Why the projections think so highly of the Mets

New York projected by FanGraphs to have 5th-most WAR

February 16th, 2020

It's projection season across America, where all of the smartest minds in baseball do their best to come up with various systems that help us look ahead to the coming season. Projections aren't predictions, to be clear, and while they do a good job, they of course can't perfectly play out the future in terms of injuries and breakouts. Still, they're valuable enough to give us a good baseline expectation of what we might see.

Now, it doesn't exactly take a supercomputer to know that the Dodgers and Yankees are expected to be very good, so it shouldn't be surprising to find that at FanGraphs (using the Steamer projections) and Baseball Prospectus (using PECOTA), using very different methods, each come up with the same four teams at the top, in slightly different orders: Dodgers, Yankees, Astros, Twins. That makes sense, obviously; each club won at least 101 games in 2019, and the group added Mookie Betts, David Price and Josh Donaldson, as well as swapping Gerrit Cole and Kenta Maeda among them.

But the fifth team, in both systems? It's not the defending champion Nationals, or the 97-win Braves, or the back-to-back 97-win A's or a very good Rays team. It's a club that's already had to hire two managers in one offseason. It's the New York Mets.

It's important to note that the projections aren't really "picking winners" so much as they are taking a median range of projected player value and making a best estimate from that. (As the PECOTA explanation notes, "our simulations do not ‘pick’ a team to ‘win’ any particular number of games. Rather, they identify an estimated range of games a team might win and tells you the average of that fairly wide range.") When we say the Mets are fifth, it's by fractions of a point over other teams, including the Nationals, and nothing more concrete than that.

Still, though: Fifth! It would seem that if you asked most fans to rank the top five teams headed into 2020, the Mets wouldn't be on that list, and that's what makes this all so interesting. It's not just the public-facing numbers that seem to like New York, either. As Ken Rosenthal wrote in The Athletic earlier this month, "the internal projections of one AL team has them finishing second in the NL East with 88.1 wins, just behind the Braves at 89.4." That's obviously different than the public numbers, but you get the idea.

But still: Why do the projections like the 2020 Mets so much? Can they really live up to that? Let's find out.

1) The projections really love their starting pitching

Steamer ranking: #3

Is it possible for a projection to be both completely unsurprising and somewhat eye-raising at the same time? Of course the Mets rotation is projected to be very good, because is the back-to-back NL Cy Young Award winner, and because has been a consistently above-average pitcher, if not quite the ace that his obvious physical talent would imply he could be. They'll get a full year of , too, and stayed healthy enough to make 60 starts in 2018-19. Over the past two years, the Mets have had the third-most-valuable rotation. This tracks!

But on the other hand ... they no longer have Zack Wheeler, who is off to Philadelphia, and he's a little better than Stroman. He's a lot better than his presumed replacements ( and ) are expected to be. The best version of this rotation is the one that had both Wheeler and Stroman down the stretch last summer, not the one they'll have in 2020. There's a concerning lack of depth, too, because beyond this group, the top starting option appears to be , who has a career 8.84 ERA in 37 2/3 innings.

So it's easy to see this going sideways, but that's also true for most teams in the Majors, and most of those teams don't get to start with deGrom, do they? The Mets rotation has been good, and so long as deGrom is in one piece, they'll probably still be good. But there's more to pitching than just the rotation, and ...

2) The projections really like their bullpen, perhaps too much

Steamer ranking: #4

... this one's going to turn some heads. Realize that if you were to look at the four highest bullpen ERAs in team history, you'll find the past three years and the legendarily bad 1962 Mets:

4.99 -- 2019 Mets
4.96 -- 2018 Mets
4.82 -- 2017 Mets
4.76 -- 1962 Mets

No, we're not accounting for differences in hitting environment across the decades by doing it this way, but you get the point: It wasn't great. The 2019 Mets bullpen had the fifth-highest ERA and the seventh-lowest WAR in the Majors last year. Not great!

So where is this high ranking coming from? Let's take the seven primary expected Mets relievers -- , , , , , and -- and compare both projection systems' 2020 numbers to the 2019 actuals:

Steamer: 5 WAR
2019 actual: 3.6 WAR

Betances threw just two-thirds of an inning in 2019 due to injuries and is projected to be healthy in 2020, so this isn't exactly comparing apples to apples, but you can see that each projection system expects a pretty strong step forward from the bullpen. Lugo was quietly excellent in 2019, and a healthy Betances would certainly help, but this really comes down to what you expect from Díaz, who had, to put it charitably, a rough first year in Queens.

So is it reasonable to expect Díaz (0.0 WAR in 2019) to rebound to 1.1 WAR (PECOTA) or 1.9 WAR (Steamer)? It depends on how you feel about what happened to him last year. The good news, if you're looking for it, is that he still missed a ton of bats -- his 39% K rate was fifth best among qualified relievers -- and that a few high-profile blowups made his entire season look worse. (He allowed 36 earned runs, but 16 of them came in just four games.) He's been working with Pedro Martinez on his slider this winter, not that the projections take that into account.

It's easy to see the bullpen being better, because Betances had been elite for nearly a decade, and because there was still a lot to like about what was otherwise a disastrous 2019 for Díaz. Even an average bullpen would be a big step forward. But fourth? We'll take the under on this one.

3) Because the lineup might be deep enough

Steamer ranking: #14

The Mets failed to get the real outfield upgrade they were after this winter, but we were interested to see their name pop up relatively highly when we looked at 2020's deepest projected lineups a few days ago. They don't have the 10 projected above-average hitters that the Twins do, but with seven -- , , , , , and -- they rank near the top of the league. Aside from Canó, this group was consistently excellent in 2019, posting a combined .274/.360/.510 line.

That was just from a hitting perspective; if we roll in projected defensive ratings as well, we get this overall positional list for the Mets:

C: 22nd
1B: 5th
2B: 11th
SS: 17th
3B: 13th
LF: 16th
CF: 16th
RF: 7th

Each of which is defensible, really. Ramos isn't known for his defense, which holds him back, and none of the outfielders are terribly strong fielders. Shortstop Amed Rosario's difficult defensive start turned into roughly average performance for the remainder of the year; he also hit .307/.337/.441 over the season's final four months.

Catcher aside, the takeaway here is that each position is ranked 17th or higher, which is to say, there's no clear holes in the lineup. Would it look better with Starling Marte in center or Mookie Betts in right? Absolutely. Could it be good enough as is? With most of the same players, the 2019 Mets -- excluding pitchers batting -- did have the seventh-best offense in baseball.

So you get the idea here, which is that a potentially very good rotation, a solid enough lineup and a ... well, not-as-bad bullpen all combines to make for a team that projects to be one of the better ones in the game.

This isn't the first time we've said that, though. What's held the Mets back in the past?

4) How have projections done on the Mets previously?

We can look back through a few previous years of projections at FanGraphs, so ... how have they done with the Mets?

2019 // 84-78 projected // 86-76 actual
2018 // 84-78 projected // 77-85 actual
2017 // 87-75 projected // 70-92 actual
2016 // 92-70 projected // 87-75 actual

Which goes to show the hit-or-miss nature of attempting to project a baseball season, though you might remember the big 2017 miss was the season plagued by somewhat unpredictable injuries to Syndergaard, Conforto, , Juan Lagares and others. There are always things that are impossible to account for, like, say, the unknown effects of a rookie manager.

Last year, despite the relief issues, they actually outdid the projections slightly. If they do that again, they'd be right there in what should be a tight three-team race in the NL East. It's sometimes easy to forget that the Mets have as much top-level star power as any team in baseball. That alone doesn't get them to October, clearly. But if you're looking for a bright sign of spring, the projections are it. The 2020 Mets ought to make some noise.