Breaking Balls: Pitches Per Out

This Sunday I headed down to Safeco Field (“The House That Griffey Built And Left”) to catch a Mariners game, and was treated to some great contrasts. Gil Meche wavered, striking out seven in only five innings, allowing four hits and three walks and taking 116 pitches to do it, of which 68 were strikes. Meanwhile, Mark Buehrle breezed, throwing just a hair over half as many pitches until he unraveled in the 6th inning. Meche was relieved by Rafael Soriano, who’s been so impressive of late I’ve given him my full endorsement. Soriano pitched two innings and struck out five of the next seven batters he faced, allowing only a solo home run–and it took him only 34 pitches, 17 per inning, or five less an inning than Meche.

Buehrle was relieved by once-closer Billy Koch, who took 34 pitches to get out of one inning, again bringing up a thought.

How many pitches does it take a guy to get an out?

The stat of Pitches Per Out is at once a cool tool and sort of dumb, like many specialized tools. (I have this bike tool that came with an expensive set I bought and while I’ve become all too well acquainted with every bike repair that can be made to a Cannondale, I still have no idea what that thing is. It looks like an 1895 can opener with a length of unsecured chain attached to it, as if the Jets had raided the Museum of Industrial History and converted it to a rumble weapon.)

Pitches Per Out seems to tell you too much and too little. It doesn’t tell you a lot about how they’re getting those outs–strikeouts or balls in play–or what’s happening to the outs they’re not getting–home runs, or walks, or singles? There are other problems here too.

With those limitations noted, I looked up some pitchers and found interesting results:

Top 10 Pitchers, Pitches Per Out 
(Min 10 starts)
                       Pitches/  Pitches/
Who	         Team   out      6 IP	 SNWAR	BB/9 (rank, ML)
Greg Maddux	 ATL	4.50	  81	 1.1	0.67 (2)
David Wells	 NYY	4.58	  82	 2.4	1.47 (1)
Roy Halladay    TOR	4.63	  83	 3.6	1.14 (3)
Mark Mulder	 OAK	4.64	  83	 4.7	1.91 (8)
Sidney Ponson	 BAL*	4.77	  86	 2.6	2.61 (42)
Tim Hudson	 OAK	4.79	  86	 4.7	2.34 (27)
Miguel Batista	 ARI	4.81	  87	 3.1	2.45 (35)
Ryan Franklin   SEA	4.82	  87	 1.7	2.21 (23)
Kevin Brown	 LA	4.82	  87	 3.9	2.10 (19)
Jeff D'Amico	 PIT	4.84	  87	 1.6	1.82 (7)

All stats as of Sunday, August 3rd.
* Mostly

I’m dying to hit the next one, so I’ll be brief. Ryan Franklin’s always been an interesting pitcher to me. He’s the kind of pitcher you could go to before a game and say “we just played a 32-inning monster and you’re the only pitcher we have left. I need a complete game.” And Franklin would nod and throw a game where he’d let the ball get hit into play early in counts and only go after a hitter if he needed a strikeout.

Anyway, one of these things is not like the others. Maddux, who’s been one of baseball’s best and most efficient pitchers for years, is still efficient. He may not be doing as well as he used to, but he’s sure not using more pitches as he struggles.

For the most part, though, these are guys who are up in the Top 30 pitcher rankings or just off the tail end, highly effective one way or another.

Bottom 10 Pitchers, Pitches Per Out
(Min 10 starts)
                         Pitches/   Pitches/
Who	          Team	  out	     6 IP	SNWAR	BB/9
Kazuhisa Ishii	    LA	  5.86	     105	1.7	5.73
Joaquin Benoit	    TEX	  5.91	     106	0.4	4.21
Jimmy Haynes	    CIN	  5.95	     107	-0.9	5.44
Glendon Rusch	    MIL	  5.98	     108	-1.5	3.61
Brandon Duckworth PHI	  6.06	     109	-0.1	3.80
Al Leiter	    NYN	  6.05	     109	1.0	5.56
Kevin Appier	    ANA	  6.08	     109	0.2	3.50
Jesse Foppert     SFN	  6.13	     110	-0.2	5.46
Oliver Perez	    SDN	  6.29	     113	-0.0	6.13
Colby Lewis	    TEX	  6.34	     114	-1.3	6.16

A diverse group. Most of these guys have done poorly this year, though Ishii’s still been effective despite having to burn a lot of pitches. There’s no obvious connection between these guys, besides the walks–they walk guys, just as the most efficient guys don’t. Well, duh, you’re thinking.

What’s that mean to in-game strategy, though? Say you’re a pitcher facing Joe Averagehitter, first inning, two outs, nobody on, and he’s worked you for three balls (because he’s batting third and is only average, you can figure you’re facing the Dodgers). At the moment, the curve’s not breaking for a strike and your change is dropping into the dirt. Do you let him walk and take your chances with the next guy, or do you serve up a meatball and see what happens?

You serve it up–depending on the game situation, serving up a hittable strike could be worth the increased risk of an extra-base hit because the out’s so valuable, and over the longer haul, so’s your arm. The problem is that pitchers having walk trouble might have trouble serving that pitch up.

Overall, pitch efficiency has a much wider effect. Say Ryan Franklin and Kazu Ishii are going up against each other. They’ve both been about equally effective this year. After six innings, the game’s pretty close, and Ishii, with 105 pitches thrown, isn’t going to head out for another inning if his pitch count’s 110. He’s certainly going to get pulled after the 7th, when he’ll be up over 120, and that’s going to be a tired inning, too. Meanwhile, Franklin heads to the mound in the 7th having only thrown 87 pitches. He’s going to be able to pitch the 7th and probably the 8th before he runs into that same problem.

If you’ve got a deep and effective bullpen, that doesn’t matter as much to you, but there isn’t a manager in baseball that wouldn’t like to see an effective starter work seven, eight good innings and save the bullpen for extra innings, or the next day. In that way, pitches/out may be a weird, fluky tool, which can still offer some use when you want to go to the trouble of calculating it.