Looking at some ZiPS projections

Monday was the start of the Winter Meetings, and it also happened to be one of my favorite days of the offseason. One of the hallmarks of baseball’s winter is projection systems dropping and projecting finishes for every team around the league. Among my favorites of these systems is the ZiPS system, developed by Dan Szymborski and published on Fangraphs every year. On Monday, the ZiPS projections for the Red Sox dropped, and you can see them in full here. Keep in mind that the WAR totals on the depth chart at the top of the page are different from the ones in the tables below. The depth chart factors in expected playing time.

Before we get into it, just a word about projections. I think most reading this site are probably on board with projections in theory, but there’s always pushback against this time of year. Every time someone discusses projections someone else has to jump in and declare that projections aren’t perfect and they get things wrong, as if that’s not a thing everyone knows. We are all well aware of the short-comings here! All this represents is another piece of information to consider when thinking about potential performances in 2019. Some players will outperform their projections and some will underperform. That’s why we watch. Anyway, don’t be that guy. Now, let’s get into some of the names I was interested to see and a few who I was surprised to see.

The one-two punch in the lineup is: Good

No surprise here, but ZiPS expects the one-two punch of Mookie Betts and J.D. Martinez to be dominant yet again in 2019. Betts in particular is an absolute monster in these projections. The 6.7 wins he is projected for may seem a bit low considering he’s outperformed that by a significant margin in two of the last three years. However, remember that these projection systems are conservative by design and will always regress towards the middle. To be projected for almost seven wins is huge praise. Overall, ZiPs sees Betts putting up a 138 OPS+ with elite defense and approaching a 30-30 season with 28 homers and 29 stolen bases. Again, this is from a system that errs towards conservatism. Mookie Betts is good.

Then, there’s J.D. Martinez, who is projected to be the best hitter on the team by a thin margin. Betts obviously outpaces him in the field and on the bases, which makes him a better overall player by a safe margin, but Martinez barely edges out the reigning MVP with a 140 OPS+. The power is obviously the carrying card, with 36 homers projected for the slugger.

Photo by Maddie Meyer/Getty Images

Dustin Pedroia is....fine

The projection I’m most curious about for every system is going to be Dustin Pedroia, because I have no idea what to expect. Now, I don’t expect projection systems to be infallible about a guy who basically didn’t play at all in 2018, but again it’s just another data point. ZiPS, for whatever it’s worth, sees him as being solid enough but certainly not the Pedroia of old. They only project him to play in 96 games — I think the Red Sox would hope for 20-30 more games than that, at least — but on a per-game basis his value is fine. Ultimately, ZiPS projects Pedroia for 1.6 WAR over 96, which is a 2.5-win season over 150 games. That’s a good, solid regular. Unsurprisingly, much of that value is carried by solid defense — they still see him as above-average with the glove — and positional adjustment. At the plate, Pedroia is projected to an OPS+ of 88. None of this would blow us away, but if the veteran could produce like this over a bit of a larger sample, it would be fine.

A step forward for Rafael Devers?

Along with Pedroia, I’m very interested in what the projection systems think of Rafael Devers. The young third baseman is one of the most exciting players on the Red Sox, but he was also very frustrating for a lot of 2018. Now, a lot of that frustration came on defense and I’m not super interested in projections for defensive stats, but he also has work to do at the plate. After seeing what Andrew Benintendi did as a follow-up to a disappointing first full season — I acknowledge that Devers and Benintendi are totally different stylistically — I am optimistic about Devers. ZiPS is, too. They aren’t calling for a full breakout, but again that’s not really what projection systems do. The power is the standout here, as they have him hitting 27 home runs with a .215 Isolated Power and a 109 OPS+, that last number being a 15-point increase.

ZiPS would get rid of Blake Swihart

The big question on offense for the Red Sox is what to do behind the plate, as they are almost certainly going to get rid of one of Christian Vazquez, Sandy León or Blake Swihart. ZiPS makes the comparison between the three pretty easy as they project similar playing time for all three. None of them are projected to do anything special, with both Vazquez and León being projected as replacement-level players. That is still better than Swihart, though, who is not only the worst defensive player of the group but also projected to have the lowest OPS+. I don’t think I’d agree with that last part — I’d probably have him second above León — but if 2019 is all you care about then ZiPS would have you move on from Swihart.

The rotation is outstanding

For as great as the Red Sox offense was last year, but the end of the season the idea of a fully healthy Red Sox rotation was mouthwatering. They never really got it all together at the same time, but if they can do so in 2019 it could be special. ZiPS is a big fan of all five Red Sox starters, projecting them as the five most valuable pitchers on the roster. That may seem like a given, but remember that ZiPS also projects free agents who left a given team, meaning Craig Kimbrel is projected here. By ERA+, ZiPS ranks the group as Chris Sale, David Price, Nathan Eovaldi, Eduardo Rodriguez and Rick Porcello and has them at marks of 168(!), 117, 111, 110 and 101. All five pitchers are above-average, even just very slightly. That’s outstanding. Rodriguez in particular stands out here, and in terms of WAR he is projected to be the number three starter in this rotation.

Ryan Brasier = Heath Hembree?

Ryan Brasier is one of those guys that projection systems are going to have a hard time with, but also the kind of guy we may overrate a bit. That’s not to say I’m not as confident in Brasier as most others. It’s hard not to believe in him after what we saw last year. That being said, projections are never going to love older players who come out of nowhere for a great season, and we’ve seen plenty of those players fall back after their breakout years. ZiPS projects Brasier to finish with a 116 ERA+ and a 3.83 FIP, which aren’t terrible but not really special either. Additionally, his projections are almost identical to those of Heath Hembree, who we’d all put below Brasier on the depth chart. I think we’d all disagree with that, and again I think projection systems in general can have a hard time with these kind of players, but I also found it very interesting.

Photo by Harry How/Getty Images

Tyler Thornburg is done

If the Red Sox bullpen jumps to the next level in 2019 without a couple of significant additions, it would likely be because Tyler Thornburg came close to approaching his Milwaukee levels as we’ve hoped for over the last couple of years. I’m not confident at all in that happening, and these projections aren’t helping those feelings. ZiPS is very low on Thornburg at this point, projecting an 88 ERA+ and a FIP of just below 5.00, putting him on the same level as recent minor-league signing Domingo Tapia. Not great!

Some interesting prospects

The Red Sox farm system isn’t great, but there are a few prospects in here that interested me. Just note that Durbin Feltman was not included, presumably because there’s not enough data for ZiPS to make an accurate projection.

• Michael Chavis is projected to be just below a league-average hitter in 2019 with 17 homers in 94 games. In fact, ZiPS projects Chavis to be better in terms of full value than Mitch Moreland in less playing time.
• Bobby Dalbec is projected to hit 21 home runs in 116 games, which is outstanding, but also the only good part of his game. There’s still development needed in Dalbec’s plate discipline, so this projection makes sense.
• Josh Ockimey is the Red Sox prospect most likely to be selected in the Rule 5 Draft, and he’s projected for an 80 OPS+. Again, I don’t think he’ll be able to stick on an active roster all year.
• I was very interested in Darwinzon Hernandez’ projections, but they come mostly as a starter. ZiPS gives him an 83 OPS+ with seven walks per nine innings, which isn’t great! But if Hernandez does see the majors this year, it will be as a reliever.
• Travis Lakins should see the majors at some point, but again ZiPS projects some starts in there because he spent part of 2018 as a starter. Overall they have him as about a league-average pitcher.
• The two most interesting and surprising names here are Matthew Kent and Dedgar Jimenez, both of whom are projected to be worth at least one win. I would bet against both of those.

The Comps

My favorite part of the ZiPS projections are the comps, when the system compares a current player to a historical one. This is also the most useless part of the system, and these are not meant to be perfect matches. They’re mostly just for fun. Here are a few of my favorites.

• Mookie Betts: Al Kaline
• Xander Bogaerts: Alan Trammell
• Andrew Benintendi: John Kruk
• Michael Chavis: Mark Reynolds
• Chris Sale: Johan Santana
• David Price: Frank Viola
• Nathan Eovaldi: Carl Pavano

What do you think of the projections? What interesting ones did I miss? What were your favorite comps?