Why Has Javy Been So Bad?

Written by Josh W. | 26 August 2010

This year, Javy has a 4.96 ERA, a 5.43 FIP, a 4.82 xFIP, and a 5.32 tERA. In other words, he's been really, really bad this year. In fact, he's been so terrible, that according to WAR, he has actually cost the Yankees a tenth of a win this year.

Should the Yankees have known that he would have pitched this poorly? Well, it turns out that in 2009, Javy was pretty darn good. He had a 2.87 ERA through a whopping 219.1 innings. It wasn't just luck either; he had a tremendous 2.82 xFIP, which is really hard to do. However, this had been a career year, so naturally we would expect some regression. Here was how he was projected to do this year (before the year began), via Fangraphs:

There's a pretty high level of agreement here; all five systems see Javy as basically a 3.63 ERA pitcher across 205 innings.

So why isn't Javy pitching as well as his projections? An obvious explanation would be that his lesser performance can in part be attributed to switching from the NL East to the big, bad AL East. Does this claim actually hold true? If so, we should see a significant increase in the difficulty of batters faced. We can check this at baseballprospectus:

2009: average OPS of opposition: .731

2010: average OPS of opposition: .727

At first glance, it appears that Javy actually faced more difficult batters in the NL East. Yet we must consider the fact that 2010 has had a depressed runs scoring environment (some call it the year of the pitcher). In 2009, the league average OPS was .751, while in 2010, the league average OPS is .732. From this information we can tell that Javy faced below average hitters in 2009, and about average hitters in 2010. It seems that the slightly lower OPS of opposition in 2010 is a combination of a lower run scoring environment and better opposition canceling eachother out. Because these two variables just cancel eachother out, we still haven't found anything that would explain why his performance is different than his projections.

When we consider ballpark factors, we find that Yankee Stadium (NYS - New Yankee Stadium) has a park factor of 1.237 in 2010, while Turner Field has a park factor of .946. If I'm interpreting the ESPN park factors correctly, NYS is 30.7% more conducive to runs scoring than Turner Field. However, park factors are not that consistent, so I'm going to average the 2009 and 2010 figures to get better park factors. This makes NYS 1.1 and Turner Field .92, which seem like more reasonable figures.

If we divide 90 runs by two, we get 45 runs (you divide so you can apply the park adjustment to runs allowed at only the home ballpark). If we divide 45 by .92 (Turner Field adjustment), and then multiply by 1.1 (NYS adjustment), we get about 54 runs allowed. Add this to the 45 runs allowed at away ballparks and we get a projected total of 99 runs allowed. Now we divide by 1.08 to scale to earned runs allowed, and we get about 92 earned runs allowed. This would give us an ERA of 4.04 for Javy.

*this is assuming the averaged projections I used don't already account for ballpark.

The next factor is the drop in velocity. Based on the pitch f/x data at Fangraphs, we know that his velo has drop 2.1 mph, from 91.2 to 89.1. How much does this make a difference? Based on Mike Fast's research, we know that on average, a drop in 1 mph for starting pitchers means an increase of .25 runs allowed per nine innings. If we calculate this for Javy, and scale to earned runs allowed, we would get an increase of .486 earned runs allowed. Mike Fast also says that based on how much the pitcher depends on his fastball, the number can vary, with young pitchers being the most dependent on changes in velocity. Javy has not relied much on his fastball in recent years, so we will throw out a number of .400 earned runs for him.

This would bump his ERA up to 4.44. Interstingly, this is very close to his current rest of season ZiPS projection of 4.50. This still leaves a difference of .52 to account for (his current ERA is 4.96), and I believe this can be mostly explained by a loss of command.

Anyway, there was no way the Yankees could have predicted such precipitious drops in velocity and command, so it is likely that they were expecting a low 4's guy across 200+ innings, which would have been quite valuable.