If I picked the All-Star Game rosters...

Talking Chop, I have a confession to make.

I really, really, really want to like the All-Star Game. I know other people think it’s stupid, but I think it’s a great idea. Take a baseball game, boil it down to superior performers from all of the teams across the league, and watch the best pitchers face the best hitters with something like the best defense behind them if the ball is put in play.

But, here’s the thing: I actually kind of hate the All-Star Game. Some of it is the rosters, but a lot more of it is that the way the rosters are used is not really in service to try and win the game. Even when it counted (and how miserable that was...), there was too much managing by rote, too much roster-injury-swap junk that gave second-string All-Stars a lot on-field time due to an unwillingness to participate in an exhibition game, and so on. In my mind, because it’s an exhibition, the All-Star Game would be a perfect time for managers to experiment with things like starting the game with a shutdown reliever (because the first inning is advantageous to the offense since they are guaranteed to bring up their best hitters).

I’m sure the 2017 All-Star Game will be much of the same as what we’ve already seen: bloated with ads and interruptions, rosters semi-unrecognizable from when they were revealed this past Sunday, and some of the game’s best players conspicuously absent. I may not watch, and I probably won’t watch all the way through — it’s just not compelling enough, and this is coming from a guy who’s watched clunkers like the Braves getting slammed by the Brewers 7-0 on a Sunday afternoon.

At this point, I don’t have a great solution for how to change the All-Star Game itself to capture more of my interest. It would have to be a subtle change in terms of the value placed on its importance and relevance, I think, and that would have to come from the players, the coaches, and the league. But, what I can do is at least offer up my thoughts about the type of roster I’d like to see that I think would improve the experience, because I think having some pretty poor choices kills a lot of the potential enjoyment and anticipation right at the outset.

The rest of this post is going to consist first of my preferred rosters for the 2017 All-Star Game. Then, there’ll be a few quick tables comparing those rosters, the actual rosters selected (pending the Final Vote shenanigans), and Dave Cameron’s selections made last week on Fangraphs, just for illustrative purposes.

---

Picking the All-Star Game Rosters

I think the All-Star Game roster selection process is a tricky exercise. Is a roster spot a measure of recognition? If so, is it a measure of recognition for performance through June, or is the right time period to consider the entire stretch since the last All-Star Game? After all, since there’s no second-half All-Star Game, how fair is it to let a superb July-September performance be swept under the rug?

Alternatively, is the All-Star Game roster meant to be a reflection of the players that give their composite team the best chance to win the game? I know that the game, currently, is not treated as such... but is it meant to? I find that possibility compelling.

As a result, I’ve tried to choose my rosters with all of the above considerations in mind. Both first half 2017 and “last calendar year” performances are considered, though the former is weighted much more heavily. Usage, platoon, and defensive considerations also factor in, as I’ll discuss more with regard to individual players. For roster contours, I use the actual rosters: 20 position players, and 12 pitchers per team. Additional specifics are listed below.

• The main criterion for selection was an indexed WAR measure that combined 2017 performance through June (two-thirds weight) and “last calendar year” performance from July 2016 through June 2017 (one-third weight). Note that since the latter includes 2017 performance, this is effectively a 75-25 weight on this year versus last year.
• Pitcher values were assessed based on a 50-50 split of fWAR and RA9-WAR. This likely did not have a substantial effect on the results.
• For position player roster spots, I selected the starting eight for each league (no Designated Hitter, due to the game taking place in a National League park, despite the weird voting process that still elected a DH for the American League), and then a backup for each position. That left four “Wild Card” spots of the best remaining position players in each league. This is consistent with the actual rosters, once the Final Vote candidates are added on.
• For pitching spots, the split between relievers and starters differs substantially in the actual rosters. The current rosters feature three relievers on the AL team, and six relievers on the NL team. I’m not really sure whether having more starters or more relievers is best to try and win a singular game, but I tried to include standout performers across both relievers and starters, and ended up with the same mix for the NL, and one additional reliever (at the expense of one starter) in the AL.
• My initial cut failed to include players from four teams: the Athletics, Orioles, Padres, and Phillies. Three of these teams had relievers added on (booting some other players off). Notably, with a crowded AL infield, the Orioles’ only standout player in 2017, Jonathan Schoop, is not on my list. What the Orioles have done thus far with really only one good performance from their position players and starting pitchers is remarkable.

Enough chatter: here’s the NL roster:

Starters are pretty easy choices - these are just the best performers at their respective positions. Yes, Ender Inciarte warrants a starting spot here, by the algorithm described above. Rather than pushing Charlie Blackmon to the reserves, though, I think having both of them roam the outfield would be great. All-Star hitters are more likely to drive the ball in the air and to the gaps, so good outfield defense is paramount. (Of course, as leedawg noted, Max Scherzer may also just strike everyone out, so in that case outfield defense would be redundant. I feel better having Blackmon and Inciarte in the outfield, though, than I would replacing Inciarte’s bat with the upgrade that an Adam Duvall, Marcell Ozuna, or Michael Conforto would provide.)

The NL’s position player strength in recent history has apparently been concentrated in its corner infielders: a bunch of sluggers comprise the “Wild Card” reserves and create somewhat of a glut there. By comparison, the outfield choices aren’t super-exciting. Generally, though, these guys provide good platoon partners for the starters. Grandal hits righties very well, and his framing is relatively comparable to Posey’s. Across Goldschmidt, Freeman, Rizzo, and Votto, you have all sorts of first base options. Third base is the only same-handedness situation here; platoon partners exist for everyone else. In the end, I didn’t feel that Tyler Flowers warranted any real consideration over Grandal for the backup spot: the framing is nice, but he’s a righty hitter like Posey, and I’m not sure he offers more offensively.

Max Scherzer should definitely be starting, and the remaining starters give the NL a number of options for how to play with the middle innings, with two lefties and three righties. Of course, they may not be needed, because there are six quite-effective relievers around, too. Pat Neshek and Brad Hand are the “undeserving” fill-ins for their respective teams, but honestly aren’t too far off from the guys they booted off (Raisel Iglesias, Archie Bradley) as far as performance goes.

For injury replacements, first nods for position players should go to Jedd Gyorko and Eugenio Suarez, or Willson Contreras if the injury is to a catcher. Trea Turner would be another choice, but he’s dealing with an injury of his own at the moment. The aforementioned Iglesias and Bradley would be my first choices for pitcher replacements would they be needed.

On to the AL roster:

There’s a pretty clear separation of quality here between everyone else and Sanchez/Morrison. Trout is on here because he’s Mike freakin’ Trout, but given his injury status, George Springer would likely take his place. I’m honestly not a fan of most of the AL first base options, but I think Morrison is the guy here. The AL will have to hope that the star power of the rest of their starting lineup can hide the relative lack of oomph from catcher and first.

Where the NL was crowded at the infield corners, the AL has a bevy of talented shortstops and outfielders. Again, there’s a bit of a focus on platoons, and the AL having a bunch of switch-hitters that have performed at high levels helps a ton in this regard. Having Andrelton Simmons for late game defense is a real luxury, too, though many of the infielders are hardly chopped defensive liver. Due to the relative weakness of the AL 1B field, I slotted Yonder Alonso in as Oakland’s lone representative. While there are cases to be made for Justin Smoak (or Jose Abreu) to relieve Morrison, I’m not huge fans of either for this purpose, and the ability to include Alonso as a backup 1B makes two problems go away, if perhaps not in ideal fashion. An alternative would have been to include Sean Manaea as a starter instead, but the AL has so many good starting candidates that someone was going to be kicked off either way. This caused kind of a cascade effect, where selecting Alonso over, say, Abreu and/or Smoak made it necessary to have Roberto Osuna on the roster for the Blue Jays (see below), and Avisail Garcia on the roster for the White Sox (because I have no idea who else would even make sense for the White Sox). So the roster kind of gets to be a mess at the fringes in the AL as a result.

This is a brutal pitching staff to contend with. Aside from Chris Sale to start, there are three other tough southpaws that the AL can bring to bear (including Jason Vargas, which is a weird thing to type). Brad Brach is the lone Orioles rep, as it wasn’t worth displacing the other reserves for Jonathan Schoop, in my mind, but he’s also not an outlier as far as quality goes.

There’s a very crowded field of possible injury replacements in the AL, because there’s a whole bunch of above-average types on the position player side that aren’t really more or less deserving of one another, including Jarrod Dyson, Justin Upton, Elvis Andrus, Didi Gregorious, Jean Segura, Ian Kinsler, Stephen Souza Jr., Jackie Bradley Jr., Josh Donaldson, Brian Dozier, Kyle Seager, and Avisail Garcia. As far as injury replacements for pitchers, I’d look first to Lance McCullers, and then Ervin Santana.

Unsurprisingly, MLB’s best teams, including the Dodgers, Nationals, and Astros, have five representatives each; other good teams like the Indians, Red Sox, and Yankees come in with a quarter of players on these rosters.

---

Comparison to Actual Rosters

So, now that you’ve seen the above, how does it stack up with other potential roster configurations? (Click to pop open.)

Blackmon, Harper, Murphy, and Posey are unanimous starters. I’m apparently missing the boat on Bellinger (lotsa pop, but not too sure how long his HR/FB north of 30% will last, or what happens when it regresses) and Zimmerman (horrible last year, not great in the field if needed). Meanwhile, MLB’s choices of excluding Freeman, and putting Rendon/Turner on the Final Vote ballot rather than just giving them spots seems very questionable. I’m also apparently too big a fan of Duvall, Grandal, and Rizzo; meanwhile, Lamb, LeMahieu, Molina, and Stanton seem like really weird choices from MLB’s perspective, and are some of those interest-killing blah choices I mentioned earlier.

Some more spread in the pitchers. My “outliers” are Chase Anderson, Gio Gonzalez, Ivan Nova, and Addison Reed. Meanwhile, I appear to be missing the boat on Zack Greinke, Greg Holland, and Robbie Ray. All of those guys were definitely up there for me but just didn’t make the cut. On the flip side, it’s very strange that the actual roster does not include Felipe Rivero and his current video game number stats, but has Wade Davis instead (good, but not quite as good). Hand, Jansen, Kershaw, Knebel, Martinez, Neshek, and Scherzer are the unanimous selections: kind of funny that it’s a mix of dominant starters/relievers, and then also the gimmes of Neshek and Hand to fill spots for their respective teams.

For the AL, the comparison looks like this (click to pop up):

Altuve, Correa, Judge, Ramirez, and Springer are unanimous starting choices, and Trout would be too had Dave Cameron included him. My main omissions are Justin Smoak (discussed above), Salvador Perez (meh choice at a relatively weak position), and Jonathan Schoop (fine in a vacuum, but wanted to fill the roster with other players). Meanwhile, I apparently like having speedy Yankee outfielders on the roster (Hicks, Gardner), a sentiment not otherwise shared; Cameron and the actual roster also could not find room for Andrelton Simmons.

Meanwhile, some notable actual roster weirdness includes omitting Alex Avila and Robinson Cano, while pushing Xander Bogaerts and Logan Morrison into the Final Vote. Meanwhile, I’m not really sure why Michael Brantley, Starlin Castro, and Nelson Cruz are on the roster. Brantley’s been pretty average, and Cruz has been a powerful DH but his numbers are not egregious.

I apparently missed the boat on Betances (he has an 18% walk rate this year), Severino (just crowded out), and Lance McCullers (who would have been next on my list). Chris Archer has been nuts this year, but not so nuts that either the actual rosters or Dave Cameron included him; and my Brad Brach choice might be weird but the Orioles needed someone, and I didn’t include Schoop.

Notable actual roster omissions include Chris Devenski (no respect, man) and Roberto Osuna (which I get, since MLB included Smoak, as a starter, no less). Meanwhile, I get why MLB included Ervin Santana, though I think there are more deserving candidates (like Archer). Overall, the AL pitching array looks pretty good, I think.

---

So, with all that, how do you feel about the current rosters? Which omissions grind your gears? Which guys do you think shouldn’t be there at all? Do you hate the requirement that each team have at least one representative, which seems to result in some undeserving players getting spots?

Note: You can find Dave Cameron’s posts on selecting All-Star Game rosters here and here. I used the current MLB All-Star Game rosters unveiled on Sunday night, and memorialized here; they are subject to change as injury replacements get added, to the point where the roster gets semi-unrecognizable.