On questioning Joe Maddon: Pinch hitting Matt Joyce, game 2 ALDS

With five innings and one out in the books, John Lackey's line stood at 4 ER, 7 H, 3 BB, 6 K, and 95 pitches.

The Rays had just scored their fourth run and had a man on first base—Yunel Escobar, who had batted in Desmond Jennings—with one out, but the Red Sox were still leading by two.

John Lackey's control had been expiring, with only 14 of 25 pitches resulting in strikes the previous inning. Already at 95 pitches after netting the first out of the sixth, he was going to be pulled in the near future. John Lackey's velocity was on a steady decline, but Joe Maddon didn't ride Lackey into the ground. He took matters into his own "hands."

With the right handed Jose Molina and then the left handed David DeJesus due up, Maddon opted to replace Molina with Matt Joyce, hoping to put two lefties in the path of John Lackey.

This would prove to be an ill-advised move.

As we discussed yesterday during the lineup analysis, and will probably discuss again, Joyce had seemed like the obvious choice to start at designated hitter against Lackey. Both players being considered for designated hitter—Delmon Young and Matt Joyce—had hit fastballs well from the locations Lackey was most likely to throw them, and Joyce had the further advantage of being left handed, but Maddon had the clear intention of "playing the hot hand" in Delmon Young, who was there to hit the ball deep enough for runs to score.

That logic aside, choosing to bring Joyce in at this moment in the sixth inning was a strange choice.

The Red Sox had lefty Craig Breslow warming in the bullpen and a few pitches away from being ready. Maddon had to know he that by pinch hitting when he did he was forcing John Farrell's hand to pull his struggling, right handed starter. Sure enough, the Boston manager would stall as much as possible, and then make the swap to bring in Breslow to face Joyce.

After Joyce flew out to center field, Maddon replaced David DeJesus with the right handed Sean Rodriguez, who grounded to short. With the Rays not scoring any more runs in the inning, Maddon immediately pulled S-Rod for Jose Lobaton, burning a possible defensive replacement without taking the field.

A few problems here:

1. Maddon lifting Molina for Joyce allowed a gifted reliever to enter the game earlier than necessary. Rewinding to that moment, if Maddon had left in both of his starters, and Farrell had not made the corresponding move, this would have been the projected wOBA, based on each player's career splits and the ZiPS RoS projections:

In any scenario we could discuss, this would have been the most ideal. Allowing DeJesus to have one more chance against Lackey would have involved a respectable wOBA; however, considering how exhausted Lackey had become, there's a near zero chance Farrell would have allowed Lackey to face the Rays' leadoff hitter.

Maddon tried to seize the moment by bringing in Matt Joyce, but was the gamble worth it? Farrell took no chances and made the change to Breslow, leading to this match up:

Note: the projection system takes in to account a 10% penalty for pinch-hitting, stemming from research in The Book.

The quick hook displayed by Farrell indicates that the Boston manager had no intention of allowing Lackey to face DeJesus, even if Maddon had left Molina in to face Lackey.

With that hindsight in mind, a better chess match may have been to allow Molina—who had already made two plate appearances against Lackey—the opportunity to bat Yunel Escobar around the bases.

Riding the earlier suspicion, even if Molina had not been able to do anything, Farrell would likely have brought in Breslow to face DeJesus, and Maddon could have responded by pinch hitting Rodriguez, leading to this more favorable match up:

Here's where the red flag should be going up again. Your brain should be telling you, "leaving Jose Molina in the line up does not seem like it would be the best choice," and you would be right.

We can even take this train of thought one stop further: swapping Jose Molina for the other catcher, switch hitting Jose Lobaton.

But even if Craig Breslow had been brought in after this switch, the numbers are still favorable to using Jose Lobaton:

Recognizing that DeJesus would never have the opportunity to face Lackey, and that Lobaton would need to be brought in as a defensive replacement one way or another, the wisest move may have been to pinch hit Lobaton from the get go.

Of course, the best situation might have been a pinch hit by Jose Lobaton and another chance for David DeJesus against Lackey, but that's wishful thinking.

Before Game 1, the TBS broadcast relayed multiple times that Maddon had considered "every possible match up" for the first game in the series. Assuming Maddon had done the same for Game 2, and assessing Joe Maddon's state of mind, he truly must have believe that John Lackey would have pitched a full six innings, or at least longer than 5.1 innings.

That was a mistake.

2. We've already established that choosing to pinch hit with Matt Joyce was, by our projections, not the ideal choice (if Maddon had any suspicion that Farrell would switch to Breslow), but there was an additional consequence—it removed a later opportunity for Joyce to pinch hit, particularly against right handed reliever Junichi Tazawa.

Joyce has an excellent projection against Tazawa, and he would have been a significant upgrade over any of the three batters who faced him in the eighth inning of yesterday's game: Desmond Jennings (.277 wOBA), Delmon Young (.264), and Yunel Escobar (.275)—all right handed batters.

3. And this has a minuscule probability, but Sean Rodriguez is the third catcher on the depth chart, and Jose Molina was lifted the batter previously for Matt Joyce. Then S-Rod was pulled for a defensive replacement in Lobaton. If Jose Lobaton were to be injured at any point between the sixth inning and the end of the game, the Rays would have needed to turn to Wil Myers, a converted catcher, to take the mask. This is not ideal, but luckily, the problem never came to pass.

As always, the wOBA projections are regressed platoon splits calculated using ZiPS RoS projections and Bojan Koprivica's platoon research.

All statistics come from a wOBA projection tool created by Jason Hanselman of Dock Of the Rays and our own Ian Malinowski.

More from DRaysBay:

• Pitcher Preview: Clay Buchholz
• Rays 4, Red Sox 7: Price, Offense Come Up Short As Sox Extend Series Lead
• Reflecting on David Price's Game 163 Dominance