Washington Nationals Stat shot: What's the deal with Stephen Strasburg?

Stephen Strasburg is a good pitcher. He is probably a great pitcher, and has had plenty pre-season buzz about All Star, Cy Young, what have you. And yet, he hasn't seemed so sharp so far this season. Yes, I know, it's only three starts, two against serious post-season contenders, but the Nats have only won 1 of his 3 starts so far. Why? Well, let's have a look at some of his top-line stats and diagnostics so far this season:

(All tables in this article via Fangraphs, click to embiggen)

First of all, I will again admit that Stras has hardly pitched at all--only 18 innings and less than 100 batters faced. Any trends I try to pick out in this article could be unlucky random variation or some early-season issue that will get worked out as he "settles in." Second, our #37 is putting up overall performance at about the same level as 2010 and 2012, with a 77 ERA- (remember, lower is better with the "-" stats, since they're directly proportional to the number of runs a pitcher gives up, with 100 being league average... Stras gives up only about three-quarters as many ER as a league-average pitcher). So, why is he only 1-2? Well, the 2 unearned runs (8 R allowed vs 6 ER) cost one game. Without Zim's unfortunate throw on Saturday, maybe I'm not even making this post.

Still, there's a little cause for concern, in that ERA- is much lower than his xFIP-. This means that he's gotten a bit lucky with how many HRs he's given up so far this year. And if we look at HR/9, that's just what we see: only 5.9% of his fly balls have gone out so far, versus about 10% in 2012 and 2010. Once that regresses and Strasburg is giving up dingers at twice his current rate, the ERA- is going to climb. What's missing? Is it luck? Well, Stras' LOB% is a bit low at 69.4%, but that's not too much worse than his career numbers. Likewise, his BABIP-against is only .281, which is a bit better than his career average. He's getting more ground balls than average (56%--almost 10 points better than his career average!). He's also not walking too many people, as his BB/9 is on part with his career average. Where's the disconnect?

Could it be, pitching to contact? Take a look at that K/9. In 2010 and 2012, Stras struck out well over a batter per inning. That strikeout rate has fallen from more than 11 K/9 last season to only 7.3 K/9 so far this season. He's only striking out about two-thirds as many batters this year as last. It's right there in his rate of swinging strikes: batters are whiffing on 8.8% of his pitches in 2013, versus 11.2% last year. It may be that Stras is making a deliberate attempt to reinvent himself as a groundball pitcher, rather than a flyballing, high-strikeout pitcher. Let's take a look at his plate discipline numbers:

Did some say P2C? In spite of throwing close to a career-low number of pitches in the zone (44.1%), he's getting the most contact against him in his career, about 80% vs. 75%. Oddly, his contact rate in the zone is about the same, while the contact rate outside the zone is up almost 15 points! Batters are swinging against him at about the same rates as always, both in and out of the strike zone. Why are batters getting on more balls out of the zone? Is it pitch selection, or speed?

Stras' 2013 pitch distribution is almost identical to 2012's, and so is his velocity--he's a tick slower on his curve, but that should only help induce misses, given that his FB is just as fast as ever. (Note that Pitch F/X's characterizations are slightly different--it splits the FB into a 2-seam and a 4-seam, and claims he's throwing the same number of FBs overall, but a much higher proportion of 4-seams vs. 2-seams.)

Okay, what sort of results is Stras getting on his pitches? Let's take a look at pitch values, both in terms of total runs above/below average and runs above/below average per 100 pitches (the "/C" columns).

The fastball and change are both plus pitches, like before, but... dat curve! It's gone from a plus pitch to a minus pitch! Let's take a look at how Stras' curve has been doing this year. This table shows AVG against and wRC+ against (wRC+ scales to total offensive output, where more is better and 100 is league average) on Stras' curve:

Well, that seems to illustrate the problem: Stras' curve has gone from essentially unhittable to a BP pitch. A batter against a Strasburg curveball this season generates 50% more offense than average! What the heck? I took a look at a few things to try and get a handle on this. It quickly dissolves into a bigger pile of eye charts and tables, but the conclusions about the curve seem to be these:

• It's moving as much or more as ever (8.7 inches horizontal, -7.3 inches vertical this season is actually more break in both directions than career average).
• It's in the zone at career average rates (about 49% of the time).
• Batters are swinging at it less (about 10 percentage points less than career average both in and out of the zone), and making more contact in the zone but less out of the zone (by about 9 points in each case).
• This is based on a total of 57 pitches thrown (against 11 batters) and is probably a way-over-analyzed, small-sample size, random fluctuation.

Bottom Line

Okay, is it time to start panicking about Strasburg? No, although the slow start isn't exactly encouraging. This "trend" may disappear as we pile up a more significant stack of pitches thrown and batters faced. Here's what to watch going forward:

• Strikeouts. Will Stras' K/9 settle out at more or less than 9?
• P2C. Will the GB% stay high? Will BABIP be low, signifying weak contact (the "good" P2C)?
• Dingers. The HR/FB rate tends to end up around 10% for everyone, making every walk and seeing-eye single hurt that much more. Can a new "P2C" Stras depress the HR rate with his increased GB%?
• Curveballs. We're used to a guy with three plus pitches. Two plus pitches is still pretty good, but it's not ace-worthy. Again, the apparent "minus" curveball so far this season might be an illusion, but keep an eye on it.

(Updated to fix a few typos and introduce some new ones, probably.)