Last night, Los Angeles Dodgers pitcher Caleb Ferguson picked up his first career save, and it happened to come on his 22nd birthday. Dodgers starter Alex Wood went six strong innings, allowing just one run on a second-inning home run to Pittsburgh Pirates catcher Elias Diaz. Ferguson took over in the seventh and pitched the final three innings to earn the save.
Oh yeah, and the Dodgers scored 17 runs.
The MLB rules allow for a save for a pitcher who pitches the final three innings or more of a victory and is not the winning pitcher, regardless of the margin of victory. As I watched the game, I wondered what the largest margin of victory in a save was in Dodgers franchise history. I knew the 16-run victory wouldn’t come close to the MLB record, because I vividly remember Wes Littleton getting a save for the Texas Rangers in their 30-3 victory over the Baltimore Orioles in 2007. But I suspected that it might come close to the Dodgers record, and I was right:
I tweeted that out after the bottom of the eighth inning, and Ferguson pitched a 1-2-3 ninth to complete the 16-run save. But as I looked closer, I noticed something a little funky about Ed Roebuck‘s game. Here’s a screenshot from the Baseball-Reference Play Index from this morning, with Ferguson and Roebuck side by side:
Do you see what I see? Roebuck only pitched 2.2 innings and the Dodgers won by 16 runs — how in the world did he get a save?
My first suspicion was that Roebuck had entered the game in a save situation and then the Dodgers exploded for a bunch of runs once he was already in the game, but that wasn’t quite right. When Roebuck replaced 20-year-old Don Drysdale in the top of the seventh, the Dodgers were leading 9-3 and there were runners on first and second. Had it been 9-4 and bases loaded, that’s a save situation — if the potential tying run is on base, at the plate, or in the on-deck circle, it is a save situation regardless of how many outs are left in the game. But in this case, due to the score and the bases occupied, the potential tying run was still several batters away. So how did Roebuck get a save?
Well, as you probably recall, the save statistic didn’t exist yet in 1957, so Roebuck’s save was a retroactive thing anyway. Sportswriter Jerome Holtzman created the save in 1960 (it became an official stat in 1969), and under the original rules, there was no innings limit. Basically, if you came into a game with the lead and finished the game, you got a save. The save rule as it currently stands has been in place since 1975 — but when the save became an official stat in 1969, saves were retroactively calculated for previous seasons. So for any season before 1969, the original 1969 save rule applies. In summary, in 1969, Ed Roebuck got a save for a 1957 game in which he would not have gotten a save in 1975 or 2018.
So Caleb Ferguson stands alone in “largest margin of victory in a save by today’s rules,” but that’s not really a thing. Ed Roebuck, though, got me interested in other extreme saves. So here are a few interesting quirks about “Holtzman’s Folly,” as Keith Law calls it:
Largest Margin of Victory in a Save: Wes Littleton, August 22, 2007
As mentioned above, Littleton got the save for the Rangers in a 30-3 victory. To be fair, it was “only” 14-3 when he entered the game, but the Rangers put up 10 runs in the eighth inning and six in the ninth. It was Littleton’s second of three saves in his three-year career.
Worst Performance to “Earn” a Save: Joe Willis, April 16, 1912
With the St. Louis Cardinals leading the Chicago Cubs, 20-1, Willis came in to pitch after starter Bill Steele allowed just one run in six innings. Willis pitched two innings, allowing four hits, four walks, and four runs. After the Cubs pulled to within 15 with a three-run top of the eighth inning, umpire Jim Johnstone called the game. Here’s how Irving Ellis “Sy” Sanborn of the Chicago Tribune described it:
In an exhibition that would have made a team of grammar school boys afraid to go home, even in the dark, Chicago’s Cubs humiliated themselves to the tune of 20 to 5 today in the final game with the Cardinals. It might have been a lot worse than that if Umpire Johnstone, inspired by humane motives, had not called the game at the end of Chicago’s half of the eighth inning.
The ostensible reason for curtailing the combat was to permit the two teams to catch their trains, but as neither club was booked for an early getaway the humane grounds were the only substantial ones for stopping the game.
About Willis and Cubs pitcher George Pierce — who allowed 10 runs in 3.2 innings — Sanborn wrote:
The game being sewed up, both managers accepted the opportunity to convert it into a spring training trip exhibition by replacing regulars with recruits until there had been a total of twenty-eight men in the mess and the score books looked as if they had been taken through an epidemic of smallpox complicated with eczema.
There were two of the wildest left handers you ever saw on the slab when the game ended for Manager [Roger] Bresnahan, for when he saw how wild Pierce was he took out Steele and sent in Joe Willis — no relation to Vic. It is possible that Pierce and Willis may turn out to be great pitchers when somebody tames them. They have plenty of pitching goods but in their present untamed condition they are merely dangerous to the skulls of opposing batsmen.
Well, I have news for you, Mr. Sanborn: In 57 years, Willis is getting a save for this performance!
Best Hidden Great Performance: Fred Gladding, July 30, 1969
In the first year of the save statistic, Gladding earned a save for the Houston Astros when they beat the New York Mets, 16-3. Gladding pitched 2.2 innings, and at first glance, it looks like another Ed Roebuck Special. But when you dig a little deeper, you find that Gladding had the best save ever in a game decided by 10+ runs. Gladding’s Win Probability Added of .315 is the best among all saves in blowouts.
Don Wilson pitched six innings for the Astros and departed with a 5-2 lead. After Jim Ray faced four batters and retired just one of them, Gladding entered with a 5-3 lead and runners on second and third with one out. He promptly struck out Ken Boswell and got Cleon Jones to ground out to end the threat. After Gladding retired Art Shamsky, Wayne Garrett, and Ed Kranepool in order in the bottom of the eighth, the Astros broke out for 11 runs in the top of the ninth, highlighted by grand slams by Denis Menke and Jim Wynn — and an RBI single by Gladding himself. Gladding then retired the Mets in order in the bottom of the ninth to lock down a well-deserved save.
Worst Performance in a Save That Shouldn’t Have Been a Save: Hi Bell, July 29, 1933
The New York Giants beat the Boston Braves, 6-5, on a walkoff home run by player-manager Bill Terry. For whatever reason, the official scorekeeper chose to award the win to starting pitcher Freddie Fitzsimmons, who had departed with two outs in the seventh inning. The other options were the three relievers: Dolf Luque, who allowed a run and three hits in 1.2 innings; Watty Clark, who faced just one batter (and struck him out); or Bell, who entered with a one-run lead, waked his first batter, allowed an RBI single to tie the game, and then got the third out of the inning. Under the current rules, the victory in this game would have gone to Bell, the pitcher of record in the bottom of the ninth when Terry homered to win the game. But the official scorer had more leeway back then, and he chose to give the win to the pitcher who pitched the best — Fitzsimmons. And when the save became a stat in 1969, any pitcher who finished a victory and wasn’t the winning pitcher got a save. So voila, Bell gets a save when all he did was blow the lead.
Worst Modern Save Performance: Scot Shields, April 27, 2004
At first glance, Shields’ performance doesn’t look too bad. He pitched 2.1 innings, allowing just two hits and one run. But by Win Probability Added, it was the worst game in which a pitcher actually “deserved” a save (second-worst only to Hi Bell’s non-save above). Why?
Well, John Lackey started the game for the Anaheim Angels, and through six innings he had allowed zero runs to the Detroit Tigers and held a commanding 5-0 lead. Lackey got the first two outs in the seventh, but then he allowed a double to Carlos Guillen and walks to Carlos Pena and Eric Munson. When Shields entered the game to face Brandon Inge, the Angels had a 94 percent chance of winning. Four pitches later, after Inge hit a grand slam, the Angels’ chances were down to 72 percent. A three-run double by Darin Erstad in the top of the eighth got the win probability back up to 97 percent, and Shields pitched pretty effectively after the grand slam to earn the save in a 10-4 victory. But by WPA, he did less to help the team win than anyone else on the field.
Easiest “Traditional” Save: Ed Vosberg, April 19, 1996
Like Littleton above, Vosberg’s save came in a blowout victory for the Rangers over the Orioles. With two outs and a 10-7 lead in the top of the eighth, Rangers manager Johnny Oates called on lefty Vosberg to face lefty Tony Tarasco. Orioles skipper Davey Johnson countered with pinch-hitter Jeffrey Hammonds, who flied out to right field on the second pitch.
It’s tough to know exactly when Oates had closer Mike Henneman sit down in the bullpen, but it was probably when Juan Gonzalez hit a two-run double off of Jesse Orosco to make it 12-7. Or maybe it was Mickey Tettleton‘s sacrifice fly. Or Dean Palmer‘s two-run homer. Or Kevin Elster‘s RBI single. Or Dave Valle‘s RBI single. Or the bases-loaded walks to Will Clark or Gonzalez or Tettleton or Palmer. Or the sac fly by Rusty Greer. Or the grand slam by Elster.
But at some point during the 16-run (!!!) bottom of the eighth, Oates decided, “I think Vosberg can finish this one off.” And after entering the game in a traditional save situation with a three-run lead, Vosberg ended up earning the save in a 26-7 victory.
Quickest Save: Mitch Williams, April 28, 1989
Paul Kilgus got the start for the Chicago Cubs and went eight shutout innings against the San Diego Padres. In the top of the ninth, Kilgus got Tony Gwynn to fly out to left, then allowed a double to Jack Clark. Benito Santiago reached on an error by third baseman Curt Wilkerson, and after a fielder’s choice by Carmelo Martinez, Clark scored on an RBI single by Luis Salazar.
With two outs and runners on first and second, clinging to a two-run lead, Cubs manager Don Zimmer decided that 112 pitches was enough for Kilgus and called on closer Mitch Williams, who had been acquired from the Rangers in the offseason (along with Kilgus, Wilkerson, and others) for a package headlined by Jamie Moyer and Rafael Palmeiro. Williams must have thought 112 pitches was plenty, too, because he didn’t throw any — he promptly picked Martinez off second base to end the game.
In the pitch-count era, that remains the only zero-pitch save in baseball. But, amazingly, it wasn’t the only save with zero batters faced … by Mitch Williams … that year! There have been just four zero-batters-face saves in history, and two of them we by Williams in 1989. On September 11, 1989, Williams entered with two outs in the ninth and picked Jeff Huson off first base to end the game, but that was only after throwing a first-pitch ball to Montreal Expos catcher Nelson Santovenia.
The save rule is one of my favorite and least favorite things in baseball. I think it can be relatively meaningless and I think people’s obsession with it leads to poor managerial choices (see Buck Showalter and Zach Britton, for example). But it also leads to enormously fun quirks like the ones listed here, and that’s not a bad thing.