Let’s have some fun with numbers, shall we?*
Here are three NL starting pitchers and a collection of their stats so far in 2013; they are joined by a common thread – they were all considered by fans and pundits alike to be good fits for the last spot in the Washington Nationals starting rotation coming into this season. You’ve read the title of my post, so you already know who one is – Dan Haren. OK, so the other two? One is former Nat and current Chicago Cub contract albatross Edwin Jackson and the other is former free agent siren, current Milwaukee Brewer, Kyle Lohse.
Now another question – who’s who in this table (courtesy of Fangraphs)?
The first stat line belongs to Haren, the second to Jackson, and the third is Lohse’s; overall, we see three relatively well touted back of the rotation guys doing what back of the rotation guys do – eating innings, doing so with occasionally shaky peripheral stats. What we also see is that, in spite of the cries of many Nats fans to designate Haren for assignment or banish him to the bullpen in favour of a more productive pitcher (albeit one without the luxurious pelt of Haren’s), the other options available in the offseason would have provided essentially the same results – underwhelming ones. Throwing salt into this wound and confirming the ‘meh’ of this trio’s output, Haren’s average game score of 44.6 ranks 110th out of 122 starting pitchers who have at least 50 IP so far in 2013, with Jackson’s not too far behind at 114th (44.3 average game score), and Lohse leading this pack with a 51.1 average, good for 75th.
Essentially, the Nats and their fans are stuck with Haren and his $15 million highwire act. With that reality posited, it’s best for all involved to accept that attention must be turned to figuring out how to get 2013 Haren more in line with the productivity seen over his career:
Comparing this table to the first shows some promise – 2013 Haren is really not that far off of the mark with a number of his rate stats. He is striking out and walking people at the essentially the same rate as he has his entire career, and the same can be said for his linedrive rates (LD%). The real buggers here are his propensity to give up more flybballs (FB%) in 2013, which leads to his propensity to give up the longball (HR/9 and HR/FB).
So we have a ‘well duh’ moment – much of what Haren is suffering from arises from his inability to keep the ball in the park. While much has been made of Haren’s declining fastball velocity as being a large part of the issue, I won’t discuss it here, as I feel that his velocity has been on the decline over several years, so much so that it plays an ever shrinking role in his problems, as he has had time to make adjustments to compensate for declining velocity. So with that caveat out of the way, let’s take a look at some other details of Haren’s 2013 season and his career averages, my fellow baseball pathologists:
Here, we have pitch types and percentages – FA is four-seam fastball, FT is two-seam fastball, FC is cut fastball, FS is split finger fastball, and CU is curveball. While Haren does have other pitches in his repertoire, these are what he’s using thus far in 2013, per Fangraphs/PITCHf/x data. Right away, we see that Haren’s 2013 is marked by change – not only the lack of a changeup, but also a drastic difference in how frequently he throws certain pitches. Known the last few year for having his success from cutting and sinking his fastball in order to set up one of the better split finger fastballs, we see in 2013 he has pretty much, now relying even more heavily on his split and all but abandoning his curveball as well as his slider and changeup. In many respects, Haren is slowly becoming a two pitch pitcher (fastball/split), with both pitches seen in the 85-89 MPH range. Not the greatest recipe for success, as most starters need that third pitch to be effective and get hitters out, especially when facing them a third and fourth time. However, his split is good and his fastballs have a reasonable amount of movement that it *could* be enough to be effective.
Let’s look at his PITCHf/x pitch values for the pitches he’s using in 2013 – briefly, pitch values are a way to evaluate the effectiveness of a pitch; a positive number is good, and negative one is bad. There are many caveats to using these numbers, but for our discussion, they help add to the gist of the matter.
By the looks of it, Haren’s best weapon in 2013 is his two-seamer and split, with his cutter being his least effective pitch. Compare this to the previous table and we see Haren is relying heavily upon a pitch that is probably his least effective one, the cutter (FC). We do, however, see Haren use his best pitch, the split, twice as often as he has previously in his career, so it isn’t all bad in 2013. In general, these two tables show us the maturation of Haren as a pitcher who must rely on guile and movement of pitches as well as a pitcher that understands that he must continue to evolve his pitch selection and approach.
So we have a reasonable idea of what Haren has and what he’s doing with it – so how does that play into the results seen thus far? Picking at the numbers, Haren isn’t atrocious and much of what he has done appears to be easily remedied with a few changes in how he attacks hitters. While the overall stats don’t show it, Haren has had a handful of games where he pitched more than capably, highlighted by his 8 innings of one run ball May 2 against the Atlanta Braves.
So what gives?
Let’s shift gears a bit here, and look at some data from Baseball Reference that shows how progresses during his appearances in 2013. These next two tables show essentially the same thing, just broken down into different forms, the first by number of times through the opponent’s lineup and the second broken down by pitch count:
|1st PA in G, as SP||13||123||117||6||2||29||14.50||.274||.289||.496||.785||.313||80|
|2nd PA in G, as SP||13||113||105||5||4||16||4.00||.305||.342||.495||.838||.321||96|
|3rd PA in G, as SP||13||80||75||6||4||12||3.00||.333||.375||.640||1.015||.333||133|
|4th+ PA in G, as SP||3||5||5||0||0||2||.400||.400||.800||1.200||.667||170|
And there lies the rub. Haren’s stuff still remains effective, but under a finite number of pitches or number of times through the opponent’s lineup. Once he hits about 60 or so pitches, roughly the third time through the lineup, the wheels fall off for Haren, and his inability to throw an effective third pitch to keep hitters off-balance and not sitting on his fastball or split becomes exposed, typically in the form of a big inning. While this is troubling enough for Haren, it is compounded by the call up of Ross Ohlendorf from Class AAA Syracuse and his fantastic first outing as a National; now, it can no longer be assumed that patience will be lended to Haren, should he suffer another shaky outing.
So how to make lemonade out of this lemon? What can be done to make the most of Haren’s fleeting but desirable effectiveness while also maximizing the productivity out of back end of the rotation?
While lineup and pitching rotation optimization is still considered a bad word in many baseball circles, the possibility of a fifth starter by committee could provide a method of making the most of Haren’s good innings, while also lessening the possibility of damage from a big inning – quitting while you’re ahead, figuratively and literally. With Ohlendorf now slotted into the long relief/spot starter role in the Nats bullpen, it would be an interesting experiment to slot both Haren and Ohlendorf for a discrete number of pitches for each of Haren’s starts, a la the 2012 Colorado Rockies, giving Haren his starts, but keeping him to his 60 or so effective pitches, with Ohlendorf coming in to take care of the rest.
It is an extreme alternative, but considering the extreme home run rates and loss of effectiveness in the later innings arising from Haren’s outings, it’s one that could give the Nats the shutdown, end of the rotation starter that they thought they had after signing Haren in the offseason.
*I know, I know, there’s a first time for everything.