Who’s an All-Star? Ah, but what numbers?

I find All-Star Games to be frustrating experiences. It’s not just that it’s an exhibition game that has had false significance attached to it in recent years (“This time it counts!”). I find it frustrating that there is seeming consent that the players who should be on the All-Star teams are those who are having the best season to date. We’re currently about 45 percent through the season, and Lady Luck still has a big say in who is doing well and who is not. Is the All-Star Game really supposed to be about rewarding the players who have experienced the best fortune in small sample sizes?

Instead, I think All-Star games should be about the best players. And so, as I did last year, I’ve put together an All-Star ballot based on rest-of-season projections. I’ll spare the nitty-gritty details until later, but essentially I’m using an average of THT’s Oliver and BPro’s PECOTA for offense and pitching (the numbers are park-adjusted), and a combination of UZR and nFRAA projections for fielding. With that said, let’s get to it!

American League

Catcher: Joe Mauer, Minnesota Twins

Mauer’s time on the DL this season still has a shroud of mystery to it, because reports conflict on whether the bilateral leg weakness he experienced was really attributable just to a late start and a viral illness. For that reason, if I’m choosing a team for a full season, I’d gladly take Santana over him. But when he’s healthy, Mauer’s combination of top-notch hitting and plus defense make him the best pick for the AL catching gig.

First base: Miguel Cabrera, Detroit Tigers

This really is too close to call. Cabrera probably has a slight advantage over Adrian Gonzalez at the plate. But depending on how large you think the fielding gap really is, you might go with Adrian here.

Second base: Ben Zobrist, Tampa Bay Rays

Cano has been better over the past year and a half. But in 2009, Zobrist was a legitimate MVP candidate. The projections remember this and peg them as equivalent hitters, which allows Zobrist’s fielding edge to carry him. I’m surprised to see Tsuyoshi Nishioka fare so well, but projections based on Japanese leagues must carry a substantial margin of error… Or, maybe he really is that good. I’m looking forward to finding out.

Third base: Evan Longoria, Tampa Bay Rays

I think Longoria would be the consensus pick here, even though he’s just the third-best hitting third baseman in his division. It’s his fielding skill that drives the difference.

Shortstop: Yunel Escobar, Toronto Blue Jays

AL shortstop has, for several years now, been a position lacking any legitimate stars. While Asdrubal Cabrera and Jhonny Peralta have had brilliant first halves, the rest of their body of work isn’t enough to get them on this list (they finished sixth and ninth, at -3 and -4 RAA, respectively). Yunel Escobar, on the other hand, combines plus fielding with solid hitting, and that was enough to get him to the top. Note, however, that just 12 runs separate the No. 1 and No. 10 names on the list…so we’re sort of splitting hairs.

Left field: Josh Hamilton, Texas Rangers

Josh Hamilton is an amazing player, as long as he can stay healthy. He’s healthy now and, while not producing at the level he did last year, he continues to put up numbers at a superb pace. The No. 2 and No. 3 players on this list aren’t in the same league with Hamilton at the plate, but their superb defense in left pushes them over the likes of Luke Scott and Josh Willingham.

Center field: Curtis Granderson, New York Yankees

Granderson’s long been a favorite of mine, and his first-half surge this season has been enough to push him over the edge. His closest competition, Grady Sizemore, may be projected a tad high given how he’s been slowed by injury issues. That said, his power surge earlier this year showed he can still be an offensive threat.

Right field: Jose Bautista, Toronto Blue Jays

You can still make the argument that the projections haven’t yet caught up with Bautista’s apparent change in talent level. But even so, he still rates as the No. 1 right fielder in the American League. Nelson Cruz makes it close, however, given his fielding prowess. I was pleased to see Matt Joyce place so well.

Starting pitcher: Felix Hernandez, Seattle Mariners

King Felix reigns supreme. But there are a bunch of top-notch pitchers in spitting distance of him, including the amazing and still-underrated Justin Verlander. On the Alexi Ogando thing: I think the projection systems are still projecting him as a reliever, and haven’t adjusted for his role change in their projections. Jake Peavy and to a lesser extent Erik Bedard are probable also controversial, but the projection systems don’t know about their injury history.

Relief pitcher: Mariano Rivera, New York Yankees

I just find it amazing that even after all this time, as old as he is, and for as long as he’s been doing it, Rivera still ranks as the best closer in baseball. And it’s really not even close. Joey Devine was a surprise, given that he missed two years and has never really been so amazing. I guess the projections like his historically high strikeout rates, but I think they’re underestimating his walk rate (career 4.6 per nine innings).

National League

Catcher: Brian McCann, Atlanta Braves
McCann’s the best catcher in the league, and has been for a while now. But after that, we have a dead heat of good quality players. Chris Iannetta and Geovany Soto are the best hitters in the group, but Chico Ruiz and Ryan Hanigan earn great marks for their fielding. I’m not showing it, but the No. 6 catcher on the list? McCann’s backup David Ross (+7 RAA). He did briefly have a starting job for the Reds, but given his power and fielding skills, I’m surprised he hasn’t gotten another shot.

First base: Albert Pujols, St. Louis Cardinals

Unfortunately, given his untimely injury, he will have to sit this one out, which grants Joey Votto the gig. I was surprised to see another injured first baseman make the list, Ike Davis. I think I’ve always badly underrated him. I didn’t realize that his fielding was plus.

Second base: Chase Utley, Philadelphia Phillies

Utley is another who may be overrated here because he’s lost a step due to injury. But even so, I’m not sure that I wouldn’t rank him as the best second baseman in the league… at least as long as he can stay healthy enough to be on the field.

Third base: Ryan Zimmerman, Washington Nationals

I may well be underestimating Zimmerman here: His UZR projection entering the season was +13 runs, but to my surprise his 3.5-year nFRAA average is just -2 runs per season. I don’t know which is right, but I’ve never heard anything but glowing reviews for his fielding prowess.

Shortstop: Troy Tulowitzki, Colorado Rockies

The King Is Dead! Long Live the King!

Calling Hanley Ramirez dead is certainly an exaggeration, but his disastrous first half pulls him below the more well-rounded (and awesome) Tulowitzki. That said, they are still the top two talents at their position. I was surprised to see Furcal rank so highly, but his fielding continues to rate very well.

Left field: Matt Holliday, St. Louis Cardinals

Holliday edges out Ryan Braun for the starting job in left field. Both are outstanding players, however, and it could go either way. If you wanted to group the two corner outfield positions together, Braun would rank above any NL right fielder. Nice to see Logan Morrison make his debut on this list: I would take him over Jason Bay any day of the week.

Center field: Andres Torres, San Francisco Giants

To say I was surprised to see Torres at the top of the list is a colossal understatement. But despite Matt Kemp‘s superior bat, the fielding metrics are not thrilled with his fielding. The same could be said about Andrew McCutchen, which was a surprise to me as well. Meanwhile, Torres gets rave reviews. The fielding difference is enough to push him to the top of the list, though all are easily within the margin of error. Almost as surprising was Cameron Maybin’s debut on the list…I’m skeptical about his projected wOBA, although he’s wOBAing .319 in Petco this season, so it might be about right once you adjust for park.

Right field: Jayson Werth, Washington Nationals

We have ourselves a tie! But I think these numbers are a tad overly optimistic about Lance Berkman’s fielding, so I’m giving Werth the nod. I’m excited about Mike Stanton, who seems to be doing better by most fielding measures than I expected of him. Jay Bruce brings the best glove of the group, and has shown flashes of brilliance with the bat. Unfortunately, his hot streaks are balanced by cold streaks, leading to a good-but-not-great projection for him.

Starting pitcher: Josh Johnson, Florida Marlins

The difference between Josh Johnson and guys like Tim Lincecum and Roy Halladay is durability. When he’s on, he is probably the best pitcher in the National League. Also, the Phillies have four of the top 11 starters in baseball according to these rankings: Cole Hamels ranks 11th.

Relief pitcher: Mike Adams, San Diego Padres

Even after trying to adjust for Petco, the Padres’ relief corps is ridiculous. Heath Bell is arguably the best closer in the league, and yet he may not be the best reliever on his team. The projections give that title to Mike Adams, who is nothing short of filthy. I was surprised to still see Broxton rank so well, but his peripherals weren’t nearly as bad as his ERA last year…and this year, he’s thrown only 12 innings. Perhaps the biggest surprise for me was Rafael Betancourt, who quietly has had himself an outstanding career as a reliever. He has accumulated nearly 12 fWAR since his debut in 2003.

References & Resources
Where these numbers come from:

Offense: I used current Oliver and PECOTA projections, updated June 24, as my data source. Batting runs were calculated using linear weights that were extracted by the +1 method from a base runs equation (found here) run on 2011 MLB totals. Park adjustments were done using Patriot’s 2010 park factors, and then the data were converted into RAA per 700 PA and wOBA. These figures were averaged across the two projection systems.

Fielding: For most positions, I used Steve Sommer’s preseason UZR projections as well as “my” 3.5-year nFRAA averages that were pulled from BPro, all converted into per-162 game season values. For catchers, I used a 2.5-year BRef’s Rtot score averages (pro-rated to 162 games) to assess impact on the running game, plus BPro’s nFRAA 3.5-year average to assess performance on bunts and soft grounders.

Pitching: The ERA* numbers above are composite statistic. I park-adjusted projected ERA (from both PECOTA and Oliver), and also calculated a projected FIP (park-adjusting the HR data). ERA* is the average between the two numbers. I did this to help remove some of the fielding influence on the projected ERA numbers, as well as force it to be a bit more DIPS-y.

No position adjustments were used above, as players were compared only within their position.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Checking the schedule
Next: Who’s an All-Star? Let the numbers decide »


  1. geo said...

    I don’t find that there is seeming consent that the players who should be on the All-Star teams are those who are having the best season to date. I find that there is seeming consent that only Yankees and Red Sox are allowed to participate.  Their fans continually gripe about how “this one counts” means that only the best players should be out there, yet continually shoot themselves in the foot by voting in Derek Jeter.  They get what they deserve.

  2. jinaz said...

    I have him 7th at -6 RAA, behind the five above list plus J.P. Arencibia (-3) and Kelly Shoppach (-3).  Neither Oliver nor PECOTA is particularly impressed: .313 wOBA and .307 wOBA, respectively.

    Avila’s interesting.  He has pretty good minor league numbers (especially with his walk rates), but has never shown power at the level he has thus far.  Kevin Goldstein had him as a 3-star prospect following 2009…
    …so he has some pedigree, but not superstar expectations.  And last year, he was pretty much terrible at the plate.  His fielding numbers are good this year, but below-average last year.  He was a little bit trendy this preseason, but I don’t think anyone could expect what he’s done so far this season.  And I don’t really expect it to continue at this level.  That said, I think he’s a solid backstop.

  3. jinaz said...

    @geo, I keep having conversations with people who argue that the all star game is about the players having the best year.  There will be innumerable articles about how x player got snubbed because he’s having a great season.  That’s where this is coming from.

    As for Jeter…there’s a sense in which if fans want to see him, they should get to see him.  I’m ok with it just being about popularity.  But if we’re going to use numbers, I’m for using projections over just two to three months of data.

  4. jinaz said...

    Well, you can’t read projection systems blindly.  In Bay’s case, there are injury issues (the concussion in particular, but also the intercostal strain this year) that could have directly caused a sharp change in his talent level that would outpace the general models the projection systems use.  Whether those changes are permanent or temporary remain to be seen, though to be sure he’s been so bad this year after a big drop-off last year that you have to be pretty skeptical about him.  That said, it’s only been 625 PA’s since 2009 when he was a very good, 5 WAR player.  A comeback, at least a modest one, is not out of the question…though I’m not going to hold my breath.

    So, what I’m saying is, yes, I think the projection systems are overestimating Bay’s true talent.  If that’s the worst miss in this whole post, I think we’re doing pretty well.  No method is perfect, but I think this approach resulted in a pretty good representation of the best players at each position across the two leagues.

  5. Trevor said...

    No Kimbrel or Venters for the NL relief pitchers? Arguably the two best relievers in the league/majors.

  6. jinaz said...

    They’ve been spectacular.  But it is in a small sample—even with Venters, he has less than 200 IP in his MLB career.  So will they continue to be Maybe.  Maybe not.  There’s a reason that relievers are perceived as being so volatile—they pitch so few innings that we almost can’t help but judge them on small sample sizes. 

    Right now I have Kimbrel 16th and Venters 40th.  With Venters, his projection gets dragged down by a lackluster minor league career as a starter.  Both are probably too low, but I wouldn’t rank them at the top of the league yet either.

  7. jinaz said...

    Check back on that at the end of the season.  Players who have bad first halves and get labeled “washed up” have a nasty tendency to bounce back in the second half.  That’s why the projections use as many seasons as they do—it allows them to be more predictive.

    There are some players on this list that may well be washed up.  Jason Bay has come up a few times and I agree he is not looking particularly good.  But I would probably have said the same thing about someone like Lance Berkman last year, and he’s shown that he is clearly not done despite his leaky knee capsules…

  8. jim said...

    I think it’s a good thing you aren’t picking the teams. Lots of washed up guys on these lists.

  9. jinaz said...

    The point of using projections in this exercise isn’t to predict the future, per se, it’s to estimate true talent levels of ballplayers.  That way, we can pick the best players for the ASG without being unduly influenced by their most recent performances. 

    No one would argue that we use them to determine the playoff winners, or even the regular season winners.  Or, for that matter, the outcome of the ASG.

  10. Ousy said...

    Well we might as well just project who will win the World Series and skip the playoffs all together.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>