Monday, April 15, 2013
“Starting” pitcherPosted by Pat Andriola
It was the top of the fifth inning and the Giants were down by one run to the Cubs last Thursday. The bases were loaded and there was one man away. Hisanori Takahashi, a soft-tossing lefty, was in for Chicago, and the ninth spot in the order was up for San Francisco. This meant Ryan Vogelsong, who had already thrown 81 pitches in the game and had given up five runs, came up to the plate.
Vogelsong battled and worked a walk to erase the last of what was a five-run deficit and tie the game. The Giants scored two more in the inning, eventually holding on to a narrow 7-6 win.
I was confused when Vogelsong came up to hit. Results aside, the Giants were down by one run and had the bases loaded with one out, which should produce, on average, around 1.55 runs. According to win expectancy, the game was completely even, 50-50, when Vogelsong hit. So the question is, should Bruce Bochy have pinch-hit?
I think it’s unquestionably yes. Vogelsong wound up going two more innings, throwing 26 more pitches, after the top half of the fifth. Last year Vogelsong threw 3,056 pitches in 31 starts for an average of 98.6 pitches per game, so based on last year’s numbers, Bochy was looking at around 18 more pitches.
Ryan Vogelsong is a pretty bad hitter. He has six career extra base hits in 209 plate appearances, good overall for a .198 wOBA (17 wRC+). Meanwhile, Andres Torres and Marco Scutaro, switch-hitter and right-handed hitter respectively, toiled on the bench. Torres had a .342 wOBA against lefties last year and is at .324 (102 wRC+) for his career. Scutaro had a .315 wOBA against lefties last year and is at .325 (96 wRC+) for his career. Simply put, both players would have been immense upgrades over Vogelsong.
So, in essence, Bochy had this calculus:
(Value of ~18 pitches of Ryan Vogelsong – Value of ~18 pitches of substitute pitcher) > (Value of ~.325 wOBA – Value of ~.198 wOBA in situation with leverage index of 3.80)
I’m sure there are ways to rationalize the decision. There was only one out and the top of the order was coming up, the Giants wanted to save their bullpen, Vogelsong was settling down and it was good for his confidence, Bochy saw an arrangement of sunflower seeds on the ground that spelled out “RYAN” and thought it was a sign from God. But whatever, managers make mistakes like this all the time; it’s not the biggest deal in the world.
But it did get me thinking about the role of the starting pitcher and the ninth spot in the batting order. As long as the National League shuns the designated hitter, this is going to be an issue. It’s long been theorized that the way in which starting pitching has traditionally worked is suboptimal, and that’s probably right.
Interestingly, the game may be reacting to the realization of this fact. Starting pitchers averaged 6.66 innings per game in 1972, 5.98 in 2010, 6.03 in 2011, 5.89 in 2012, and 5.71 thus far this season (sample size warning, of course). Maybe managers are starting to realize that the word “starting” in starting pitcher is the most important part of the title: They start the game, but that doesn’t mean they need to be around forever. Dave Cameron talked last year about some of the play-in teams starting the game with a closer, which definitely would’ve been neat.
Interestingly, the shorter the average starting pitcher goes, the more important the ninth spot in the lineup becomes. The data behind pitchers hitting eighth have already demonstrated some possible underlying importance out of the ninth spot (probably because it immediately precedes players who typically have high OBPs). So maybe one benefit of being flexible with who's on the mound is that you can take advantage of high leverage index situations early in games. You also will probably have fewer pitchers reach the plate on average, which will unquestionably help the offense.
Bochy’s error is easy to pick out because of how crazy the situation was: a huge leverage index spot pretty early in the game, a starting pitcher who already had a high pitch count, and pretty good pinch-hitting candidates. But what if we tweak the formula above? Instead of 18 pitches of Vogelsong, make it 30, and instead of a 5-5 game with the bases loaded and one out, have the Giants up 5-4 with a man on third and one out. It gets trickier, and the sooner managers start experimenting with shorter stints from their starters, the sooner we’ll see some interesting managerial maneuvers.