May 23, 2014
This Week's New Pitching Lines, 5/23
Welcome to what will likely be the penultimate This Week's New Pitching Lines! This week's is pretty fun. Glad you both made it today.
Best new line of the week: Chris Sale's 6/1/0/0/0/10. While doing this, I often ponder which lines are the most tied to this era; which lines have the greatest ratio between post-2000 frequency and pre-2000, or pre-1990, or pre-1970 frequency. I've narrowed it to one of two categories. The first is the five-inning start. Wily Peralta, for instance, this week broke in a new pitching line: 5/9/3/2/4/5. Now, he's the first person to produce that line, so by definition it can't be that common these days. But compare it to Drew Smyly's line this week, which is very similar to Peralta's: 5/7/3/3/5/6. Almost the same, really. Five innings, bunch of hits, not many runs, bunch of walks, bunch of strikeouts. Smyly's line had never been produced before 2007, when Kip Wells debuted it. But now it's almost an annual event; Smyly is the second pitcher to produce that line just this year, in fact. I'd bet anything that Peralta's unique line won't be unique within a couple years, and that we might have 10 of them by 2030. (Excepting the unearned run, at least, which wasn't a factor in Peralta's new line being new.)
The other most-2000s line is the six-inning, eight-strikeout, no-walks game. This isn't what Sale did. Sale did much better. But the idea behind it is similar: Pitcher pitches very well, gets loads of strikeouts (because strikeouts are cheap these days), but gets pulled well short of a complete game (because pitch counts, because bullpens, because of the multiple-times-through-the-order penalty, etc.). So here's how common these games are now, relative to before now:
If you include 5 1/3 and 5 2/3 inning starts in the accounting, it's 52 percent. So I'm going to call this the most modern pitching line. Call it the Phil Hughes. Call it the Dan Haren. Call it the James Shields.
Worst new line of the week: Scott Kazmir’s 1.1/1/1/1/3/1. This is an artificially unique line. Kazmir was ejected. Most pitchers who have this line get to keep going, so that they soon don't have this line. But Kazmir's day was ended against his will due to an external force. This is no guarantee of newness. Matt Cain left a start with a hamstring injury this week, and his 3/0/0/0/0/3 line seemed certain to be new; what could possibly prompt a pitcher to get pulled from a start after just three innings with a perfect game going? And yet, somehow, it's not unique; it's a third-timer! This sort of surprise repetition happens constantly when I look up these games. But ejections, man. Ejections are so random and so rare. So how often do ejections produce new lines? A quick look at the 2013 ejections:
This gives you a pretty good idea of how often everything has happened. Even with the intrusion of a random disruptive element, we saw nothing new. All repeats. All pretty common.
Most surprisingly new line of the week: R.A. Dickey’s 6.1/3/4/3/3/3. Ignore the unearned run; it's a macguffin, because 6.1/3/4/4/3/3 would also be new. Move any one of these numbers up or down by one (or, in the case of innings, 1/3) and you'll find a predecessor, except for removing one walk. Everything else has been captured; this line is thoroughly surrounded. What's truly fascinating to me is this, though: If you adjust Dickey's line by a single hit upward--6.1/4/4/4/3/3--you get only one result: Dick Lange, June 1974. And if you adjust Dickey's line by a single hit downward--6.1/2/4/4/3/3--you get only one result: Dick Lange, July 1974. WHAT THE HECK BASEBALL?
Stephen Strasburg line of the week: 7/6/2/2/1/4. Well this pretty much kills the Strasburg-lines-are-all-unique-because-Strasburg-is-unique conceit of this series. There are very, very few lines more common than Strasburg's line this week. It's a 93rd-timer. That's not the most common line, but it's close, and raises the question of which is the most common line. I can't answer that, precisely, but what I can do is go step by step to get close to the most common line: The most common innings total is nine; the most common hits total for nine-inning starts is six; the most common runs total for nine-inning, six-hit games is one; the most common walks total for nine-inning, six-hit, one-run starts is one; and the most common strikeouts total under these circumstances is three. There are 128 starts in history with a 9/6/1/1/1/3 line.
But complete games were such the norm for so long that they skew everything. What if we do the same process but ignore complete games? Step by step we go, until we end up with this as our most common pitching line: 7/6/2/2/1/4. Would you look at that. Strasburg's line. Literally the commonest line you can come up with. 93 times. Crazy. So, yeah, no more Strasburg line of the week. Maybe we'll switch to Zach McAllister, who produced two new lines this week, and has produced new lines with five of his 10 starts this year. But he's so booooorrring.
Asterisk denotes a line that is unique only because of the classification of some runs as unearned. All lines are compared only to previous starting pitching lines; reliever lines are not considered.