Baseball Notes

On Sunday afternoon, the (almost) complete All-Star game rosters will be announced for the 2007 game in San Francisco on Tuesday, July 10. A buddy and I got into a discussion last night about what we would change in the All-Star game if given the chance. I don’t know that my response would be the same as most people out there, but all of my answers hinged on one important factor: the game actually counts for something. As long as we have the All-Star game count for home-field advantage in the World Series, the current system is flawed. Fans vote for who they want to see play in the game, but 100% of the time, they do not always pick the most deserving players at every position.

If the game does not count for home-field advantage, I have almost no problem with the current system, and fans can vote for whomever they want, because it will be for them, not for the World Series teams. We joked around, saying why don’t we just run a poll on MLB.com that reads, “Who would you rather see have homefield advantage? NL or AL?” As long as fans vote for their “favorites,” and not the “best,” you can’t have a fair representation of the elite of the AL versus the elite of the NL.

So that got me thinking, “What does make fans vote the way they do?” Does it have to do with stats, the most recognizable players, the all-time greats, or something else? Looking at the most recent voting results that were released Wednesday, you can eliminate a couple of potential voting factors:

NL
C - Russell Martin
1B - Prince Fielder
2B - Chase Utley
SS - Jose Reyes
3B - David Wright
OF - Carlos Beltran
OF - Ken Griffey, Jr.
OF - Alfonso Soriano

AL
C - Ivan Rodriguez
1B - David Ortiz
2B - Placido Polanco
SS - Derek Jeter
3B - Alex Rodriguez
OF - Vladimir Guerrero
OF - Magglio Ordonez
OF - Ichiro Suzuki

Looking at these lists, if they stay this way, there are a few mistakes. David Wright should really be Miguel Cabrera and Carlos Beltran and Alfonso Soriano should really be Matt Holliday and Carlos Lee. In the AL, Ivan Rodriguez should really be Jorge Posada, Derek Jeter should really be Carlos Guillen, and someone who is actually a first baseman should play first (probably Youkilis since Morneau has been hurt a couple of times). But all-in-all, it is a pretty good list. But looking at it, you just don’t have the all-time greats or your most recognizable players if they don’t truly deserve it.

For example, there is no Barry Bonds, Manny Ramirez, Craig Biggio, Sammy Sosa, Albert Pujols, or Andruw Jones on these lists. The fans did a good job of leaving off these big names who do not deserve to be starters this year. Because of this, I am going to give these millions of voters the benefit of the doubt and assume this was not a popularity contest, but rather they actually considered stats to help with their voting (after all, since this counts for something, you HAVE to consider stats, right?).

So I thought it might be interesting to show some familiar stats and who the “All Stars,” or the statistical leaders, are for each position, divided up by NL and AL. Then, we will see which stats most closely correlate with who the fans actually voted to start. Using eight recognizable stats (all numbers as of July 28), this is what I found:

There are a couple of things to note here. I chose 150 plate appearances to signify playing about half the time for a player’s respective team in the first half. There were plenty of players with 50 or 60 PAs who were hitting .380 or whatever, but this weeds those out and favors the players who have been making significant contributions. Also, when you search statistical databases, you are never going to find DHs in a list of first baseman stats, they have their own listings. So an All Star ballot with David Ortiz, Travis Hafner, Jason Giambi, et al as first basemen will never be able to reflect against true first basemen statistics, which is shown above. Those DHs are nowhere to be found.
Out of the 16 categories mentioned above, 12 of them match somewhere between two and four of the fans’ picks; four of them match five picks. Those four are: NL VORP, AL Avg., AL VORP, AL RC.  -  and are the stats most closely correlated to the players the fans selected.

So whether the fans realize it or not, they are really falling in line with all of us sabermetric nerds and going with the players whose VORP and RC say they deserve to be All Stars. That’s a scary thought. Or maybe these “rarer” stats are catching up with the mainstream…who knows?

But let’s not stop there. Let’s look at the leaders of each category again and compare them to who SHOULD be All Stars.  I won’t touch the whole AL DH/First baseman thing; there is just no way around that when you are in an NL park. My changes would be:

Matt Holliday for Carlos Beltran
Carlos Lee for Alfonso Soriano
Miguel Cabrera for David Wright
Jorge Posada for Ivan Rodriguez
Carlos Guillen for Derek Jeter
By that estimation, VORP runs away with the NL, getting seven of the eight starters correct, only missing on Barry Bonds. The AL is a bit more convoluted. VORP still leads here with five correct, but Avg., OPS, SLG, and RC also have five correct.

I attribute this to one major factor. While (in my opinion) the NL is pretty cut and dry about who should be starting, the AL has a lot of question marks. Four out of eight if I am counting correctly.  Posada could be Martinez, Ortiz could be Youkilis or Morneau, Polanco could be Roberts, and Ichiro could be Torii Hunter. You have so many AL hitters doing so well in so many categories (on base, power, speed, run production), that the stats have a hard time differentiating between all the great players when you start to look at five, six, seven, eight unique stats. You will get a wide range of players doing well across the statistics. For what it’s worth, AL RBI and OBP also matched four of the should-be starters.

So the fans almost can’t go wrong in the AL. Hey, we all know the game should be for them anyway.