- FanGraphs Baseball - http://www.fangraphs.com/blogs -

Toronto’s Alvarez Abandons Change, Succeeds Harder

See bottom of post for note on what is likely not an abandoned — but, rather, a distinctly harder — changeup.

Here’s an exchange that could very well occur between two mostly knowledgeable baseball fans:

Person No. 1: “A young pitcher whose fastball sits at 92 to 93 mph — and touches around 97 or 98 — struck out seven batters in as many innings yesterday.”

Person No. 2: “I have no reason to doubt it.”

And here’s another, also entirely possible, exchange on a similar theme:

Person No. 1: “Toronto right-hander Henderson Alvarez struck out seven batters in as many innings yesterday.”

Person No. 2: “Remove yourself from my sight, you gutless liar!”

One reason why that first conversation might exist is because many pitchers — like, Aaron Harang, to name one — have struck out seven batters in as many (or fewer) innings this season. Another reason is because pitchers who throw harder also tend to post both more strikeouts and lower xFIPs (a metric informed in no small part by strikeout rate).

The reason why the second conversation might exist is because Toronto right-hander Henderson Alvarez, despite possessing a fastball that sits at 92 to 93 mph — and which touches 97 or 98 — struck out seven batters in a game precisely zero times in his first 38 career starts. His 39th career start — on Wednesday night against the Yankees — marked the end of this streak, as the reader has every opportunity to discern from the following sortable table (which includes Alvarez’s top-10 starts by strikeout rate):


Date Opp IP TBF SO K% xFIP FIP
9/19/2012 @NYY 7.0 25 7 28.0% 2.35 1.10
7/16/2012 @NYY 6.0 25 6 24.0% 4.32 5.26
8/26/2011 TBR 6.0 25 6 24.0% 3.99 9.03
9/27/2011 @CHW 7.0 27 6 22.2% 2.21 3.17
8/31/2011 @BAL 8.0 26 5 19.2% 2.09 1.78
9/05/2011 BOS 6.0 22 4 18.2% 3.24 2.19
7/05/2012 KCR 5.1 26 4 15.4% 3.27 2.72
7/28/2012 DET 7.0 27 4 14.8% 4.07 2.81
8/10/2011 OAK 5.2 27 4 14.8% 3.78 4.97
9/10/2011 BAL 7.0 28 4 14.3% 2.85 2.31

So, this thing happened with Henderson Alvarez, and now some of us — and, by “some,” I mean, “like five” — are curious as to how or why it happened.

To understand that sort of thing, we begin — as one does in these situations — by turning our attention to Alvarez’s pitch selection, movement, etc., for his Wednesday start relative to the other ones he’s made.

Here we have Alvarez’s PITCHf/x data (pitch selection, along with horizontal and vertical movement) for the entire season:

And now here we have the same chart, just for his Wednesday start:

Here we find at least two points worthy of comment. First, there’s the matter of classifying Alvarez’s different fastballs. Some are classified by PITCHf/x as two-seamers; others, four-seamers; a few even, cut fastballs. As one can see from the bottom of the two charts, however, the labels given to the pitches do not necessarily reflect the way one might categorize them based on the clusters present — in particular, we find that there are pitches labeled as four-seamers, but which appear to be thrown more similarly to two-seamers. In practice, this is no great problem: for the purpose of this post, at least, the distinction is moot.

The second point is this: while the top chart indicates five different pitches being thrown, the bottom one clearly omits two of them. One of those, the cutter, constitutes less than 1% of Alvarez’s total pitches thrown this season, suggesting that it’s likely a fluke of classification. Alvarez’s changeup, though — a pitch he’s thrown ca. 11% of the time this season and threw ca. 16% of the time last season — is entirely absent.

In fact, upon closer inspection, we find that a trend exists. Here are the 10 career starts in which Alvarez has thrown the changeup least often:


Date Opp IP TBF FA% FT% SL% Other CH%
9/19/2012 @NYY 7.0 25 47.8% 43.3% 8.9% 0.0% 0.0%
8/31/2011 @BAL 8.0 26 83.5% 5.2% 6.2% 5.1% 0.0%
9/13/2012 SEA 7.0 29 50.0% 43.5% 5.6% 0.0% 0.9%
5/10/2012 @MIN 7.0 27 70.3% 5.9% 21.8% 0.0% 2.0%
8/14/2012 CHW 7.0 27 48.8% 33.3% 15.5% 0.0% 2.4%
8/19/2012 TEX 4.1 25 44.1% 28.6% 23.8% 0.0% 3.6%
8/09/2012 @TBR 4.2 27 71.1% 1.2% 21.7% 0.0% 6.0%
5/04/2012 @LAA 9.0 32 28.9% 47.4% 17.5% 0.0% 6.2%
7/28/2012 DET 7.0 27 53.9% 26.9% 12.5% 0.0% 6.7%
7/22/2012 @BOS 5.2 26 68.3% 8.7% 16.4% 0.0% 6.7%

Of the 10 starts here, two of them are from September of this year. Another three are from August. Two occurred in July, as well. All told, seven of Alvarez’s most recent 11 starts — again, out of 39 in his career now — are on this list. He’s throwing the change less, in other words.

The numbers suggest that the trend might be a smart one. Below are Alvarez’s lines for the 10 games during which he threw the change least often, and then the 10 during which he threw it most often.

Ten Games with Lowest CH%
66.2 IP, 271 TBF 32 K, 16 BB, 5 HR, ca. 3.75 FIP

Ten Games with Highest CH%
63.1 IP, 278 TBF, 39 K, 19 BB, 14 HR, ca. 5.40 FIP

The primary difference here is in home-run allowance — something that doesn’t become entirely reliable for pitchers even with 500 batters faced — so we should resist making any strong conclusions about it. Even with regression added in, though, there are more home runs being allowed. Are all those extra home runs coming off the changeup? No. Of the 26 home runs allowed by Alvarez this season, for example, only three (or 11.5%) have been hit off the changeup — a figure almost precisely the same as Alvarez’s usage.

Of course, mysteries abound with regard to pitch sequencing, so none of this strictly rules out the possibility that throwing the changeup has somehow led to home runs. In any case, the pitch has been his least valuable this season on a rate basis (-2.21 runs relative to average for every hundred thrown) by some margin, and we’ve just witnessed Alvarez’s best game coincide with a complete suspension of the pitch.

What’s unusual about Alvarez’s lack of success with the changeup — and, on Wednesday night, the amount of success in its absence — is that the pitch was so roundly praised before the right-hander’s major-league debut. There’s more than one available scouting report that refers to the pitch either as “plus” or even “plus-plus.” In practice, that hasn’t proven to be the case: Alvarez has induced swinging-strikes on only 5.8% of changeups this season, while an average figure would be roughly twice that; an elite one, about four times as much. Moreover, Alvarez has been able to induce a significant percentage of grounders (just 47.7%) of batted changeups, while throwing the pitch for a strike about 13 percentage points below average. Indeed, the pitch could very well have been a plus at some point. For the moment, however, suspending its use seems like an entirely reasonable decision.

*****

As a couple of commenters have noted, it appears that, rather than abandon his changeup, Alvarez threw it much harder than usual on Wednesday — so much so that it appeared like a different pitch. Alvarez has generally thrown the pitch at 85-86 mph. This graph, relabeled in the most amateur of fashions, indicates a changeup more in the 87-91 mph range: