NL Central Predictions, or: How Not to Make Projections

David Schoenfield has been doing a position-by-position preview of every division in baseball to predict the winners. I admit I haven’t been following every part of the series, so I’m not sure if these are actually predictions; however, I read the NL Central edition out of curiosity, and I figured I had to address it.

For those who haven’t read it, Schoenfield basically looks through every team’s projected line-up and awards points based on who is the best at each position; in the case of the NL Central, 6 points for the best, 5 for second, and so on. He then adds the total points for each team and voila!-a prediction. By this system, he has the Brewers beating the Reds and Cardinals, 77-73-68. This actually shocked me; I think the Reds and Cardinals will be battling for the top spot, with the Brewers needing some luck or injuries on the other teams to make a run.

Really, though, the system seems very flawed. I could have gone through this process with any other division, more than likely; I just so happened to see the NL Central piece the other day on ESPN’s front page and read it. First of all, the premise has issues; because the system awards points on a positional basis, you wind up with very strange quirks. For example, Jay Bruce (third best right fielder, going with Schoenfield) is worth just as much as Alex Presley (third in left field); or, even worse, Lance Berkman (second best first baseman after Joey Votto) counts for less than Chris Narveson (the fifth starter of the Brewers). I shouldn’t really have to explain what’s wrong with this, but in case you missed it, it treats every position as equal and every spot as equal. The third best first baseman is given the same weight as the third best closer, and so on. In reality, if Randy Wolf (best number four starter) is worth anywhere near as much as Andrew McCutchen (best center fielder), I will be shocked.

The next big issue is that Schoenfield tries to account for intangibles. Normally, this wouldn’t be too bad, except for two reasons. First, they’re applied in an erratic fashion. For example, the Brewers are ranked first in part because Prince Fielder leaving will give them “a chip on their shoulders”. Meanwhile, the Cardinals are ranked fourth partly because they lost Albert Pujols. In case you were wondering, losing a 5.5 WAR player (2011, by FanGraphs) gives you a chip on your shoulder, but losing a 5.0 WAR player at the same position is a huge deficit to overcome.

The other problem with the intangibles is that they are given just as much weight as the players themselves. So, going by this method, Prince Fielder leaving the Brewers (6 points in intangibles) will do more for the Brewers chances than Rickie Weeks (5 points) actually playing for them. Yes, you read that correctly. When teams are finishing 4 or 5 points apart in your system, but you’re giving teams 6 points for having players leave, you may need to re-evaluate what you’re doing.

Then, there are the matters of the rankings themselves. You can have minor quibbles, like whether Ryan Braun should really be in the first spot when it seems almost guaranteed he will be missing time (and possibly a third of the season, at that). I would argue that the top 3 spots at third base (Aramis Ramirez/David Freese/Scott Rolen) and right field (Corey Hart/Carlos Beltran/Jay Bruce) are probably too close to provide a meaningful distinction. I can see a case for Freese and Bruce being first if you like upside, or Rolen and Beltran if you like their chances to avoid injuries.

But really, the third starter spot is the only position that looks outright wrong. Schoenfield ranks Wainwright as the division’s third best third starter, claiming we shouldn’t bet on him being as good as he was before. Well, that’s fine I suppose, but two straight top 3 finishes in the Cy Young are pretty hard to top no matter what. I would say he’s still better than Shaun Marcum and Bud Norris, at least. Tommy John Surgery is a lot safer than most people give it credit for, I think. It’s easy to say that a player just won’t be the same afterwards, but recent players seem to be able to pick up where they left off fairly quickly. Just looking through Wikipedia, I found that John Smoltz, AJ Burnett, and Marcum himself all did fairly well coming back from the procedure. Jaime Garcia returned from Tommy John to finish behind Jason Heyward and Buster Posey in Rookie of the Year voting. Jordan Zimmermann put up an All-Star level year last year. Chris Carpenter returned to finish second in Cy Young voting, while Tim Hudson finished fourth the year he came back. Stephen Strasburg came back midway through last season, and proceeded to embarrass hitters and steal their lunch money (24 strikeouts! In 24 innings! And only 2 walks!). The point is, there are many cases to suggest that Tommy John surgery is no longer the huge, scary thing it once was. Now it just is, for most pitchers. There’s a good chance Wainwright will rank above some of the number two or one starters.

Any way, back on topic, none of these individual moves will cause more than one or two point swings, but add them up, and those 4 or 5 point gaps don’t look so solid. Also, when you actually start to weight the positions somewhat accurately, the Brewers come out looking even worse. They got a combined 24 points in Schoenfield’s process from their three-through-five starters and their closer spot, which is really sort of unrealistic. The Brewers were very good last year-as a Cardinals fan, I was more worried about them going into the year than the Reds. This year, though, the Reds seem to have filled most of their holes, while the Brewers just seem to have acquired them. This year looks to be another two-team race, but this time, I’m thinking it’ll be between Cincinnati and St. Louis again.