Baseball BeatDecember 16, 2008
We Shall Not Be Saved
By Rich Lederer

Phil Wood, the co-host of Talkin' Baseball on Radio America, invited me to be a guest on the network's show last Sunday. The purpose of the interview was to discuss an article ("We Shall Not Be Saved") that I had contributed to The Hardball Times Baseball Annual 2009, which is available directly from ACTA today.

A replay of the interview is available via podcast for your listening pleasure. Wood, the host of the program Tim Donner, and I discussed relief pitchers, the whys and wherefores of saves, optimal usage of closers/firemen, and various metrics to evaluate bullpen performance. This segment is 15 minutes long and starts about halfway through the first hour of the show. You can pull the button to the midway point when Donner mentions the Mets adding two closers and the Indians signing Kerry Wood before introducing me.

The following excerpts from my article give you a flavor for the content of our discussion.

Before 1969, a save was either something Jesus did or a hockey statistic used to measure the value of goaltenders. In fact, the two were mixed in a famous bumper sticker that could be found on cars of Boston Bruins hockey fans of that era: “Jesus saves! And Esposito scores on the rebound!”

Phil Esposito, who led the NHL is scoring in five out of six seasons during the late 1960s and early 1970s, would plant himself in the slot near the net and score goals from all angles. The big center won the Hart Memorial Trophy as the MVP of the league twice and was inducted into the Hockey Hall of Fame in 1984.

Forty years ago, there was no mention of the word “save” in Major League Baseball’s rules book. That’s right, the save, which was the brainchild of sportswriter Jerome Holtzman, didn’t become an official statistic of MLB until 1969. It was the game’s first new statistic since the 1920 introduction of runs batted in (RBI).

I defined the save as detailed in 10.19 of the Official Rules of Major League Baseball, excerpted a classic article written by Holtzman that appeared in Baseball Digest in May 2002, and quoted Roland Hemond, Steve Stone, and Bud Selig as to the legendary writer's importance on relievers. Interestingly, Holtzman later told Chicago Sun-Times columnist Bill Gleason that he was "sorry he'd come up with the (save) concept" because "it wasn't necessary." Imagine that!

Another way to illustrate how the usage of top relievers has changed over the past four decades is to compare firemen of the past, such as Hall of Famers Rollie Fingers, Bruce Sutter, and Gossage, all of whom pitched predominantly in the 1970s and 1980s, with Trevor Hoffman, the all-time career saves leader who made his mark in the 1990s and the first decade of the current century.

To wit, of Fingers' 341 career saves, 135 (or nearly 40 percent) entailed pitching two or more innings, including 36 of three or more innings. Sutter and Gossage recorded 130 and 125 saves (or 43 and 40 percent), respectively, of two or more innings. Hoffman, on the other hand, has earned just seven saves (or 1.2 percent) of two or more innings out of a total of 554. Only two of his saves have exceeded two innings and none have been as long as three innings.

[Gabriel] Schecter reported that Sutter, in a matter of 39 days (from May 27 through July 4, 1984), had “more saves (nine) where he pitched at least two innings than Hoffman has in his whole career. Gossage did the same thing from Aug. 15, 1980, through the end of that season, and Fingers accomplished it in a 53-day stretch in 1979.”

As Schecter pointed out, “The earlier pitchers acted as their own setup men. These firemen put out the fire and cleaned up after themselves.”

Similarly, Fingers was credited with 101 of his saves (30 percent) when he entered the game with either the winning or tying run on base, while Hoffman has pulled off this feat in only 36 of his saves (less than 7 percent) and only once in the past seven years.

The biggest difference between yesteryear’s firemen and the current crop of closers is the number of times they enter the game to start the ninth inning with no runners on base, “the easiest situation for a reliever to face,” according to Schecter, “even with just a one-run lead.” Thanks to Tom Ruane of Retrosheet, “if the home team starts the ninth inning with a one-run lead, it will win roughly 85 percent of the time ... Start the ninth inning with a two-run lead, and you’ll win about 93 percent of the time; with a three-run lead, it jumps to a 97 percent win rate.”

Hoffman has been used in the latter situation 142 times over the course of his career, while Fingers (11), Sutter (16) and Gossage (14) were rarely used in this manner.

I also covered optimal usage patterns, quoting Bill James from The New Bill James Historical Baseball Abstract and Whitey Herzog in You’re Missin’ a Great Game.

James argued that “when you’re defining the most effective use of your closer, you should START with the tie games. That is when the impact of a run saved is the largest, when the game is tied. If the manager wants to win as many games as possible, he can get a lot bigger bang from his relief ace by pitching him in tie games than he can by pitching him with a three-run lead--eight times bigger. As percentage baseball goes, 800% is a big percentage.”

Herzog wrote, “It’s better to have your closer go two innings every other day than one inning every day.”

After discussing newer and better metrics to judge relievers, including Win Probability Added (WPA) and Fielding Independent Pitching (FIP), I concluded my essay by stating, "When it comes to measuring relievers, there’s a lot more to consider than just the raw number of saves recorded. Even Jerome Holtzman, save his soul, would agree with me on that point."

Be sure to order The Hardball Times Baseball Annual 2009. While you're at it, pick up a second book for your dad or another baseball fan in your family. It will make an excellent holiday present for you and your loved ones.


I am waiting for some low budget team to try to keep reliever salaries down by refusing to designate a closer and instead using the bullpen to get the best matchups. So nobody would accumulate many saves but they might get maximum benefit by bringing their best reliever into a tie game in the 7th rather than holding him back until the 9th.

The Rays seemed to creep in that direction in 2008, and it is why I continue to hope they do not sign anyone (like Fuentes) who has the designation closer. (That is unlikely anyway.)

Wasn't that what Boston's closer-by-committee was all about a few years ago?

Some smart (or lucky) teams have already made an adjustment to the new reality.

They make a "decent" reliever into their closer (ala Todd Jones) and then have some "up and comer" be the setup guy who coincendentally pitches in a lot of tie games (e.g. Joel Zumaya when healthy).

I am still curious to the stink that JJ Putz is going to raise about being a setup guy when lo and behold he is better than the closer.

Dunno if this has been proposed elsewhere, but I wonder if the following simple rules tweak would be a Good Thing: change the definition of a save to specifically exclude 1.0 IP (or less) appearances with a 3-run lead (would not affect the existing save rules concerning facing the potential tying run). This would have the effect of reducing the number of standard 9th inning save opportunities by approximately 1/3 (i.e. 1, 2 or 3 run leads -- I suppose it's a bit less because there are probably more 1-run leads) and hopefully encourage managers to use their closers in other situations.

The save rules have been tweaked so often it's hard to tell what they mean anymore. Were K-Rod's 62 saves in 68 IP from 76 appearances with a 2.24 ERA a better feat than John Hiller's 38 saves, 10 wins, and many other effective outings among his 125 IP in 65 appearances with a 1.44 ERA?

A save meant more in Hiller's day because ace relievers came in from the 7th on, to put out fires, not just to pitch a fresh 9th. In fact, some of the best relievers were the guys who were most effective when coming in to face a threat. Their mentality and adrenalin worked best under fire.

Some of today's closers would be more effective if used to get out of jams. Some of them are good at putting out fires, but if there is no fire, when they come in for a fresh 9th, they are liable to start a fire. Todd Jones was the epitome of such an "old school" reliever. That earned his "Roller Coaster" moniker.

Though saves are overrated stats today, there is still something to be said for a guy who can get out #27 consistently. If it was so easy, there'd be more guys who get 30+ saves on a regular basis.

I do not agree with your last sentence, Teno. It is possible, even probable, that some pitchers do have trouble getting the last out. It is even possible that some failures have arisen due to a self-fulfilling prophesy in that pitchers are increasingly being told what an achievement it is and how tough it is.

But I have never seen evidence that the capacity to do it is reserved for a few elite pitchers. Historically it is only recently that such an ability was even at issue. And each year, due to injury or ineffectiveness of experienced closers, new ones emerge who seem just as able to get out #27.

As for more pitchers doing it, the maximum is pretty much 30 each year, and quite a few teams had closers with that many and more. In most cases, the reason a team did not had more to do with issues like injuries or simply the lack of good relievers rather than an inability to get out #27 specifically.

The only pitcher I know of that has proven he can pitch, yet constantly fails when they try him at closer is Arthur Rhodes.

Anyone know any other examples of a pitcher who pitched well in relief at least 2 times, but bombed at closer 2 or more times?