Analyzing the Matt Harvey decision

On Sunday night, Terry Collins initially made the decision to remove Matt Harvey from the game after eight shutout innings. Harvey had pitched phenomenally, giving up four hits (all singles) and one walk while striking out nine. He had done all the Mets could have asked for, but with Royals hitters set to face him a fourth time with a pitch count already at 101, Terry Collins knew that it was time to remove Harvey from the game. The decision was made easier by the fact that he had a reliable closer in Jeurys Familia warming up in the bullpen, ready to protect a two-run lead.

Leaving aside concerns about pitch count and protecting Harvey from injury, removing him from the game after eight innings made sense because of the times through the order penalty, which is the idea that pitchers do worse against hitter the more times they face them in a single game. Mitchel Lichtman, who originally wrote about this idea in The Book: Playing the Percentages in Baseball, sent out several informative tweets Sunday night explaining the math behind the decision to remove Harvey and bring in Familia. Using projections, he determined that a fresh Familia is a better option in that situation than Harvey facing the Royals for the fourth time. With that being said, the difference between the two pitchers was unlikely to have an enormous impact on the game.

Once again probably about a 93% chance of Mets winning with Harvey pitching the 9th and a 94% maybe 94.3% with Familia.

— Mitchel Lichtman (@mitchellichtman) November 2, 2015

Still, every percentage point counts, and Terry Collins initially made the decision that gave his team the best chance to win the game.

It should also be noted that Harvey may have been showing signs of fatigue in the 8th inning, despite getting three relatively quick outs. I went back and re-watched the top of the eighth inning, and I was stunned to see how badly Harvey was missing the catcher's target. He threw nine pitches in the inning, and he missed his spot in a noticeable way on every single pitch.

Here's the Brooks Baseball plot from Paulo Orlando's at-bat against Harvey leading off the inning.

Harvey started the at-bat by throwing two sliders, both of which were supposed to be at the knees on the outside corner, based on where Travis d'Arnaud was set up behind the plate. Harvey yanked the first slider more up and away than he intended, but he was fortunate, as Orlando bunted the pitch foul. The second slider missed the target even more, sailing well above the strike zone. On the 1-1 pitch, d'Arnaud set up low and away once again, this time for a curveball. Harvey's pitch ended up being a hanger in the top of the strike zone, but Orlando could not quite center the pitch, hitting a fly ball to medium right field.

Next up was Alcides Escobar.

For the first pitch, d'Arnaud was set up at the knees on the outside corner, and Harvey missed low and away with a curveball. d'Arnaud was set up low and in on the second pitch, and Harvey missed his spot once again, throwing his changeup in the dead center of the strike zone. Like Orlando, Escobar was unable to center the pitch and hit a routine popup to the shortstop.

The last batter of the inning was Ben Zobrist.

Travis d'Arnaud was set up low and away for all four pitches of the Zobrist at-bat, and Harvey missed badly on each one. His first two pitches (a curveball and a fastball) were up and in, and Zobrist took them to go ahead in the count 2-0. (d'Arnaud was actually set up off the plate away for the second pitch, and Harvey still threw it inside.) Harvey then threw a fastball for a called strike on the inside half of the plate, but he once again missed his spot badly, and d'Arnaud had to awkwardly reach across the plate just to catch the pitch. Harvey's final pitch, a changeup, still missed the low and away target by over a foot, and it was also very close to the center of the strike zone. Nevertheless, Zobrist hit a lazy fly ball to center field, ending the inning.

What we have are nine pitches and nine missed spots, including several that missed by more than a foot. Harvey could not locate any of his four pitches, and he missed his spots both inside and outside the strike zone. He fell behind in the count to two of the three Royals hitters, and he failed to induce a single swing and miss. Harvey's eighth inning is a great illustration of the difference between process and results, because anyone who looked solely at the results would assume that Harvey was as good as ever in the eighth inning. In reality, he was probably lucky to get out of the frame unscathed.

I presume that the manager and the pitching coach are responsible for paying attention to these types of details so that they know when a pitcher is losing effectiveness, even if his results may say otherwise. Perhaps Terry Collins and pitching coach Dan Warthen took this into account when they initially decided to remove Harvey from the game. Nevertheless, Collins ended up changing his mind after Harvey insisted on staying in the game.

In reality, this move may be more about the psychology of reason versus emotion than anything else. I have presented what I hope is a logical argument on why Harvey was not the best choice in that particular situation, but we cannot ignore the emotion of the situation, especially with Harvey pleading with his manager to give him a chance and Mets fans chanting his name. Ultimately, Terry Collins sided with his emotions and let Harvey stay in the game, which initially pleased the home crowd at Citi Field. Collins probably figured that it was possible to get the best of both worlds, making Harvey and the fans happy without losing the game. After all, the difference between Harvey and Familia in that situation was unlikely to have an effect on the outcome of the game (or the series as a whole).

With that being said, Terry Collins' goal on Sunday night was to win a baseball game, and he clearly made a move that did not give the Mets the best chance of winning. The difference between Harvey and Familia may have been small, but in Game 5, we may have witnessed the rare situation where such a move actually made the difference between winning and losing the game.

Nick Lampe is a featured writer at Beyond the Box Score and Viva el Birdos. You can follow him on Twitter at @NickLampe1.