The Hook

An article appeared on Fangraphs today questioning the pitching decisions of Charlie Manuel and Tony La Russa—specifically that pitchers Kyle Lohse and Cliff Lee were not pulled early enough in games one and two, respectively. The author, Eric Seidman received a lot of flak in the comments for his assertion. Articles like this appear all the time, utilizing hindsight to pass judgment on managerial decisions. Had Lohse and Lee individually pitched fine the following inning, would an article have appeared at all? Should an article have appeared? I think so.

Pitchers get worse as the game goes on. Whether it’s because batters change their approach by the third or fourth time through the lineup, or pitchers get tired, or some combination of the two, the stats are clear.




SP 1




SP 2




SP 3




SP 4





This is data from the entire NL in 2011. “SP 1” simply means the average of all starting pitchers when facing batters for the first time in a game. The first three times through the lineup, pitchers as a whole get progressively worse. This makes sense if you buy into the logic above. Then there’s a strange blip the fourth time through the lineup. Why? Only about a third of all starters even make it to the leadoff guy four times, let alone all the way through the lineup. Basically only the best pitchers are allowed to pitch four times around, so there’s some selection bias going on. But still, that fourth row is heavily weighted by pitchers like Tim Lincecum, Chris Carpenter, Cliff Lee, Roy Halladay and the gang, and their stats are still worse the fourth time than average starters facing batters for the first time. The primary conclusion here is that pitchers simply become less effective as the game progresses. This information can help us determine when starters should get the hook.

The Fangraphs article was written about two above-average pitchers in the NL (Lee is even in consideration for the NL Cy Young). Maybe the rules don’t apply to good pitchers, so let’s look at those two pitchers’ splits. These are their career numbers.

OPS (against)








Based on the original chart and the graphs above, we can see that A) ALL Pitchers get significantly less effective the fourth time through a lineup, and B) Average relief pitchers are arguably more effective than even Cliff Lee his third time through the order (.686  vs .726 OPS).

This is a little unfair to Lee, since he is a much better pitcher now than his career averages suggest. In fact, this season Lee posted a .606 OPS against batters faced a third time in a game. However, when Lee went back out to the mound to pitch the seventh yesterday afternoon, he would be facing the heart of the Cardinals order in Freese, Pujols and Berkman…the fourth time through the lineup. Whether you’re looking at his career or his season, there is no argument that Cliff Lee is significantly worse facing batters a fourth time, and he is less effective in that scenario than average NL relievers (let alone Philly’s best relievers). He gave up the eventual game-winning run before being yanked.

A day earlier when Kyle Lohse went out to start the sixth inning, he would be facing the top of a Phillies order that features Rollins, Utley, Pence, Howard and Victorino. He would be facing them a third time. Take a look at the graphs above to see what happens to Lohse between the second and third times through the order. His K/BB ratio drops precipitously from 2.15 to 1.35, and his OPSa spikes from .800 to .841. Even this season, despite improving his K/BB, batters have found him much more hittable after seeing him twice, improving their BABIPs from .281 to .287, and their OPS from .686 to .758. Lohse got hammered that inning for five runs, and the Cardinals ended up losing by five.

But even if we ignore the results—which we really should anyway being just a couple of many potential outcomes—both these pitchers were being thrown to the wolves when smarter options were available. Fangraphs writer Seidman was hardly off-base when he suggested that the two starters be pulled earlier. He may not have found quite the right data to articulate his point, but the conclusion remains logical.

During a 162-game season, getting additional innings out of a starter, even if he becomes less effective, has positive residual effects. Relievers are saved for another day. However, in short playoff series managers cannot afford to use 162-game logic. In the playoffs, starters should get the hook even earlier than in the regular season, and definitely earlier than when Manuel and La Russa finally pulled the plug.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: