11 August 2008

Saving Jerome Holtzman

When Chicago sportswriter Jerome Holtzman died this week at the age of 82, it was a shame that he didn't take his best-known invention with him.

In 1959, Holtzman created a statistic designed to measure -- with sharper focus than won-loss record and ERA -- the effectiveness of relief pitchers. With the save, which was recognized by MLB in 1969 as an official stat, he succeeded.

In the early days of the save, relief pitchers entered the game when it was close and late, often in tight situations. That's when a team's best reliever was called a "fireman," not a "closer," because he entered contests to put out fires. A fireman like Goose Gossage or Rollie Fingers could throw two or three innings at a clip. Almost no one earning a save threatened the minimum save requirement of protecting a three-run lead or beginning with the tying run on deck.

The heroes of the story so far, Holtzman, Gossage and Fingers, are all Hall of Famers. The villain in the story is a future HOFer. If he didn't invent the notion of saving his best reliever exclusively for the 9th inning, Tony La Russa popularized it. All of a sudden, nearly every save was cheap -- meeting the minimum save requirement.

The fate of the save disgusted even Holtzman, as managers corrupted the stat by turning it on its head. Rather than reflect pitcher value, the save has become the guide for relief pitcher deployment. Managers won't insert their closer unless it's a "save situation." Effect has become cause.

Imagine the batter analogy. Threatening the single-season walk record, Jason Giambi gets the night off and Joe Girardi declines to use him as a pinch hitter in key situations until there are second and third with less than two outs -- an obvious walk situation. Giambi pads his stats, but doesn't help the team. His agent is ecstatic, as are the Red Sox.


I recognize that I've beaten this dead horse into steaks, but Holtzman's death (and the imminent shattering of the saves record) reminds us that the save needs to be updated or eliminated to reflect the new closer reality. Better yet, it's time that clubs recognize that reality bites and there's a better way to employ relievers.


There's a saying in economics that applies to baseball: "We only value what we measure." In 2008, we value a pitcher who enters a game with one out and a three-run lead in the ninth and proceeds to give up two hits, a walk, a hit batsman and a pair of runs before retiring two batters. It's time to stop measuring the irrelevant. Maybe then we'll get our firemen back and Jerome Holtzman can rest in peace.

2 comments:

Anonymous said...

This is my favorite braindrizzle. For umpteen years I have been so frusturated by how relief pitchers are rated. "17 saves in 19 appearances. This guy is great." But in how many situations was the tying run on double deck? How often did the pitcher give up 2 runs and still get a save. Kudos to you Waldo for making this the focal point of your column. I have a simple suggestion that needs work but it is better than what we have now. If the tying run is on 3rd base= 4 points; 2nd base = 3 points, 1st base = 2 points; other than that the save is worth one point. And - somehow we need to reflect blown saves so that All-Stars like Billy Wagner who have 19 saves at mid season have their 6 or 7 blown saves reflected in their numbers. Even the great Mariano who has not blown a save this year has lost several games where he has entered when the Yankees are tied.

Thank you Waldo for allowing me to rant.

Mitch from Merrick

Waldo said...

The funny thing about Mitch's comment is that it's a universal sentiment, yet no one has made a serious effort to replace or rehabilitate the save. You'd think that teams that have to contend with this in arbitration might want to create a stat more reflective of actual value.

You're not just good-looking Mitch; definitely onto something.

By the way, we're beginning to see how the save is distorting our understanding of value this year. Twice already I've heard sports talk radio dolts mention Francisco Rodriguez as an MVP candidate because he's likely to break the single-season save record. But a close examination of his actual performance reveals that the record is more about opportunity than anything else. He's been about the fourth best closer in the AL this year.