When I was five, you could have told me that Santa Claus and the Easter Bunny are real and I can meet them at the mall. I would have believed you. Now that I'm older, I'll still go to the mall to see Santa, but only if he has a friendly helper monkey with him (actually happened to me in college; the mall Santa had a monkey with him).
That's sort of how I am with those baseball intangibles. Once upon a time, I just accepted that the front office knew best or that pitching coaches held important positions. I may not have been sure how they did their jobs, but I knew they had to be effective.
Now, I question everything about that. Why did the team make a particular move? Can we see the results now, or do we have to wait three or four years? Do coaches really have a tangible effect on a team?
Of course they do. There are plenty of things in the game that can't be quantified. For instance, I know Jeff Bagwell did a great job as hitting coach, but I also know Sean Berry was good too. There are plenty of intangibles Bagwell brings to the dugout that don't show up on the box score, but that are essential to a winning team. I know all that.
That's a really roundabout way of saying I know that Brad Arnsberg made an impact in his first season as pitching coach under Brad MIlls. My question is how did he impact the team and what can we expect moving forward. I don't think I'll be able to analyze his whole tenure in this one article, so I thought I'd break it up into a series. From now until I run out of ways to look at the pitching staff, I'll write one article a week looking for Arnsbeg's fingerprints. We may not find anything useful; his impact could be more on the psychological side. I'm fine with that, but I also want to know. This week, I want to look at pitch selection and pitch values, which I'll do after the jump.
We know two things for sure about Arnsberg. One, he taught Roy Oswalt a new grip on his slider and Roy went on to have a bounce-back year on the mound. Roy credited him with it in spring training and often during the season. Two, he had Matt Lindstrom rely on his slider/sinker more, leading to a solid first half for the closer. Lindstrom admitted as much back in May. We also know that Brett Myers loves Arnsberg for his preparation and game plans, but that's a fact to be analyzed next week.
For now, let's delve into some pitch selection data. I went over to Texas Leaguers to get the Pitch F/X data from the past two seasons on eight Astros pitchers. My thinking with choosing these eight were that they were all guys who had been in spring training with Arnie, thus giving him the most opportunity to influence them.
Not all of these eight players have the best sample size, but they were the best bets out of the staff. Those eight are Roy Oswalt, Brett Myers, Wandy Rodriguez, Bud Norris, Felipe Paulino, Brian Moehler, Brandon Lyon and Matt Lindstrom.
I looked at each pitcher's repertoire from Texas Leaguers in 2010, getting the total number of each pitch thrown and a velocity. I did the same for 2009 and compared the two. As I suspected, velocity didn't change much at all and if it did, it was probably more a product of the sample size than of anything Arnsberg did.
What did show up was a big change in almost every one of these pitchers selections. Basically, Arnsberg emphasized more two-seam fastballs and cut fastballs. He also had guys use their sliders and changeups more often. Each time, it had the effect of balancing a guy's arsenal out, keeping a hitter on their toes as to what to expect.
That isn't rocket science, I know. Most pitchers throw a fastball about 50 percent of the time and it's not like Arnie changed that dramatically. He did even out Brett Myers, who ended up throwing about the same number of pitches in four different categories. He also had Bud Norris start using his changeup more and had Brandon Lyon rely more on his cutter and slider.
I'm sure he takes each pitcher's stuff into account when coaching them on what to throw. I mean, it's not like he taught them an entirely new pitch. But, I'm confident enough in the data to say that Arnsberg definitely has a philosophy of throwing two-seam and cut fastballs to hitters at the expense of a four-seamer. Whether that's from Mills or from Arnie, I don't know. It's good practice to increase ground ball rates, sure, and it doesn't always apply to those two pitches. With Felipe Paulino, Arnsberg seemed to emphasize his curve a little more than the four-seamer. Is that because Paulino couldn't throw a decent two-seamer or because he believed in his curve more? Again, I don't know. I do know there is enough evidence here to suggest a trend.
That's all well and good, but were his pitchers more successful this season than they have been in the past with their new arsenals? To find out, I went to FanGraphs and looked at the pitch value data to see how they did.
That wasn't a great method, though, because FanGraphs and Texas Leaguers have different pitch characterizations. In fact, FanGraphs doesn't have the granularity to the data to sort out a two-seam or four-seam fastball. At first, I looked just at the past two seasons worth of data for each of these pitchers. In that data set, there were some interesting jumps and losses. However, on a hunch, I included 2008 data as well and saw that the trend lines weren't as exact as a two-year window would allow.
For one, pitchers like Wandy lost quite a bit off their curve or fastball or slider, but if you look back another year, Wandy's curve wasn't that far off from his 2008 numbers. Same goes for Roy Oswalt, who saw a marked jump in his fastball effectiveness over last season, but didn't quite get to his 2008 effectiveness with the pitch.
There were some small things to glean from the data. If we assume that cut fastballs were sometimes lumped into the slider category on FanGraphs, then both Lindstrom and Lyon saw increases in value over the past two season in 2010. Lyon's slider/cutter jumped up almost 5 runs from where it was with Detroit last season and Lindstrom's reliance on that two-seamer/slider gave him his only positive value pitch in 2010.
Both Brett Myers and Bud Norris saw the effectiveness of their sliders rise, while Wandy, Paulino and Norris had increased success with the change. I know it's grasping at this point, since the data isn't conclusive in the macro-sense, but I do think that Arnsberg has helped a number of these players with their slider/cutter/two-seam fastballs. Oswalt indicated it was a grip he had shown them, and that may be the case. A slight modification of how they hold the ball wouldn't be hard to learn and could be repeated enough in spring training to carry over to the season. It also would account for the increased usage of those pitches.
Given this first set of data, it's safe to say Arnie did have an impact and it involved him teaching players a new way to look at the slider/sinker/cutter. Not only is that backed up by quotes from the players during the season, but we have statistical evidence to support it. Consider this Impact No. 1.