Zero, Interrupted: Pulling Starters From No-Hitters

Trevor Williams opened his 2018 season with six no-hit innings before being pulled from the game. (via Editorsaurus)

It was on the third day of baseball this season, still not out of the month of March, that Minnesota Twins manager Paul Molitor set a tone. His starting pitcher, Kyle Gibson, had thrown six no-hit innings against Baltimore, but had walked five and was sitting at 102 pitches. Molitor took him out. Gibson’s relievers could not hold the no-hitter or the shutout, but Gibson and Minnesota did get the 6-2 win.

The next day, Pittsburgh Pirates manager Clint Hurdle reinforced the pattern. His starter in the first game of a Detroit doubleheader, Trevor Williams, had pitched six no-hit innings, but likewise had walked five. Despite a pitch count of just 85, Hurdle wielded the hook. Again, the no-hitter didn’t last in the bullpen’s hands, but the 1-0 shutout did.

Over the following five weeks, it would happen three more times. Jarlin Garcia of Miami, Walker Buehler of the Dodgers and Domingo German of the Yankees all pitched six no-hit innings, then watched as relievers opened the seventh. A few weeks later, still in the month of May, Nathan Eovaldi joined the season’s six-and-out ranks.

Buehler’s Dodger teammates did keep the Padres out of the “H” column, giving him a piece of a combined no-hitter. This only underscored how something quite unusual was happening.

Historically, a no-hitter has been a demonstration of a pitcher’s dominance, even of his invulnerability. Over 140 years of professional play, it has developed a mystique—one that is dwindling. Today’s managers do not feel an obligation to let a starting pitcher work past his established limits to pursue this piece of baseball history.

What does this new trend look like? What were the old trends that it is supplementing or superseding? I had these questions, and I found some answers, and more.

The History of Interrupting History

When searching for interrupted no-hitters, I chose five innings as my cutoff point. It strikes a good balance between the awareness and the likelihood of the developing no-hitter. It also trims out many games where an early injury to the pitcher causes his removal, rather than the discretion of the manager. This doesn’t mean that late injuries don’t play their role: we will see that they still do.

My search for no-hitters carried back to 1908, the current limit supported by Baseball-Reference. It turns out 1908 is a good year to look for no-hitters. There were six completed that season, including Cy Young’s third no-hitter, an Addie Joss perfect game thrown in the homestretch of a torrid AL pennant race, and Hooks Wiltse’s Fourth of July 10-inning gem that would have been a perfect game had he not plunked the opposing pitcher with two gone in the ninth.

That last item is not merely colorful detail: it shows the different mindset of the day. The Phillies’ manager let hurler George McQuillan bat for himself in the ninth inning of a scoreless tie, where virtually any manager of our century would yank him for a better chance to get somebody, anybody, on base. The pitcher was still effective, and his team wasn’t losing: ergo, he stayed in.

Despite this mindset, there was an interrupted no-hitter that year. Gus Dorner of the Boston Doves held the Giants hitless for six and a third innings on April 28, but it was no gem. He walked six, struck out none, and allowed an unearned run in the fifth frame via two walks and an error. He came out after a batted ball struck his right foot in the seventh. While we might imagine that his manager was glad for the excuse, the interruption came from injury, not ineffectiveness.

(The Giants, down 2-1 when Dorner departed, would come back to win the game. The winning pitcher: Hooks Wiltse.)

From 1908 through 2017, there have been 90 instances where pitchers were removed while throwing no-hitters (of at least five innings). In that same period, for comparison, there have been 228 no-hitters completed by the starting pitcher*. So far in 2018, there have been six interruptions and two complete-game no-hitters. (The Buehler interruption did lead to a combined no-hitter.)

* This excludes games that ended before the regulation length. These used to be counted as no-hitters, but no longer are.

The great majority of interruptions come early in the time frame. This is natural, as no-hitters getting broken up takes away opportunities for managers to interrupt them. Looked at from that angle, the prime time to interrupt a starter’s no-hitter would be during the seventh inning, in the second row on the table below.

A Hardball Times Update
Goodbye for now.
Lengths of Interrupted No-Hitters, 1908-2017
Innings No. Innings No. Innings No.
5 37 5.1 5 5.2 1
6 23 6.1 6 6.2 4
7 9 7.1 1 7.2 1
8 2 8.1 0 8.2 1
SOURCE: Baseball-Reference

While current interruptions can be blamed on workload fears, historically performance concerns drove many instances of starters being hooked out of no-hitters. In those 90 interrupted games, 26 starters gave up a total of 44 runs, 30 of them earned. Contrast that to the 228 completed no-hitters, in which 14 starters allowed 15 runs, all but two unearned. The difference grows sharper when you consider that the completed no-hitters have almost a four-to-one advantage in innings pitched over the interrupted ones (2,053 IP to 522.2 IP).

Other performance data follow the pattern. Walks are 333 in interrupted no-nos versus 578 in completed ones, working out to a 9/4 rate ratio. Hit by pitches are 24 to 37, a 5/2 rate ratio, while wild pitches are 17 to 21, more than 3/1. Strikeout rates are virtually identical: 408 in 522.2 innings against 1,604 in 2,053. However, interrupted no-hitters are front-loaded toward modern times, when strikeout rates are higher, so this even match is really an underperformance for the interrupted starters.

This leads us to the historical distribution of no-hitter interruptions. They are more common today than in the past, though there are factors others than starter endurance influencing that. There are more games played in a current season than in the past, by almost a two-to-one ratio for most seasons before expansion in 1961. The frequency of no-hitters has fallen and risen, and if there are fewer no-hitters thrown, there should be fewer that could be interrupted.

The table below tracks both interrupted and completed no-hitters for the period being studied, in 10-year increments.

Interrupted and Completed No-Hitters, 1908-2017
Years Interrupted Completed
1908-’17 6 31
1918-’27 0 11
1928-’37 0 6
1938-’47 1 14
1948-’57 2 17
1958-’67 7 24
1968-’77 9 37
1978-’87 8 15
1988-’97 19 25
1998-2007 12 14
2008-’17 26 34
SOURCE: Baseball-Reference

After a flurry in the Deadball Era, there was not an interrupted no-hitter for 30 years, between Babe Ruth being pulled after six in a May 1916 game and Art Herring getting the hook after five in June of 1946. The dearth of actual no-hitters can only explain this partially. Something else changed.

My fill-in hypothesis is that it’s a result of the shift from a low-scoring to a high-scoring environment. In a time when hits were scarce and the strategic emphasis was on manufacturing runs, a pitcher allowing baserunners on everything but hits might have felt more dangerous in what was probably still a close game. Once the Ruthian bombardments began, a starter allowing several walks but no hits would have felt comparatively more dominant, less deserving of a precautionary hook.

That mindset eroded gradually as the boom era of the 1920s and ’30s faded into history, and the low-scoring 1960s arose. From there, the tide of more careful use of pitchers began carrying the movement. The last five years, though, have seen a further acceleration.

Interrupted and Completed No-Hitters, 2007-2018
Year Inter’d Comp’d Year Inter’d Comp’d
2007 3 3 2013 1 3
2008 2 2 2014 4 4
2009 0 2 2015 5 7
2010 2 5 2016 5 1
2011 2 3 2017 4 1
2012 1 6 2018 (to 6/8) 5 2
SOURCE: Baseball-Reference

The record for interrupted no-hitters in a season is six, set in 1991, a year with five completed no-hitters). The 2018 season matched that number on May 30, a date by which there had been two completed no-hitters. This lopsided ratio tracks fairly well with that for the last two years.

Perhaps the true peak of the trend came not this year, but last. On April 15, 2017, Sean Manaea was lifted after five innings of no-hit, two-run ball in a game his Oakland A’s lost. The next day, Miami’s Dan Straily was pulled after five and a third, having allowed five walks but zero hits or runs. Two days after that, Straily’s teammate Wei-Yin Chen pitched seven no-hit innings with just two inconsequential walks, and watched the eighth from the bench. Or perhaps a clubhouse TV, if he was ticked off enough at manager Don Mattingly.

Highs and Lows

For an occurrence as uncommon (at least formerly) as this, the trivia surrounding it is attractive. How bad can a pitcher be before the manager gives up and hands the no-hitter to somebody else? How good can the pitcher be and still get relieved? I have a few answers to such questions.

If we use Win Percentage Added (WPA) as our benchmark, the worst interrupted no-hitter was thrown by Dennis Blair in June of 1975. His five-inning effort culminated in a fifth with two walks, a hit-by-pitch, an error, and two wild pitches that produced three runs, one earned. His WPA for the game was -0.112. The only other negative no-hitter was by Charlie Hough on Opening Day in 1985. He tallied -0.019 WPA for six innings that encompassed eight walks and two runs, one earned.

The highest-scoring interrupted no-hitter was crafted by Aaron Harang in April of 2014. He posted a 0.464 WPA in seven innings—which included six walks and only five strikeouts. This exposes the problem with using WPA in these cases: all scoreless innings register the same in WPA terms. The score then becomes dependent on how little the pitcher’s own team scores. WPA has some use in measuring poor no-hit efforts, but not great ones.

If a developing no-hitter is not a great pitching performance, it’s probably walks that make it so. Twice, a pitcher has been pulled from a no-hitter after handing out 10 free passes. First was Joe Engel of the Washington Senators, who walked 10 Red Sox (and hit another) before getting his marching orders with one out in the seventh. Engel left with the score tied, 1-1. His reliever, Walter Johnson, held down Boston while Washington tallied the game-winning run.

The second instance came in the latest that a pitcher has ever been pulled from a no-hitter. In 1967, Steve Barber of the Orioles went his first eight innings dodging raindrops in the form of seven walks and two hit batters against the Tigers. His O’s pushed across a run in the home eighth, but Barber could not make it stand up. Three walks in the ninth, surrounding a run-scoring wild pitch that tied the game, exhausted manager Hank Bauer’s patience. Barber was pulled with two outs in the ninth.

When his reliever, Stu Miller, coaxed a ground ball to short, second baseman Mark Belanger booted the toss, and a second run came home. Miller got the final out, but the Tigers retired Baltimore in order, ending the game. It was a combined no-hitter, and a loss for both Baltimore and Steve Barber. Of all the no-hitter interruptions, this is one of the most unfortunate, and easiest to justify.

A 10-walk no-hitter effort is not confined to interrupted attempts. Jim Maloney gave up 10 walks in a complete-game 10-inning no-hitter he pitched in 1965. I will have more to say about Maloney and his experiences with no-hitters soon.

Another sign of a sub-par pitching performance, especially nowadays, is lack of strikeouts. Four interrupted no-hitters have featured a starter failing to strike out anybody. The first was by Gus Dorner, in the 1908 game already covered. The most recent was by Toronto’s Luis Leal, who pitched five hit-less, K-less frames on May 14, 1983 before an early shower, literally. A long rain delay intervened to end his day, and Toronto completed an 8-1 thrashing of Cleveland without him.

The other two, weirdly enough, were teammates doing it, not just in the same season, but in the same month. First was Bill Stafford of the Yankees pitching on May 1, 1962. He threw four hitless innings against the White Sox, pulled a muscle slightly in the top of the fifth while batting, then threw a hitless fifth. He then mentioned the muscle pull, and manager Ralph Houk removed him. Reliever Rollie Sheldon held the no-no for three innings, losing it, but not the game, in the ninth. The lack of strikeouts was obviously not crucial to Stafford’s departure.

Three weeks later, Whitey Ford took the hill against the Angels. He pitched seven hitless innings, though he yielded a first-inning run on a walk, steal, and two productive outs. Houk then lifted him for a pinch-hitter after he complained of a pain in his pitching shoulder. This hook was slightly more strategic: the game was tied at the time. The no-hitter ended up a 12-inning one-hitter that New York pulled out 2-1.

Of the four pitchers pulled during no-hitters in which they struck out nobody, three hooks were motivated by injury causes, and the fourth by an arm-stiffening delay. If lack of whiffs motivates managers to interrupt no-hitters, it makes no obvious impact.

To take the opposite angle, the most strikeouts in an interrupted no-hitter is 11, by Cleveland’s Trevor Bauer on April 9, 2015. This came in six innings, during which he also allowed five walks, and piled up 111 pitches. You’d have to be very old-school to castigate Terry Francona for this move. Bobby Witt had 10 strikeouts in five-innings of no-hit ball on April 17, 1986. He walked eight. His pitch count is not available, a relief for those with weak hearts.

Looking at walks, there have been five times when a pitcher was lifted from a no-hit, no-walk effort. Andy Benes (1991) and Ben McDonald (1997) did it for six innings, though McDonald hit a batter and Benes had a batter reach on an error. Bob Knepper (1986) and Daniel Norris (2015) did it for five innings, and both were perfect game efforts while they lasted. Knepper’s came in the last game of the Houston Astros’ regular season, before they faced the Mets in the NLCS. Clearly he was not going to be taxed in a tune-up for the playoffs.

The most famous instance was in 2016, when Rich Hill of the Dodgers pitched seven perfect innings in Miami. Manager Dave Roberts, concerned about a blister on the pitching hand of a starter who had missed a lot of games that year due to blisters, took away the ball. Reactions to Roberts’s action were all over the map, but the event brought the trend of interrupting no-hitters to the attention of fans.

The Reinforcements

The starting pitcher is gone, but the no-hitter remains, for the moment. It falls into the hands of a reliever, or more than one, to lose as most no-hitters are, or to preserve. How well do they manage in this role, where the team’s fans are anything but pleased to see them?

Out of the 90 instances since 1908, nine times the bullpen has completed the no-hitter. Twice the pen supplied 12 outs, five times it got nine, once it got six, and once—as related above—was Stu Miller’s one-out relief of Steve Barber. Three innings appears to be a butter zone for no-hitter relief, with a success rate twice that of two-inning efforts and four times better than four-inning attempts.

The two attempts at a one-inning close-out of a no-hitter both failed. Those two games, in 1970 and 1974, are strongly linked. The starting pitchers, Clay Kirby and Don Wilson , had both allowed runs during their eight no-hit innings, and were pulled with their teams trailing. The real coincidence is that the same manager pulled the pitcher in both games: Preston Gómez.

Gómez took heat for the move both times, especially for the first one. The San Diego Padres in 1970 were a second-year expansion team, and could have used the good publicity of a no-hitter. Gómez decided they could use the win more. (They’d finish the season 63-99, an 11-game improvement over ’69.) They got neither, and to date the Padres have never had a no-hitter, solo or combined.

When given a no-hit, no-run game, the bullpens preserved the shutout 25 out of 64 times. This includes six of the no-hitters they successfully preserved, but not all. Three times relievers were handed no-hitters with at least one run on the board. They were starts by Blue Moon Odom on July 28, 1976, Jered Weaver on June 28, 2008, and the well-trodden Steve Barber game.

Sometimes, of course, the bullpen can’t preserve the starter’s great performance. Sometimes they can’t get close. There have been inherited no-hitters that went south in truly impressive fashion.

The greatest number of hits allowed in an inherited no-hitter is 11, by the Texas Rangers bullpen that took over Sonny Siebert’s five-inning no-no on September 22, 1973. This came, however, in a 14-inning game, so giving up 11 hits (and five runs) isn’t as disastrous as it originally appears. A better candidate is Tommy Boggs’s six-inning no hitter for Atlanta on September 6, 1982. Five relievers labored to give up nine hits and eight runs in the final three innings, turning possible history into an 8-2 Giants romp.

Eight is the greatest number of bullpen runs surrendered in an interrupted no-hitter, though the Boggs game has company. When Sean Manaea left an April 15, 2017 game after five hitless innings, the Oakland bullpen gave up eight runs in his absence. Manaea had allowed two during his start, so he had less than impeccable cause for complaint. He did learn that if you want it done right, you should do it yourself, so he pitched a complete-game no-hitter on April 21 this year.

Celebrity Interruptions

All major-league pitchers are notable, but some are more notable than others. The surprise we feel when a journeyman gets detoured on the way to a no-hitter is that much greater when it’s an ace instead. I have already mentioned Babe Ruth and Whitey Ford being stopped short, but there have been other big-name pitchers enduring this indignity.

Either the biggest or the smallest surprise on the list would be Nolan Ryan, lifted after five no-hit innings for the Texas Rangers on April 9, 1990. Over 91 pitches, he walked four and struck out four. Is it no surprise that the Ryan Express was putting up zeroes yet again, or a huge surprise that he submitted to leaving the game? Little matter: he would record his sixth career no-hitter two months later, and his seventh the following season.

Then there’s Greg Maddux, late in his career but still formidable. He threw six no-hit innings for his Dodgers on August 3, 2006, in a Maddux-like economical 72 pitches. Alas, the rain falls on old aces too: a 46-minute rain delay took him out of the game. Maddux never did have a complete no-hitter, which may be a bigger surprise than Nolan Ryan leaving a no-hitter without his heels leaving grooves in the turf.

Stephen Strasburg is on the list, lifted after 6.2 no-hit innings in a July 2016 game. Cole Hamels got interrupted after six in September 2014. Earlier that same year, Tim Lincecum was removed after five, a month before completing his second no-hitter (shades of Nolan). Stars of an earlier generation with an interruption include John Tudor (six innings in 1988) and Charlie Hough (six frames in 1985, but with eight walks and two runs).

Kent Mercker had a pair of no-hitters interrupted at six innings, once in 1991 and again in 1993. Both were his Braves against the Padres; both ended 1-0 Atlanta. He would put those disappointments behind him and complete a no-hitter in 1994. David Cone also went through two interruptions: once after seven innings in 1996, and again after five the next year. He improved on Mercker’s mark two years later by finishing off a perfect game.

I cannot let this section close without recognition of a pitcher who isn’t on the celebrity level of a Maddux or even a Cone, but whose career is bound up in the matter of no-hitters, interrupted and completed. Jim Maloney’s whole baseball life was a war against injuries, and this conflict hit several climaxes during his pursuit of no-hitters.

His first near-miss was on April 18, 1964, when he threw six no-hit innings versus Sandy Koufax. A muscle strain, though, caused manager Fred Hutchinson to pull him, saying that one doesn’t take chances with an arm like Maloney’s. In June of 1965, he’d no-hit the Mets through 10 innings, but lose it and the game in the 11th. Two months later, he’d pitch 10 hitless innings again against the Cubs, though he struggled to do so, allowing 10 walks. His Reds got him a run this time, and he had his first official no-hitter.*

* The one he lost in the 11th was considered official at the time, but has since been expunged by the same scoring reform that removed no-hitters that went fewer than nine innings.

Two years after that, he had another no-hitter working against the Pirates through six innings. In the top of the seventh, he hurt his ankle running out a triple. After a pop-out and a walk in the home seventh, the ankle was too painful to let him continue, and Maloney took himself out of the game. That was two out of a potential four no-hitters he had lost due to injury.

The bug would bite yet again on April 30th, 1969. He pitched eight innings of no-hit baseball against the Astros, but in the bottom of the eighth badly pulled a groin muscle running out an RBI double. He stayed in to be driven home himself—Cincy’s 10th run of the day—but was left barely able to walk. He was still able to pitch, and this time refused to come out. He got two outs, walked Jim Wynn (an excruciating self-inflicted torment), then struck out Doug Rader to finish the no-hitter.

It’s tough to imagine a manager permitting that today. It’s also tough to imagine a pitcher having to face getting yanked out of a third no-hitter due to injury. There is pain, and then there is pain. That day, Jim Maloney chose the pain he was willing to endure.


Jim Maloney’s tale emphasizes a theme running through the history of interrupted no-hitters. While there have been instances of starters being pulled for shaky control or high-leverage stakes, very often the move was made when the pitcher somehow got hurt. The recent surge of incidents, though, comes from managers’ worries about excessive fatigue.

One can argue that this is merely an expansion and modernization of the old injury concern, depending on how one feels about the trend of limiting starters’ exposure. What is tougher to argue is that we should expect the trend to strengthen. Barring a surprising new turn in pitcher usage, more no-hitters are going to be handed off to the bullpen. The solo no-no could become an endangered species.

Could this lead to an increase in no-hitters? Could fresh pitchers throwing all-out for an inning at a time get more no-hitters from five, six or seven innings to nine? Whether it happens or not, it would be hard to claim authoritatively that bullpenning no-hitters was raising or lowering the rate. The sample sizes would be too small, at least until we had decades of results. By that time, some fresh trend would likely have arisen to confound the analysis.

Instead, we will just have to watch the games, and argue the point among ourselves. I think we can manage to do that.

References and Resources

A writer for The Hardball Times, Shane has been writing about baseball and science fiction since 1997. His stories have been translated into French, Russian and Japanese, and he was nominated for the 2002 Hugo Award.
Newest Most Voted
Inline Feedbacks
View all comments
5 years ago

I personally hate the idea of most of these interrupted potential no-hitters. These guys are likely to never get such a change again, so you almost always need to let them go for it. Unless there’s an injury, a long rain delay, a ridiculously high pitch count, or you’re actually losing the game (or tied after 9 innings with a high pitch count, or tied late with an good chance to score by pinch-hitting for the pitcher in the NL), keep him in!

Mike Matheny of the Cardinals recently handled it the right way. Michael Wacha had a no-hitter through 8 innings with a pitch count over 100, but Matheny sent him back out for the 9th. He then immediately pulled him after he gave up a leadoff single.

5 years ago

Since you brought up Maloney’s 4/30/69 No Hitter, you should mention that the team he beat, the Houston Astros, beat the Reds the next day with a No Hitter of their own, pitched by Don Wilson,. Now that’s an amazing stat!