Spring Training is underway and that means two things. One, baseball is back! Hooray for that. And two, prospect ranking season and projection season have arrived. We’ve already seen a ton of prospect rankings. Earlier this week, Dan Szymborski released his ZiPS projections for the 2019 Yankees. The fancy graphic above shows the team’s WAR projection by position. Spoiler. The Yankees project to be very good. I don’t think we needed ZiPS to tell us that.
Obligatory reminder: Projections are not predictions. They are an attempt to estimate the player’s current talent level. There are more projection systems out there than I care to count and they all go about it in different ways. Is one better than the other? Eh, not really. ZiPS is my preferred projection system so that’s what I’m going to write about. Here are five things that stand out to me about the 2019 Yankees ZiPS projections.
1. LeMahieu should play over Tulowitzki. Not sure you need a projection system to see this. Tulowitzki hasn’t played since July 2017 and, the last time he did play, he wasn’t very good. He hit .249/.300/.378 (79 wRC+) in 66 games before getting hurt. Also, Tulowitzki is 34 now. Even if he were perfectly healthy, he’s at the point where you’d expect age-related decline. Aren’t teams avoiding free agents this age? Not when they’re essentially free, I guess.
ZiPS pegs Tulowitzki as a below-average offense (77 OPS+) and league-average defense (+0 runs) player who projects out to +0.9 WAR per 600 plate appearances. It has LeMahieu as a below-average hitter as well (93 OPS+), but not that far below average, and it likes his defense too (+5 runs). The total package projects out to +2.2 WAR per 600 plate appearances. That is not an insignificant difference. Not in what could be a tight AL East race.
I totally understand why the Yankees are rolling the dice on Tulowitzki. I don’t understand why they’re seemingly so dedicated to him as their starting shortstop. “We’re planning on him playing shortstop and focusing solely there,” said Aaron Boone last week. In a perfect world the Yankees would give Tulowitzki a look in camp and a few weeks in the regular season. If he performs well, great, keep him. If not, I hope they don’t hesitate to move on. Not with LeMahieu ready to step in.
2. First base should be okay. Not great, but okay. PECOTA loves Luke Voit. That system has him as the 22nd best hitter in baseball in 2019. ZiPS is not that optimistic, though it sees a .264/.344/.474 (116 OPS+) batting line with 22 homers and +1.9 WAR in just under 500 plate appearances. That slash line looks like peak Joey Votto compared to what the Yankees have gotten at first base since Mark Teixeira’s last great year in 2015:
- 2016: .212/.293/.380 (71 OPS+)
- 2017: .244/.317/.444 (83 OPS+)
- 2018: .234/.309/.453 (96 OPS+)
ZiPS has never been all that high on Greg Bird — it projected him for +1.9 WAR total from 2017-18 — and this year it projects him for a .224/.317/.444 (101 OPS+) line and +0.6 WAR in 350 plate appearance. Probably not good enough to start at first base for a contending team, in other words. First basemen collectively hit .253/.333/.438 (112 OPS+) last season, their worst season since 1953 and fifth worst season on record. They (relatively) stink right now. Voit’s projection is middle of the pack relative to the rest of the league. For the Yankees, middle of the pack at first base would be a massive upgrade over the last three years.
3. A historic bullpen projection. The +7.9 WAR projection you see in the graphic atop the post is the highest bullpen projection ZiPS has ever spit out*. The previous record? The 2018 Yankees at +7.7 WAR. ZiPS projected the bullpen for +7.7 WAR last year and they finished the season at +9.7 WAR, making it the best bullpen in baseball history (per fWAR). (The Yankees also set a record with a 30.2% bullpen strikeout rate last year.)
* For what it’s worth, Szymborski says the bullpen projection would’ve been +8.3 WAR had the Yankees kept David Robertson rather than sign Adam Ottavino. Zack Britton (+0.8 WAR projected) over Robertson is the one that irks me though, not Ottavino (+0.9 WAR projected) over Robertson.
Bullpens are notoriously volatile but that doesn’t reduce the value of having high-end talent. Ottavino is more likely to be an above-average setup man this season than, say, Joe Harvey. Britton is a better bet to be a high ground ball lefty than Stephen Tarpley. When you’re the Yankees and you can throw money at the best available players, you should do it, and they did it with the bullpen. No, the bullpen is not guaranteed to meet (or beat) projections. ZiPS tells us they have an awful lot of relief talent though, and I’d rather bet on expensive talent than cobbling together seven or eight cheap arms and hoping for the best.
4. Does Green’s projection match reality? The best individual player projection in that stacked bullpen belongs not to five-time All-Star Aroldis Chapman or high-priced free agent Zack Britton. It belongs to Chad Green. ZiPS pegs Green as a true talent 2.70 ERA (2.77 FIP) pitcher with great strikeout (30.7%) and walk (6.4%) numbers. I have to think his +1.7 WAR projection is among the very best for relievers around the league this year.
ZiPS uses statistically similar players (weighing more recent seasons the heaviest) and aging curves to generate its projections. The system is aware of injuries but not necessarily how a player produces his results. For all ZiPS knows, Chapman throws 89 mph. In Green’s case, he is a one-pitch pitcher, and that one pitch was less effective last year than it was the year before. The breakdown of Green’s heater:
|Avg Velocity||Spin Rate||Whiffs-per-Swing||xwOBA|
|2017||95.8 mph||2,484 rpm||39.8%||.216|
|2018||96.1 mph||2,444 rpm||27.9%||.290|
Same velocity, same spin, fewer whiffs, better contact allowed. It should be noted Green’s fastball still had a much better than average whiff rate (20.5% league average) and expected wOBA allowed (.347), but the pitch was not as effective as it was a year prior. Watching him pitch, it seemed to me the book was out on Green. Hitters knew they were likely getting a fastball and they geared up for it.
ZiPS doesn’t know Green lives and dies by his fastball. It doesn’t know he could use a better second pitch to keep hitters honest. He doesn’t need a Dellin Betances curveball or an Adam Ottavino slider. Just something good enough that hitters must respect it. I expect Green to be very good this season, but unless he comes up with a better slider or changeup, I don’t think he’ll be as good as he was in 2017, and maybe not as good as he was in 2018. He strikes me as the reliever most likely to fall short of his ZiPS projection. (Also, ZiPS says Green’s top statistical comp is Hall of Famer Rollie Fingers, and he should absolutely grow the mustache.)
5. Projections as a measure of depth. I like using ZiPS to estimate each team’s depth. Last year the Yankees had 27 (!) players projected for at least +1 WAR. I know a +1 WAR player isn’t all that exciting, but when you have a lot of them, your internal replacement level is pretty high. It means you have some good players stashed in Triple-A. You’re not scraping the bottom of the barrel when you need an injury replacement.
I like to look at each team’s depth at three levels: +1 WAR (okay players), +2 WAR (league average players), and +4 WAR (comfortably above-average players). Simply put, how many of each does each team have? The more the better, obviously. Here’s my spreadsheet (we’re still waiting for the White Sox and Padres projections) and I should note I removed unsigned free agents. The Red Sox don’t get credit for Craig Kimbrel, the Astros don’t get credit for Marwin Gonzalez and Dallas Keuchel, etc. Here are the Yankees and their ranks:
- +1 WAR players: 22 (tenth most)
- +2 WAR players: 12 (third most)
- +4 WAR players: 3 (fifth most)
This isn’t a perfect measure because ZiPS doesn’t worry about doling out playing time (are Jed Lowrie and Todd Frazier both really going to get 450 at-bats?) and the timing can be not great (Edwin Encarnacion was still with Cleveland when the Indians ZiPS were released) but it’s a good ballpark estimate. I’m surprised the Yankees are only tenth in +1 WAR players, especially behind teams like the Twins and Diamondbacks. Huh. Only the Cardinals and Mets have more +2 WAR players. They have 13 apiece. With the farm system thinned out a bit, especially at the upper levels, the Yankees don’t have quite the +1 WAR depth as the last two years. The top of the roster is very strong though, and the Yankees have good depth, at least according to ZiPS.