Is Joe Girardi letting his starters pitch too deep into games?

In three of the past five games, Phillies manager Joe Girardi has allowed his starting pitcher to go into the eighth inning. On two of those occasions, he allowed the starter to pitch into the ninth inning and attempt a complete game. Here’s what happened in both of those instances.

Those two home runs have spurred plenty of discussion on Twitter, in The Good Phight’s comment section, and pretty much any other place where the Phillies are a topic of conversation. Our own Smarty Jones shared his two cents on the matter in his Diamondbacks series preview.

Personally, I feel a little conflicted about the whole thing. On the one hand, I love complete games and I’m always rooting for a starting pitcher to finish what he started. Moreover, the Phillies bullpen has been unreliable to say the least.

On the other hand, there’s a reason why complete games are so rare. It’s really hard to pitch into the eighth and ninth innings and still get batters out. In addition, the Phillies have some pretty durable starting pitchers, but even the most durable pitchers need a break every now and then.

In these two particular instances (Aaron Nola against the Padres and Zack Wheeler against the Rays) I feel confident in saying that Joe Girardi should have pulled both pitchers after the eighth inning. On Saturday, Aaron Nola had already thrown 100 pitches when he went out back for the ninth. He had already shown signs of tiredness in the seventh. It was a close ball game, and Fernando Tatis Jr. and Manny Machado were both due up in the inning.

In Wheeler’s case, he was already close to 100 pitches. He had started to show some cracks in the top of the eighth. He was facing one of the best lineups in baseball in a tie ballgame. Ian Kennedy was fresh and ready to come in.

Both Wheeler and Nola tend to struggle the fourth time through an opponents batting order. Nola has a career 6.28 ERA and Wheeler has a career 7.65 ERA the fourth time through the order.

I love complete games and I’m very wary of the Phillies bullpen. But the manager has to use better judgement to pick the right moments to go for a complete game, and I have more faith in a well-rested Hector Neris or Ian Kennedy than I do in an exhausted Aaron Nola or Zack Wheeler. These are far from Girardi’s worst moments as a manager, but they were still bad decisions.

While I was doing the research for this piece, I started to think about a bigger question. I’m confident that Nola and Wheeler should have been taken out of those games sooner, even though I love a complete game as much as anyone. But what about in general? Has Joe Girardi been leaving his starting pitchers in too long all season?

Overall, Phillies starters have a 6.31 ERA in the 7th-9th innings this year, which doesn’t look great. It’s a lot worse than the Phillies 4.59 bullpen ERA. That certainly makes it seem like Girardi should be taking his starters out of games sooner. However, when you look a little bit closer, it isn’t as bad as it seems. For one thing, Phillies starters have a 4.21 FIP in the 7th-9th innings, which is significantly better than the bullpen’s 4.65 FIP.

Aaron Nola has only pitched into the 7th inning 7 times in 25 starts, and he has a very respectable 3.65 ERA when doing so. Zach Eflin has pitched into the 7th innings just 4 times in 18 starts, and while he has a 13.50 ERA in the 7th-9th innings, it’s too small of a sample size (3.1 IP) to mean anything. The same can be said for Kyle Gibson; he’s pitched into the 7th in 2 out of 5 appearances with the Phillies, and he has a 12.00 ERA in 3 IP.

What particularly concerned me were Zack Wheeler’s numbers deep in games this season. He's pitched in the 7th, 8th, and 9th far more than any Phillies starter, and he’s allowed a concerning amount of runs.

• Innings 1-3: 2.19 ERA (78 IP, 26 games)
• Innings 4-6: 2.47 ERA (73 IP, 25 games)
• Innings 7-9: 6.31 ERA (25.2 IP, 18 games)

It doesn’t look much better when you break it down by inning.

• 7th inning: 5.29 ERA (17 IP, 18 games)
• 8th inning: 5.40 ERA (6.2 IP, 9 games)
• 9th inning: 18.00 ERA (2 IP, 4 games)

Using the Baseball Reference statistic sOPS+, Wheeler has performed 61% better than league average in innings 1-3, 32% better than league average in innings 4-6, and 11% worse than league average in innings 7-9.

I found this particularly surprising, because I had though that so much of Wheeler’s value this year came from his ability to pitch deep into games. He’s gone into the 7th inning in 18 of his 26 starts. He leads the league in innings pitched. Is it possible that he’s actually been hurting the team by pitching deeper into games?

Thankfully, there is reason to believe that Zack Wheeler has just had some pretty rotten luck in the later innings. His BABIP allowed in innings 7-9 is .343, compared to .276 in innings 1-3 and .291 in innings 4-6. For that reason, his FIP in innings 7-9 looks a lot better than his ERA. His 3.21 FIP in innings 7-9 is significantly higher than his 2.41 FIP in innings 1-6, but it’s still excellent.

As for his ‘fourth time through the order penalty’, Wheeler still has a pretty atrocious 6.57 FIP the fourth time through the order this season, but his xFIP is a much more promising 3.49. In such a small sample size (27 batters faced), I put more stock in xFIP than FIP. His K/BB ratio the fourth time through the order is 6.00, which is basically as good as ever.

What this shows us is that the Phillies haven’t really benefited all that much from Wheeler pitching deep into games, since he’s given up 18 late-inning runs in the 18 games he’s pitched into the seventh. However, that doesn’t mean Wheeler shouldn’t continue to pitch into the seventh inning going forward. His durability is still an asset, and he can still be an excellent pitcher late in ball games. He just maybe shouldn’t aim for a complete game every single time, okay Joe?