June 24, 2004

Breaking Balls

Framing the Debate

by Derek Zumsteg

Printer-friendly

If you watch the home team broadcast, almost every start by a home pitcher isn't just good, it's great. No, it's outstanding. Just plain fantastic. It was a gritty, gutty start. And that's for a six-inning, 9-hit, 4-walk, 2-strikeout start where the pitcher sees four runs cross the plate. He worked himself out of some tough jams. He literally put out some fires (the use of literally to mean figuratively causing writers and English majors across the country to literally grind their teeth).

That's to be expected. After all, the broadcasts are, first and foremost, marketing tools for the team. I shouldn't get frustrated when baserunning gaffes are excused, or a hitter's awful hacks are ignored. I do, but I shouldn't. When we find nuggets of serious analysis, or discussions that aren't flattering to the team, or even criticism of botched plays, it's a bonus.

The problem with that kind of slide towards boosterism is that it ends up making actual great performances unremarkable. If an average outing is the subject of a five-minute game of one-upsmanship between announcer and color man, like a crazed manic version of Statler and Waldorf, then what's left to say about a one-hitter?

How can you argue against pitcher abuse if racking up huge pitch counts means that a pitcher is praised for his durability? I saw someone, with a straight face, talk about what a workhorse C.C. Sabathia was becoming by throwing a ton of pitches. Sabathia was stepping up to be the leader, a big man who wanted to be the Big Man on the staff, and when he lost a season to a blown elbow, he'd be in the dugout providing clubhouse leadership whenever his rehab schedule allowed.

Okay, I made that last part up.

Or what about the use of statistics, when they're used, as a way of defining character? In an article praising a player (Raul Ibanez), I read that nine of his eleven home runs had either tied the game or given his team the lead. This was used to support the contention that Ibanez was a monster clutch hitter and the heart of the offense, and the GM who picked him up was brilliant for doing so.

It's all part and parcel, this kind of statistical laziness, changing the dimensions of the debate, misapplication of labels. It's the reinforcement of a viewpoint the spouter had arrived at long ago--the home team is tops, great pitchers carry great loads, or home team management is good and makes good acquisitions.

That said, nine of eleven home runs can set off an interesting line of thought. Most games are pretty close. Scoring a run, any run, should be pretty likely to tie or put a team on top. Home runs should be particularly good at that, since they're often going to put more than one run on the scoreboard.

I asked super-dude Keith Woolner about it, and through the games of 6/16 (which is when I asked):

• 30% of home runs put the team ahead
• 12% tied the game
• 23% only cut a deficit
• 35% were hit with the batter's team already ahead

In all of 2003, it was close to that: 31% took the lead, 9% tied, 23% made up ground, 38% came with the team already ahead.

Now, given a team like the Mariners, who score almost no runs, lose a lot, and play in a low-scoring environment, it's going to tilt even further toward making home run hitters look clutch. A Seattle hitter finds his team behind by close margins far more often than (say) an equal hitter for the Yankees.

That article got me to think about something interesting about the particulars of clutch statistics on different teams for a while, so I didn't mind at all.

But what about when sloppy debate harms people? Take Sabathia and his high-pitch outing. So we're clear, he threw 123 (see ESPN's box score) pitches. Sabathia didn't even pitch that well. He went eight innings, which is good, only one walk, good, struck out four... that's okay, and gave up a home run. Sure, it's a one-run win, but it's not a dominating start in the way that, say, one of Roger Clemens's 20-K complete game performances were.

This kind of start somehow gets turned into an argument against pitch counts, which seem to be perceived as some stathead invention that pulls pitchers automatically at a hundred pitches, no matter how good they look, or how the game is going. If Sabathia can go 123, why not let him go 130, or 140? I've seen a number of articles, even at mainstream sites like ESPN, that argue pitch counts amount to a lot of excessive handwringing by people who think, but don't know, that high pitch counts are bad.

It's easy to make that argument, if you're lazy. Find a pitcher who threw a lot of pitches (look, there's Sabathia). Then point out that they're not injured. By shifting the debate, and mis-characterizing a rational viewpoint by claiming something like "They says the 101st pitch causes a pitcher to explode," people who don't want to think too hard don't have to. A modestly complicated argument requires nuance, and nuance often loses space-confined arguments.

The point here at Baseball Prospectus has never been that Sabathia or any pitcher should be pulled at 100 pitches. It's that, generally, pitcher performance tends to decline after the 100-pitch mark, high pitch outings tend to result in short-term performance drops in the weeks after a high-pitch start, and continued long-term stress on a pitcher's arm increases the risk the pitcher will have a major injury.

What's more, the constant assertion that there's been no serious work on the subject, that no one can show a serious correlation between stress and injuries baffles me. Woolner, when he's not looking up crazy stuff about home run effects for me, did a huge amount of work to look at this, using ten years of game and injury information. Anyone can read his work on the short-term impacts of high pitch counts, as well as the long-term impacts of high pitch counts. Or his response to questions, including some from Sean Forman.

It's a complicated thing to understand, for sure. All pitchers aren't alike, and some will certainly be more prone to injury than others. Some will not do badly after a long outing, while others will.

This is where the terms, and the way the debate is framed, become important. Is it worth it to continually pimp the home team if it means that worthy accomplishments seem pedestrian? Does that help or hurt the sport? Does a statistical fluke change the value of a player to his team? Is being a macho, macho man and throwing a lot of pitches important enough that it's worth risking a player's next couple starts or his career?

As baseball gets smarter, the quality of play improves and the game is better able to compete for fans and the future. Keeping an eye on how debates start and how they're defined will help advance the game.