justin-verlanderYou can still hear Peter Gammons after Grady Little failed to lift Pedro Martinez before the eighth inning of Game Seven of the 2003 ALCS.

“Why not just use your bullpen?”

Pedro Martinez and the Boston Red Sox had a 5-2 lead heading into the bottom of the eighth inning. Martinez was due to face New York’s 2-3-4 hitters for the fourth time that game. Martinez was able to induce a pop fly to Nick Johnson to begin the inning, but four batters later, after a single and three doubles, the lead had evaporated. The Sox would lose the game, and the chance to break the curse of the Bambino on Aaron Boone’s solo home run in the 11th inning.

Obviously one game in April is not Game Seven of the ALCS, and the Kansas City Royals are not the 2003 Yankees, but this anecdote should always serve as a warning to managers who insist on keeping their starter on the mound late in a game*.

*Not that one anecdote should ever cause us to conclude anything. It’s just an anecdote, no more, no less.



I understand that Justin Verlander is a phenomenal pitcher, perhaps (probably) the best in the game right now, but even Verlander is susceptible to the “times through the order effect”. As I tweeted out when Verlander stepped on the field in the bottom of the ninth: “4th time thru the order is rarely a good idea.” Pitchers (generally) lose effectiveness as the game wears on, but perhaps more importantly than that, hitters learn the pitcher.

Here’s a look at Verlander’s career on-base percentage allowed per time through the batting lineup (data from Baseball Reference):

Split OBP
1st PA in G as SP 0.293
2nd PA in G as SP 0.280
3rd PA in G as SP 0.316
4th+ PA in G as SP 0.341

For his career as a whole, he’s allowed a .298 OBP, but it’s easy to see how that number changes as he progresses through the game. It’s probably worth noting that the league’s average on-base percentage was .321 in 2011 (and it’s been .333 or below since 2008). So clearly Verlander is a dominant pitcher the first two times through, and he’s still above average the third time, but, in terms of recording outs, he’s below average the fourth time through the lineup or later.

This is huge. It means if you’re given the choice between Verlander facing hitters for the fourth time and an average reliever, you probably choose the average reliever. And if the choice in that same situation is Verlander or one of the premier closers in the game, you go to Papa Grande without batting an eyelash.

Watching Valverde blow the opening day save after Verlander pitched a gem probably made Leyland a bit skittish (and understandably so), but that was definitely the right move. Things don’t always work out and closers sometimes blow saves, but you’ll get burned far less when going with an ace reliever than you would with leaving your starter in too long.

Matt Snyder is the editor of The Tigers Den. He can be reached on Twitter @snyder_matthew.