Are the Oakland A’s using their starting pitchers differently?

In a column on Wednesday, Eno Sarris of the Athletic pointed out that offense is down so far in the 2020 season. He explored several factors, but one in particular was titled “pitching strategy is being optimized.”

Within that section, Sarris illustrated that starting pitchers are not as often being asked to face opposing lineups a third time in a game. Limiting repeat exposure in that way is not a new idea and the league-wide trend has been moving in that direction for a while, but it has accelerated greatly in 2020. He included a relevant quote from the Oakland A’s own Matt Chapman:

“... [W]hen you haven’t seen guys — there are a lot of new faces on a lot of these teams we’re playing — sometimes it just takes at-bats to get used to it. I definitely feel more comfortable at bat after I’ve seen someone a few times. Even when we played the Angels, it seemed like they were throwing out different relievers every inning, so nobody really got to see anyone twice.”

All of that makes sense, and the data clearly shows the recent shift. However, that’s a big-picture look at the whole sport, and individual teams could vary in their behaviors. How are the A’s using their own starters so far this summer?

First, some background. As you might expect, Oakland was on the strategic forefront as usual, in terms of avoiding the third time through the lineup. In 2018, they were famously among the first to try out the bullpen “opener,” whose purpose was related to this concept, and it (arguably) helped them maximize a ragtag group of under-the-radar arms into an effective unit. In 2020, though, the A’s have a much stronger rotation, and thus less need for gimmicks.

There are two ways to begin looking at this numerically. One is the percentage of games in which the starter was allowed to venture into the third time through at all (binary: yes or no), and the other is what percentage of plate appearances were made by hitters facing the starter for the third (or more) time in a game (note that this leaves out pinch-hitters who subbed in for that third plate appearance).

Here are the last three years of the binary question of how many of the 162 games included the starter facing at least one batter a third time:

• 2017: 88.8%
• 2018: 80.8%
• 2019: 90.7%
• 3-year total: 86.8%

* Note: Some of the 2018 “starts” were by one-inning opener relievers, so a few more games may have included a long man clearing the 18-batter threshold.

And the last three years of the percentage of plate appearances (against starters only) that were that batter’s third (or fourth) time facing that starting pitcher:

• 2017: 24.7%
• 2018: 21.1%
• 2019: 23.0%
• 3-year total: 23.2%

* Notice the outlier 2018 on both lists, the Year of the Opener, when the A’s went hard in terms of pulling starters early; that also accounts for some of the ‘18 difference in this second list, since no one-inning opener ever had a chance of contributing in this department but they still added PAs to the total

** More context: In 2013, the rates were 96.2% and 28.6% ... in 2010 they were 95.7% and 29.8%. ... in 2005 they were 96.2% and 30.5%

As for 2020, things get a bit trickier. Due to the unprecedented delay of the season and the short preseason period, starters were uniquely not ready to go their full normal workloads on Opening Day. Therefore, I’m removing the entire first turn through the rotation, because it’s skewed data — the starters were pulled early (which supports the league trend!) but not for the relevant reason (it was for health, not matchups/performance).

Starting July 29, with the second turn through the rotation, and also removing Jesús Luzardo’s first start (same principle; he was still ramping up), here are the 2020 numbers:

• Begin 3rd time thru? 83.3%
• Rate of PAs as 3rd/4th of game? 18.3%

This covers a dozen games. The only two in which the starter didn’t reach the third time through were two of Sean Manaea’s disaster starts, in which he was removed not to avoid future matchups but because the past ones had gone so poorly. In all the other outings they at least faced one batter a third time, and in seven games they saw the whole top-three spots a third time.

As for the percentage of total at-bats, that’s a bit lower, and it’s even got a bonus over the previous years because it includes any pinch-hitters (due to differences in how I compiled them). On the other hand, it’s also got a penalty compared with those other years, because pitch counts are lower and thus health precautions are still confounding the comparison — nobody on the A’s has thrown 100 pitches in a start, and only twice have they cracked 90, which is costing them extra at-bats at the ends of games.

* On that last note: In ‘05, ‘10, and ‘13, A’s starters faced around 68% of the team’s hitters; in 2017-19 is was down to 59.4%, round up closer to 60-61% due to 2018 openers; in 2020 it’s at 57.3% in these dozen selected starts, but that new drop could possibly be due more to pure health and pitch count caution than anything

We don’t really need the small-sample numbers to make this call for us, though. We can see it ourselves, when A’s starters are allowed to go about their business as normal. A few times they’ve been pulled amid jams, like Mike Fiers in his disaster Wednesday, but when they’re plugging along well they get to keep going until they reach a reasonable workload — and they’ve been efficient enough with their limited pitch counts to still eat mostly normal amounts of innings.

Montas has gone seven full the last two times out, Fiers did six full twice, and Bassitt did seven last time. In Bassitt’s appearance before that, in Seattle last week, he was doing nicely but was pulled from a tight game in the 6th — but he’d let the tying run come to the plate in the form of a tough lefty slugger, so a call to the pen made sense on its own merit. That might have been a bad matchup regardless of how many times they’d faced that day.

None of this changes Sarris’ point about the league-wide trend of starting pitchers being used differently. And indeed, the initial numbers suggest that even Oakland has slightly reduced its exposure to the third time through opposing lineups, unless that’s just the effects of decreased pitch limits. But either way the bottom line is, now that the A’s finally have a quality rotation with arms they trust, they seem to be zagging back toward relatively traditional use of them — or at least holding steady, as everyone else is rushing in the opposite direction to lean on their bullpens.