Matthias ZiPS through the All-star Game

All-stars are generally chosen for what they’ve done so far—whether the deciding metric be batting average, OPS, WAR or something else. Using Dan Szymborski’s ZiPS projections, let’s take a look at how good we expect these players to be in the future.

First, the starting lineups.

Player Pos Proj. WAR/600 Player Pos Proj. WAR/600
Joe Mauer C

4.2

Yadier Molina C

5.5

Chris Davis 1B

2.7

Joey Votto 1B

5.1

Robinson Cano 2B

5.1

Brandon Phillips 2B

2.7

J.J. Hardy SS

2.7

Troy Tulowitzki SS

6.1

Miguel Cabrera 3B

6.3

David Wright 3B

4.5

Mike Trout OF

6.6

Bryce Harper OF

3.8

Adam Jones OF

2.9

Carlos Gonzalez OF

4.1

Jose Bautista OF

4.7

Carlos Beltran OF

2.5

Total

35.1

34.4

It’s hard to get much more even here, with the American League enjoying a statistically insignificant 0.7-WAR advantage. There’s a trio of Orioles that seems to be hindering the AL. Despite putting up 9.2 WAR in the first half, ZiPS expects just 8.3 WAR from Davis, Hardy and Jones over roughly their next full season. Though I’m picking on the Birds, it’s not surprising that many of these players have outpaced their projections—that’s why they’re in the All-star game. But there’s no guarantee of continuing this level of production. In fact, I would bet most of these guys will regress some in the second half due to natural causes.

On the National League side, ZiPS is unkind to aging stars, Phillips and Beltran. But the NL gets huge advantages at first base and shortstop, where Tulo and Votto go Orange Bird hunting. On a side note, it’s crazy to think that Trout and Harper are facing off in the All-star game for the second time already. Harper can’t even drink legally yet! Oh, wait! Never mind.

Skipping over the nitty-gritty of bench players, it suffices to say that the AL averages 3.2 projected WAR per bench player, and the NL 3.3. On to the pitchers.

Player Pos WAR/200 Player Pos WAR/200
Max Scherzer SP

4.6

Matt Harvey SP

3.5

Felix Hernandez SP

4.8

Clayton Kershaw SP

5.1

Clay Buchholz SP

3.5

Madison Bumgarner SP

3.4

Bartolo Colon SP

3.2

Cliff Lee SP

4.9

Yu Darvish SP

5.0

Adam Wainwright SP

4.8

Justin Masterson SP

2.9

Jordan Zimmermann SP

3.7

Chris Sale SP

4.8

Patrick Corbin SP

3.1

Chris Tillman SP

1.4

Jose Fernandez SP

3.0

Mariano Rivera RP

2.9

Jeff Locke SP

0.9

Jesse Crain RP

3.2

Travis Wood SP

2.4

Joe Nathan RP

3.6

Aroldis Chapman RP

2.4

Glen Perkins RP

3.0

Jason Grilli RP

4.0

Brett Cecil RP

1.1

Craig Kimbrel RP

5.2

Grant Balfour RP

1.7

Mark Melancon RP

3.0

Steve Delabar RP

1.4

Sergio Romo RP

4.3

Greg Holland RP

4.0

Matthias Kullowatz TDI*

0.0

Total

51.0

53.7

 

Like the lineups and bench, the pitchers are almost dead even. Though there are always quibbles about who got snubbed—often legit quibbles—and how each league could have picked better players, these are two very talented, very evenly matched teams. The AL went with more natural relievers, which I guess makes sense since a large percentage of pitchers in the All-star come in as relievers. The NL gets them back later in quality, though.

ZiPS seems to like the depth of the AL starting pitchers slightly more, with 3.8 projected WAR per starter versus the NL’s 3.5. But the NL gets them back bigtime in the ‘pen, where Chapman and company dominate 3.8 to 2.6. We all know if big-game Felix pitched the whole game, it would be a landslide for the AL squad, but at least we can be pretty sure he’ll get an inning.

Of course, WAR has all its shortcomings, especially in how relievers are valued, and simple totals are somewhat misleading since some players will play more. To conclude, I created some logical weights for each player, giving more to the guys will probably play more in the All-star game.

And the final tally is:

NL – 3.9, AL – 3.7**, which agrees nicely with Vegas’ 51% NL edge.

 

*TD stands for “turned down invitation.”

**Those units, by the way, would be average weighted WAR per player, per 600 plate appearances or 200 innings pitched.