Arizona Diamondbacks starters: going deeper, or just not as short?

By fWAR, the Diamondbacks currently have the best rotation in the major-leagues. This is quite a turnaround, considering that last year, they ranked 21st by the same metric. It’s even more striking, given the lack of changes this winter: Taijuan Walker replacing Archie Bradley, is the only significant alteration from the rotation last year. [90% of non-Archie starts in 2016 - all of them bar fourteen - were by people who have also started games for Arizona this year]

One of the reason for this is sheer volume of work. At 190.1 innings, it’s the most by any National League team, and just two outs behind the major-league leaders in Houston. If this pace is maintained, our rotation will post 934.1 innings this year, fifty more innings than last season. They’re also on pace to strike out 967 batters; for K’s by Arizona starters, this would trail only the monstrous 2002 season. That’s the year Randy Johnson and Curt Schilling both passed 300 strikeouts, which remains the all-time high for K’s by any rotation.

Here’s a comparison of the stats from 2016 and 2017 for our starting pitching, each through the first 33 games of the campaign:

This shows numbers trending positively almost across the board. We’re averaging an out per game more, while allowing fewer hits - particularly home-runs. Strikeouts per start are more than one-third up on 2016, and ERA has plummeted from 5.20, all the way down to 3.69. About the only black-mark has been against the defense more than the pitching staff: unearned runs allowed by our starters are up 60% over the same period as last season. I haven’t got fielding game logs, so can’t see what the total was to this point in 2016, but we’re currently on pace for 108 errors, which would be Arizona’s worst tally since 2009.

However, it’s worth noting that the increased length from our rotation is due almost entirely to one things - the absence of the “disaster start”. If we break down the first 33 starts this year and last by length, here’s what we find.

Once you get past the fifth inning, the numbers are almost identical between 2016 and 2017. It’s a decrease in the really short outings which has been the main driver to increasing the total number of innings pitched by the rotation. The graph above, if anything, undersells this aspect, because all four of the “4 innings or less” outings this season went exactly four innings. Most of the seven such starts last year fell short even of that mark: three by Shelby Miller (1.2, 2 and 3.2 innings) plus one each from Robbie Ray (3) and Rubby De La Rosa (3.2). That’s five times in the first 33 games, our starter couldn’t even get through four innings; we haven’t seen that at all so far in 2017.

The impact is not limited to the rotation, it also helps the bullpen out. This is apparent most dramatically in the sheer number of relief arms we’ve needed to use. Since Opening Day 2017, the only additions to the bullpen have been Silvino Bracho and T.J. McFarland. Over the same period last year, the D-backs went through some incredible roster instability, bringing up nine relievers at one point or other: Enrique Burgos, Matt Buschmann, Zac Curtis, Rubby De La Rosa, Kyle Drabek, Keith Hessler, Dominic Leone, Evan Marshall and Tyler Wagner. Most are names you’ve likely forgotten, and with good cause.

While the number of long (> 3 outs) relief appearances has been the same this year and last, in 2016, Chip Hale seemed to treat those relievers very much as a “churn and burn” utility. Bring ‘em up, throw them into mop-up duty, then ship them out for a fresh arm. This year, Randall Delgado and Archie Bradley have executed the majority of them, and it feels like their appearances have been much more a conscious choice, rather than a case of emergency use. Again, the lack of disaster starts is likely helpful. As we saw with Braden Shipley and Patrick Corbin’s outings, if your starter gets through four, then your bullpen can usually survive, especially on the road.

If we look at length in pitches, we see something similar. Only three times has a starter not reached 80. One was Miller’s final outing, and another a Corbin start that lasted six IP, ending because his batter’s spot came up in the 7th, with Arizona down two. The only “true” sub-80 pitch start was again Corbin, on Saturday in Coors. But even including all three, the only NL teams with fewer are the Nationals + Cardinals. Conversely, we have 13 starts lasting 100 pitches or more, behind the Nats’ 21 - the only team averaging six innings per start. At least in our case, we have a high K-rate partly responsible. In Washington, it seems more a case of Baker gonna Baker...

Certainly, if the team is to compete this year, the improvement is going to need to continue coming from the rotation. Win or lose, having your starter going close to six innings allows you to line up your bullpen for optimal efficiency, allowing you to save your best relievers for the most important innings and games.