2013 and 2015, he made contact and got on base about 15-20% more often than the past two seasons.  It is hard to be productive when you are not getting on base.  Second, his immense .300-some ISO has dropped to a solid .200 ISO, but that still makes it difficult to make up for his lack of contact.

The second aspect that was concerning was that even in his great years, Davis was still rather adequate against left handed pitching.  As the rest of his game has fallen, his performance against left handers has cratered.  It is so poor that it is hard to think that Davis has gotten 365 PA these past two seasons against left handers.  Buck and the Orioles have shown zero interest in protecting him from pitchers that he is unable to handle.

Chris Davis Platoon Splits

Split	PA	BA	OBP	SLG
vs RHP ('16)	464	.223	.341	.487
vs LHP ('16)	201	.216	.313	.398
vs RHP ('17)	360	.218	.317	.468
vs LHP ('17)	164	.208	.293	.326

Provided by Baseball-Reference.com: View Original Table
Generated 11/1/2017.

In 2017, Chris Davis had a .118 ISO against left handed pitching.  That is bad, but really how bad is it?  Off the 25 first basemen who had more than 125 PA against left handed pitching, Chris Davis second worst.  What is the median value? .205.  Chris Davis' power performance (which is what his value is) is 50% of what the average first baseman does against lefthanded pitching.  Now, let us only consider left handed batters who play first base.  There were 11.  What was that median? .186.  Baltimore has a problem.

If the club can stomach Davis as a first baseman against right handed pitching, then the question comes to who plays first base when Davis sits.  As I noted, Mark Trumbo fits that bill and he is also a problem for the Orioles.  The club signed Trumbo after he found great difficulty in getting decent offers as a free agent.  The Orioles placed a qualifying offer on him and he refused it, thinking he had enough leverage to secure a deal elsewhere.  Instead, he came back to the Orioles at what honestly was a bargain price though the Orioles already had players cluttering up his best defensive positions: 1B and DH.

As I have often mentioned here and on Twitter, Mark Trumbo suffers from playing designated hitter.  Some scoff at the idea that a player could be impacted based on whether or not he plays in the field.  However, this has been shown to be true as a population effect.  The next statement I receive is that 14 points of wOBA are not that much.  Yes, that is correct.  It is not much of a difference even though it is a significant population effect.  One needs to go to the next step though.  The population is full of players who are found to be worse than or better than that average.  No one is that average.

On that spectrum, Mark Trumbo finds himself on the very bad side during his Oriole career:

Mark Trumbo 2016 and 2017
In field 280/338/543, 263 ISO, 133 wRC+
at DH 220/278/413, 193 ISO, 81 wRC+

And even worse last year:

Mark Trumbo 2017
In field 331/366/504, 173 ISO, 132 wRC+
at DH 207/268/368, 161 ISO, 66 wRC+

It is plainly obvious that Trumbo simply does better when he gets to use his glove.  Why?  I do not know.  There certainly are ideas out there to explain it.  For instance, when is the least safe time when you are in a plane?  When the pilot disengages from autopilot.  Having spent up to several hours not flying a plane, that moment where the pilot is suddenly engaged winds up causing significantly more flight errors than at any other point and even when correcting for pilots not using autopilot.  That would seem to suggest that maybe why some batters do worse when riding the pine is that their mind is not engaged.  That playing the field may prime a player for when he steps into the batters box even though that activity is far different from what one does in the field.

Regardless of why, Trumbo has repeatedly struggled as a designated hitter for a long while.  This leaves the club starting off with a lineup against right handed pitchers with Chris Davis at first base and Trey Mancini at designated hitter.  If pinched, Mancini moves to left field, Davis to the designated hitter slot, and Trumbo goes to first base.  Against southpaws, Trumbo starts at first and Mancini once again is the designated hitter.  If pressed, Trumbo stays, Mancini moves into the field and Davis gets the designated hitter slot.  This si what the plan should be on opening day.

If Davis is setting the world on fire, he can press for a full time role.  Against lefties, I think this would likely mean that Mancini gets more time in left field.  Likewise, if Trumbo reflects his 2016 self then that probably means squeezing Davis into more designated hitter time than first base time.  Yes, Davis is a very good first baseman.  Yes, I am ignoring last year.  His average play last year was likely impacted by core injuries that not only limited his acceleration, but also his lateral movement.  From that, I would think he would bounce back in 2018, defensively.  I also think that Trumbo's ability there (rather decent in his youth) is not much of a drop from where Davis is.

So what we are left with is Chris Davis making 17 MM in 2018 and Mark Trumbo tacking on 12.5.  Yes, I am suggesting the Orioles start off with a 29.5 MM first base platoon.  It is not anything that any of us wanted, but it is what we got.  Time to make some lemonade.