Looking Forward to: Chien-Ming Wang

The general consensus is that Wang is a very good deal for the Nationals. He once was a decent mid-rotation pitcher, and his 2009 season looks like an aberration, possibly the result of some injury troubles. (You can find a very good summary of all this here.) Consequently, he projects none too badly, with a Marcel* ERA of 4.86, as the worst case.

We can use some new-fangled sabermetrics to take more measured assessment, perhaps. One of the problems with the widely available projections is that they are mass-produced. They take a bunch of assumptions and throw them at all players indiscriminately. There is nothing wrong with this. Bespoke projections would take a lot of time, and not actually add very much to the accuracy. In the case of someone like Wang, though, it's worthwhile to dig deeper.

Last year, David Gassko of The Hardball Times introduced something he called Luck-Independent Pitching Statistics (work out the unsightly acronym for yourself). I'm going to look at Wang using some of Gassko's ideas, with a leavening of my own 'keep it simple, silly' mentality.

What I'm going to do is to use the ratio of different types of outcomes to a plate appearance. I'm limiting myself to balls hit on the ground, balls hit in the air, line drives, strikeouts and walks. Enough work has been done to tell us what a given batted ball is worth. Using this data, and imposing it on a pitcher's own career patterns, can give us some idea of what is likely to happen.

First of all, what patterns can we deduce from Wang's career so far? The interesting ones for our purposes are a steady fall in groundballs in play, a steady rise in strikeouts and walks. His line drive rate started low and has risen to around the average. His outfield fly ball rate has been in the low 20s except for 2009. My educated guesses based on the data is that these trends are going to produce in 2010 something like:

LD%      GB%      OFFB%      IFFB%      K/9      BB/9      
18.8    52.0      22.4        6.8       7.7       4.9
(MLB 09 
median)

For the strikeouts and walks I used the average rate of change on a year-to-year basis and added it to the 2009 ratios. That may overestimate things, especially if Wang alters his approach under a new pitching coach. That's an eye-watering BB/9, and I won't be surprised if he does.

To translate that into something more meaningful in terms of traditional statistics, we need to take a guess as to how much playing time Wang will get. I've assumed 15 starts, which converts on the basis of his past seasons to about 97.3 innings pitched. His worst-case projected Batting Average Against is .279 (Marcel again). That altogether means:

In-Play Outs     Strikeouts     Walks     Non-HR hits     Home Runs
  209                83           53         93              10

Yesterday you caught me using Runs Created, but this time I'm going to use a different run estimator, because it's better. This is BaseRuns. Converting those totals of Strikeouts, Non-HR hits, etc into runs, we get an estimate of 57 runs allowed by Wang, in 97.3 innings.

Which works out to an ERA of 5.28. Ugh.

Riggleman's comments are interesting. He's basically saying that he anticipates Wang to be a good fall-back option if Balester and/or Mock don't turn out well. Balester's 2010 Marcel projection (since that seems to be the projection du jour) is for a 6.72 ERA. Mock comes in at 4.81, but his projection doesn't show a full-time starter's innings, and relievers generally have better ERAs than starters. Mock's starting ERA might be a lot closer to 5.28 than 4.81.

Wang's an interesting gamble, but he's still a gamble. If he can cut his walk rate, the Nationals might win the gamble. Otherwise, it's more of the same again in NatsTown.
________

* Marcel is a very basic projection system, monitored by TangoTiger of The Book Blog.