Torey Lovullo and the everyday line-up

Personally, I don’t see very much grounds for complaint about Torey Lovullo. Today sees game #100 in the schedule, and the Diamondbacks find themselves 15 games above that, in the first wild-card spot, with a record that would lead half the divisions in baseball. Given this is a team few expected even to be at .500, it’s a remarkable feat, and Lovullo seems likely to be a front-runner for Manager of the Year. But some fans still take issue with his line-ups - in particular, his tendency to sit those they consider should be “everyday starters”. Let’s look at the data and see what’s going on.

The myth of the “everyday starter”

The baseball season is more of a death march: at 162 games, it’s twice as long as any other sport, a daily grind played out over the course of six months (or more, if you make it to the post-season). But recent years have shown an increasing trend to give players more rest over the course of the campaign, in an effort to reduce the stress and keep them fresh. Last season, in the National League, only six players started more than 153 games. That’s one-third of the number less than ten years ago; there were 18 such in 2007, but the trend over the past decade has been almost relentlessly down, as the chart below shows.

If we look at last season, you will find far fewer who were an “everyday starter” than you might think. With fifteen teams and eight position player slots in the NL, that’s 120 in total. But only 107 players, across the entire league, started even more than half of their team’s games. The players outside the top 120 for starts last season - roughly the top eight on each team - made more than a quarter (26%) of all starts in the league. Put another way, in the average NL line-up last year, two spots were occupied by players who should not be considered everyday starters.

Now, there are a range of reasons which go into that - injury obviously plays a part. But we are also seeing teams much more aware of fatigue. Stan Conte, who was the Dodgers’ trainer from 2007-15, said, "I think we're getting more analytical about this and asking ourselves how we can preserve these players over the course of a season. If a guy does great playing four games in a row but doesn't do well on the fifth day, why are we playing him when we have a fresh guy on the bench who could be better? I think we definitely see a decrease in performance when people get fatigued. And then maybe you get injuries after that."

Analysis has shown a measurable difference in performance, apparently resulting from fatigue. “The number of games that a hitter had played in the previous days had an across-the-board negative effect on all sorts of hits and increased the number of outs in play that a hitter made.” In particular, tiredness increased the proportion of balls in play that turn into outs. “Ground balls that used to be smacked hard enough to get past the shortstop now end up in a simple 6-3 putout. Baseball is a game where a small fraction of a second can make a huge difference on a ball. Miss on timing that swing by a little bit, and you'll roll the ball over to the second baseman.”

Is Torey Lovullo over-resting players?

We’ve hopefully established the importance of rest over the course of a season. Is our skipper giving the everyday Diamondbacks more rest than his colleagues? First, let’s clarify, what we mean by “everyday Diamondbacks”. LF and CF have been significantly impacted by injury to Yasmany Tomas and A.J. Pollock respectively, while catcher was always intended to see a split of playing-time between Chris Iannetta and Jeff Mathis. So it’s the core five of Paul Goldschmidt, Jake Lamb, Chris Owings, Brandon Drury and David Peralta. These have easily been our most regular players; including today, they are at 96, 93, 86, 80 and 74 starts respectively; no-one else is even at 50.

How does that compare? A quick straw-poll on other teams in the playoff hunt shows it seems pretty reasonable. Going into play here today, the Dodgers have also played 100 games. They have no-one at ninety starts, and their top five are 87, 84, 77, 72, 70. The Nationals? Similarly, even Bryce Harper has made only 88 starts for them, and their everyday players are at 88, 88, 88, 71 and 70 starts. Crunching the numbers for all 15 National League teams, the starts by their top five ranged from 378 (Mets) to 460 (Reds). The D-backs’ total of 425 (not counting today) is seventh - but certainly worth noting, only one of the more “regular” teams above them has a winning record.

Looks like there’s nothing to suggest Lovullo is out of line with his colleagues, in terms of resting players. Indeed, he’s being a bit more aggressive in his use, when you compare the number of games played by our core, against the teams with whom the Diamondbacks are directly competing this season.

Are his line-ups costing the D-backs wins?

Probably the most frequent complaint is any game where Chris Herrmann is in the starting line-up. Expect to hear comments such as, “Torey conceding another one” on social media, perhaps with something mentioning Herrmann’s batting average. But there’s a very strong amount of confirmation bias going on here. If the team loses, it’s Lovullo’s fault for the crappy line-up. But if they win, he’s never given any credit for it: this is truly a lose-lose situation. And here’s the reality. The 2017 Diamondbacks are currently 21-15 when Herrmann starts. That’s a fractionally better win percentage than without him in the line-up.

Even when we have multiple “bench guys” in the line-up - which happens less than you’d think from the complaining - it hasn’t hurt the D-backs. There have been seven occasions so far when our starting line-up has included Hermann, Gregor Blanco AND Daniel Descalso at the same time. The team is a perfectly respectable 4-3 in those. And it wasn’t all about the pitching, with the “B line-up” scoring an average of seven runs per game. Here are the results in those:

• 5/7 at COL L 2-5
• 6/6 vs SDP W 10-2
• 6/11 vs MIL W 11-1
• 6/24 vs PHI W 9-2
• 6/30 vs COL L 3-6
• 7/20 at CIN W 12-2
• 7/23 vs. WSN L 2-6

Part of this is that, as we noted a month ago, Herrmann is a much better hitter when he starts than coming off the bench. with an OPS currently 268 points higher. As a third catcher, he does help, not even directly, because it means we can pinch-hit for Jeff Mathis in the later innings, then use Hermann or Chris Iannetta behind the dish, and still have coverage in the event of an injury. But we can count on the fingers of one hand, the games so far where Herrmann has come into the game off the bench, and stayed as a catcher. If there is legitimate cause for complaint, I think it’s more likely his use as an outfielder.

But in turn, I think that was more a case of necessity than choice, a result of the injuries to Pollock and Tomas. With the former now healthy, and the arrival (after a brief detour to the bench) of J.D. Martinez, I’m inclined to think we’ll be seeing a lot less of Herrmann in the outfield, all being well, with Gregor Blanco now the fourth outfielder.

Conclusion

There’s no doubt, I’ve been fairly scathing about the critics. That’s because, while I may not agree with every decision Lovullo has made, he is party to far more data than we have on the outside, and I think the results so far speak for themselves. It’s over the last couple of months where we may start to see the benefit, especially for our three most regular starters. It’s worth noting, across his entire career, Goldschmidt’s second-half OPS is almost a hundred points below what he does in the first half. Owings shows about the same decline. For Lamb, the drop-off is even bigger: 232 points (last year, all NL OPS was basically unchanged between the two halves).

If giving them all more regular days off helps avoid this, and keeps them performing at a higher level down the stretch, than it’s something we should probably embrace, rather than complaining about. Because we’ll reap the rewards in the post-season.