“He was doing so good, why did they take him out?”. I heard this grammatically butchered phrase more than any other this postseason when discussing the games with other fans, both in person and online. In spite of the growing trends in the game as well as many hard stats, in general folks still overwhelmingly prefer the old school mentality of not wanting to remove any pitcher from the game “when he is doing well”. I propose the counter-argument – do we really want to wait until the guy is getting lit up before we make the call to the bullpen?
If I ever land a job as the manager of a major league baseball team, my number one goal will be to remove every single pitcher from the game while he is doing well. If you could manage that trick, you would never lose! Obviously in a perfect world, we’d take every pitcher out one batter before he lost it. Clearly that will never happen. No manager is going to be able to perfectly toe the line between squeezing the last “good” pitches out of a starter while still getting in a top-notch reliever before anything goes wrong. In close and high-stakes games, like say, Game 7 of the World Series with a 2-0 lead, the margin for error is extremely small.
The general Monday-morning quarterbacking consensus from Astros fans is that Hinch blew the game by taking Zack Greinke out “when he was doing well” with 1 out in the 7th. Now I will certainly hear an argument that with the benefit of a robotic umpire, Mr. Greinke could potentially have been in a better position to finish the inning. However, missed call on Soto aside, he had just given up a home run and a walk. With the pitch count being what it was I don’t think it was necessarily a stamina or fatigue issue, but to me, regardless of the pitch count, 6 innings was plenty for him. I don’t think anyone expected him to be on the mound in the 7th before the game started, and AJ Hinch was going to need to utilize his bullpen to win the game in any scenario. I posit that Hinch’s mistake was not removing Greinke from the game too early, but rather letting him begin the 7th inning in the first place.
Whether it was Cole, Harris, Smith, or Osuna who came in first, it was going to take most or all of them to nail down the win. Only needing to cover three innings was already a huge advantage for Houston. We can certainly argue that Hinch could have picked someone other than Harris who was showing signs of simply running out of gas for the season as a whole, but I think the pressure on Hinch placed by the low pitch count forced him into a mistake. If Cole comes out to begin the 7th, there’s a great chance he shuts it down and they can go to either Smith or Harris in the 8th. They would also have Osuna at the ready to polish off the game at any point. The most logical move was having Cole follow Greinke’s outstanding effort, but by letting Greinke start the 7th Hinch removed that option for himself, as he needed the “fireman” skill-set to come in mid-inning with a runner on.
Before the game started, all I heard was how shaky Greinke was, how they’d be lucky if he got through four solid innings, how they were going to count on the bullpen to get it done, etc. All of a sudden everyone wanted him running through the order a 3rd time just because his pitch count was sort of low? He looked great for sure, dominant even. I can’t remember if it was Smoltz or Buck who said it was the game of his life, but it really did feel that way. I can feel the temptation to want him in there, both for Hinch and for the fans. From a pure statistical standpoint, if you have a 2-run lead and 9 outs to go for the World Series Championship, ranking extra rested Osuna, rested Smith, hopefully good Harris, and gritty Cole with 2 days off ready to leave it all on the field, I think they would all grade out ahead of Greinke’s last 15-25 effective pitches even in the game of his life. With so many weapons available, I’d rather leave a quality inning from Greinke on the table to avoid the risk that comes along with wondering if that last inning is going to go well. Especially when you factor in the decrease in expected quality from a reliever who comes in with men on base.
Game 7 is hardly the only situation where people have complained. I heard lots of chatter about Hinch having a quick hook with Urquidy in Game 5. To me, that was the perfect call. He set down the side 1-2-3 in the 5th, but why push it? His pitches in that inning were not as crisp regardless of the result, and regardless of his crispness and pitch count, there was no reason to give any of the Nationals hitters another look at him. Again, too many weapons available. No matter how well he was doing or what his pitch count was, there was no reason in the world to allocate more than 5 high-leverage World Series innings to Urquidy. Hinch knew it, which is why they won that game in spite of a bullpen that wasn’t perfect. He may have sacrificed a few outs that Urquidy could have gotten, but he managed the game as a whole in a manner that yielded a victory.
Relief pitchers are not always perfect. Just like starters they are going to give up runs, but for relief pitchers their runs can be magnified drastically. Even a super-star closer with a 1.80 ERA is going to give up a run in approximately 20% of his outings. Just like no manager will ever be perfect in removing the starter from the game one batter before he isn’t doing well, no relief pitcher will ever shut it down every single time he comes in. The knee-jerk reaction any time a reliever gives up runs is that the manager screwed up by taking out the pitcher before him, especially if the previous hurler hadn’t fully lost his stuff. The end result may not always work out for the best, but the manager’s job isn’t to strike anyone out or hit home runs, it is to put the entire team in the best possible position to win the game. I propose the theory that putting the relief pitcher in the best possible position to succeed – bringing him in to begin a clean inning – is even more important and beneficial to the winning cause than squeezing the last few outs from a starter who is running on fumes.
Looking back at Greinke in Game 7, he was crushing it. He looked awesome. I would have given him a really good chance of getting through the 7th inning with no runs allowed. Certainly better than 50/50, I might have gone as high as 65 or even 70%. I think if you played that inning 100 times, he absolutely gets it done in at least 60 of them. But how many times do we think Cole doesn’t give up a run? 75%? 80%? 90%? Even if we only assume 75%, wouldn’t we prefer to have a 75% chance of taking a 2-run lead into the 8th than a 60 or 65% chance?
No pitcher is a sure thing. Both the “doing well” guy currently on the mound and the fresh arm in the bullpen have upsides and downsides. Both of them are going to give it up sometimes. I think the manager’s job is simply to put the guy on the mound who has the best chance to succeed, even if that means taking out a guy who has only a slightly-less-good chance at success.
