Are short outings an issue for Diamondbacks starters?

The Diamondbacks bullpen has already had to work harder than that of any other team in the major leagues, Coming in to yesterday's game in Pittsburgh, our relievers had thrown a total of 407.1 innings, which is 11.2 more than the second most-taxed bullpen, the Colorado Rockies. While they've stood up to the workload fairly well, with a season ERA of 3.36, good enough for sixth in the National League, this would seem to suggest Arizona's starting pitchers are not going deep enough into games. But are things quite as obvious as that?

Extra innings, brought to you by....

It certainly isn't fair to put all the blame on our starters. Because a significant factor in the workload is the number and duration of extra-inning games played this year. Here are the five longest games by time, played in National League ballparks this season:

1. July 19, NYM @ STL 18 innings, 355 minutes
2. May 31, ARI @ MIL, 17 innings,349 minutes
3. August 18, ARI @ PIT, 15 innings, 311 minutes
4. July 17, SFG @ ARI, 12 innings, 311 minutes
5. July 11, STL @ PIT, 14 innings, 304 minutes

Yep. The Diamondbacks have been involved in three of the four longest, and have played sixteen extra-inning games overall. Only the Pirates (17) have more, and is six above league average. As a result, some of the workload is because we've seen more than most in the way of extra-innings. In total, our bullpen has had to work through 166 unscheduled plate-appearances. That's a lot more than the rest of the NL West, none of whom have even reached a hundred extra-inning PAs, with the number ranging from 94 for the Padres, all the way down to just 37 for the Rockies. You can't say our starters are responsible for that.

A plethora of former rotation members

One thing which has helped, is the presence of arms who are former starters, capable of throwing multiple innings. Josh Collmenter was our 2015 Opening Day starter; Andrew Chafin pitched in the rotation virtually his entire career prior to this season, and Randall Delgado through the end of 2013. As a result, while we lead the majors in innings, we are fifth in the NL for relief appearances; Braves have come in from the bullpen, for example, two more times, but have thrown 61 fewer innings than Arizona. That's because our pen are masters of the long game: we have 90 appearances recording four or more outs, compared to an NL average of only 66.

This matters, because it seems that for bullpen arms, getting the time to recharge between appearances is more important than the length of the outing [I read something to this effect recently, but damned if I can now find it!]. Despite the number of innings, Chip Hale appears to have done a pretty good job of managing his way through it; While 80 of our relief appearances have come with the pitcher concerned on zero days' rest, that's actually below league average (83). Only eight times this season have pitchers had to go back-to-back-to-back, and three of those belonged to LOOGY Oliver Perez.

The incredible shrinking start

Declining offensive numbers should make us readjust out expectations e.g. the average major-league hitter this year is hitting .254. And the same is true for pitching. There have been only 27 complete games thrown in the National League so far; just four years ago, there were 80. That's the sharp end, but the truth is, most starters get pulled before the end of the sixth. The average start in the National League this season has lasted 5.87 innings and 93 pitches. There is precisely one starting pitcher in the league who is at seven innings per start (Clayton Kershaw, naturally) and only thirteen - so less than one per team - averaging an out into the seventh or beyond.

The days of expecting a normal starter to go seven innings and toss 110+ pitches should be as dead as the 4-man rotation. Sure, these games will still happen. But they should not be considered as the "norm". It is true that the D-backs are a short of the average, at 5.67, but that works out at one fewer inning from the rotation every five games. So, while there's certainly room for improvement, it doesn't seem like it should be a overwhelming problem at the team level. And it doesn't appear to have hurt the bullpen performance, at least thus far.

Dissecting the D-backs

However, let's dig into things a bit further. Here are the numbers for all our ten starters used by Arizona, this season to date.

Some interesting numbers to chew over here, though it's worth noting the overall figures are likely skewed by things like the pitch-count limitation on Corbin, the bullpen start for Hudson and the more or less unadulterated suck of Webster and Bradley in terms of starting stamina. If you look at our current rotational five (including, for now, Anderson), the average innings per start is 5.82, just four-hundredths of an inning per game below the average - and that's still including Corbin's limitations.

Surprised to realize that De La Rosa has actually done best at working deeper into games. That's mostly because he needs fewer pitches to get through the batters he faces: Rubby averages 3.67 pitches per plate appearance, compared to team and league averages of 3.86 and 3.78 respectively. Corbin's number of 3.61 is better still, but save for Webster, all the others are above league average, going up to Ray at the top with a 4.07 P/PA stat. That's why his starts are our longest by pitch-count, if not innings, and the one we saw from him yesterday (not included in these stats( exemplified this: six innings, 99 pitches, twenty-six of them coming in the first frame.

Digging a bit further into Ray's stats, the figures seem to bear out what we've seen in his starts. While the number of strikes is perfectly fine (at 63.9%, virtually at league average of 64.5%), a lot of Ray's pitches end up getting fouled off for a do-over. That happens to almost one-third (32.4%) of his strikes; for comparison, only one of the 60 currently qualified pitchers in the NL has a foul rate for 2015 which is as high. On the other side, the percentage of swinging strikes Ray gets is below average; if our young pitcher can develop and/or use his repertoire in a way which shifts that balance a bit, it should help him pitch deeper into games.

Conclusions

Yes, the Diamondbacks' starters aren't going quite as far into games as some. However, this is most likely the result of what we knew coming in: the rotation was a cause for concern and work in progress. The current incumbents are now very close to league average in this department. Also, this may have been a factor in bullpen construction, including multiple players there who can give us extended outings. Chip Hale has done a good job of spreading the relief workload thus far, and the relief corps has been mostly solid. Through three-quarters of the season, there's not much evidence the length of outings produced by our rotation has been a significant problem.

[All stats through end of play, August 18]