2010 Retrospective: David Aardsma

A review of Aardsma's 2010 season was partly done already near the end of July. As a summary of the above linked article, looking at Aardsma's pitch repertoire, pitch results and batted ball results showed little deviation from 2009 to 2010. What mainly went away was the good luck on keeping home runs in the park. Nothing else had principally changed.

With the season now officially closed, the comparison between 2009 and 2010 illustrates a bigger contrast between the two years. I will begin with pitch selection.

The unknown pitch is something in the two-seam/split-finger/change-up realm and differentiating between those are difficult even with the pitcher's help.

While Aardsma did not begin the 2010 season much differently in this measure, he sure ended it with a new strategy. Up through July 23, when I made the post above, Aardsma had gone with the heater 84% of the time. From that day onward, Aardsma threw a four-seamer just 66% of the time. That is a remarkably big shift in strategy. When exactly it occurred is hard to say. Near enough that I can tell, it changed over the All-Star Break. Was this a conscious choice? I believe so. Did it pay off?

That is a loaded question and truthfully one that I cannot truly answer because the data supports wildly different conclusions based on when you split 2010. If you split it at the July 23 mark then a downward trend in swinging strike rate on fastballs appears. Aardsma goes from 11% in 2009 to 10% in the first portion of 2010 and 9% from July 24 on. However, if you split it at the All-Star Break--a mere 10 days and only three appearances earlier--then the results look completely different with 9% in the first half and 12% in the second. Such is the nature of small samples and a biased cutoff date.

I do think the All-Star Break is the better splice point though and what I can say that is consistent across both cuts is that Aardsma threw his fastball for strikes more often with the decreased usage and that nothing in the makeup of the pitch (speed, movement, etc) dramatically changed. 

Examining David's other pitches is even more fruitless of a task given the absurdly small samples. I can tell you that his swinging strike rate on his slider dropped from 14% last year to 8.5% this year and that would seem alarming, but he threw just 82 and 80 sliders respectively. Whiff rate stabilizes quickly, but that's still too small to say anything of consequence.

Overall, Aardsma's swinging strike rate did drop nearly two points this season. His strikeouts fell by a similar amount and his walks tracked upward a fraction. Those are the bad signs. The good sign is that Aardsma sustained his regressed ground ball rate throughout 2010. From 25% of all batted balls a year ago, Aardsma kept about 35% on the ground this year. The line drives even dropped too. That helps to balance out the smaller strikeout per walk rate and we can see from xFIP and tRAr that 2010 Aardsma approximated 2009 Aardsma's performance quite well.

What changed in FIP and tRA were the home runs, which doubled in rate and ended up right at the league average of 10% of fly balls and 6% of all non-groundballs. Back in July, I noted that Aardsma's strand rate of 70% was poor for a high-strikeout releiver. He finished at 77.5% in 2009 and that's generally where we would have expected him to fall. Aardsma ends 2010 with a 77.4% strand rate. Regression can work both ways.

Aardsma's xFIP rose from 4.12 last year to 4.19 in 2010 and his tRAr dropped from 4.52 to 4.45. Both small changes and both figures paint Aardsma's talent level as roughly average for an American League relief pitcher.

That's not bad to have around, but not worth paying a lot of money for. With Aardsma's second 30+ save season and a good end to the year getting his ERA down to the mid-3s, I will be more surprised than not if come April 2011 David Aardsma is still wearing a Mariner uniform.