Why a 21-K game is MLB's best record that can actually happen

G Fiume/Getty Images

The coolest individual accomplishment we could reasonably hope to see this baseball season is a 21-strikeout start.* This is just my opinion, which you might not share, and ordinarily that disagreement would be fine and we'd walk away friends.

But the weird thing about this particular record is that our opinion of it could have a tremendous influence on whether it gets set. So I'm going to have to convince you, and if I do you'll be glad I did, and then it'll be up to you to convince three friends, and then they each convince three friends, so that one day we all get to share the best day baseball can presently offer.

The strikeouts-in-a-nine-inning-game record has a fun history: Somebody sets it, then (usually) decades pass during which a few pitchers tie it, giving the impression that whatever the record is at the time is an almost impenetrable barrier, able to be reached but not surpassed. Then it gets surpassed!

By 1933, four pitchers in the modern era had struck out 16 in a game. Then Dizzy Dean bested it, and in 1936 an 18-year-old Bob Feller matched it. Two years later, on the cusp of superstardom, Feller broke the record, striking out 18 in his final start of the 1938 season.

The record then held for more than three decades -- Sandy Koufax matched it twice -- until Steve Carlton whiffed 19 in 1969. Then that record held for almost two decades (Tom Seaver and Nolan Ryan matched it), until a young Roger Clemens, lightly accomplished at the time but on the cusp of his own superstardom, punched out 20 in 1986. It has been 32 years, and 20 has been repeated by Clemens, Kerry Wood, Randy Johnson and Max Scherzer. We're due.

The means

I will allow that a five-homer game by a batter would probably be cooler than a 21-strikeout start. But a five-homer game is outlandishly unlikely. Batters don't even always get five plate appearances in a game, let alone five at-bats with five pitches to hit. In the history of the sport, there have been only four plate appearances that could have resulted in a fifth home run of a game: two for Mike Cameron in May 2002, two for Lou Gehrig in June 1932. No batter in major league history has even hit five homers in a row across games. That means a five-homer game would likely have to involve somebody doing something totally unprecedented, and fitting it within the narrow confines of a single game, with no margin for error or an intentional walk.

By comparison, a 21-strikeout game fits perfectly within the established limits of human performance and baseball's natural variation. Scherzer and Clemens each struck out 20 batters with at least one out to go (both got ground outs), and Johnson, Ramon Martinez and Wood all entered ninth innings with 18 strikeouts. Corey Kluber and Johnson both struck out 18 batters in eight innings before being pulled from starts. We've been close, so close. Indeed, there have been 26 starts in which a pitcher has struck out at least 10 batters and maintained at least a 21-K pace before being pulled from the game, and scores more pitchers have been pulled with shots at 21 still mathematically intact.

And, most importantly, we've seen pitchers strike out 21 (or more) batters across 27 outs -- "hidden" 21-K games, spread out over multiple outings. They've mostly been relievers, such as Ken Giles (23 K's in 27 outs in 2016) and Edwin Diaz (24 K's in 27 outs the same year). (A minor leaguer named Jacob Webb even had 25 K's over 27 outs in Class A.) Kluber did it, striking out the side in the first inning after his 18-K start, for 21 K's across nine innings. The sequence of events necessary for 21 strikeouts in a game has been proved possible.

Almost every record is the product of its environment. Nobody is going to hit 73 homers in a low-homer environment, just like nobody saved 62 games before modern bullpen usage solidified. This is a strikeout era. In the past week, we've seen three credible 21-K alerts: James Paxton had 16 in seven innings, with six outs to go; Gerrit Cole had 12 through six innings, but ended up with only 16; and Scherzer had 15 in 6⅓ innings before he was pulled. The means are there.

The motive

Remember the list of names up there? Clemens, Seaver, Koufax, etc.? There's not a dud in the bunch, unless you count the unfulfilled expectations we had for Wood. And, OK, I cheated and left out Don Wilson, a very good pitcher whose potential was ruined by injuries and a mysterious early death. But Wilson (who struck out 18 in 1968) and Wood were both fantastically exciting young starters with incredible potential, while the rest of our record-holders were all-timers, inner-circle dudes. Lots of pitchers have thrown no-hitters -- some even have thrown perfect games despite being not actually very good. But in the live ball era, every pitcher who has ever held the strikeouts per game record has been good; nine out of 11 were either Hall of Famers or Roger Clemens.

Further, every start that has matched or set the strikeouts per game record is visually extraordinary, punctuated by wildly varied flails and increasingly emphatic umpire punchouts. Wood's 20-strikeout game -- celebrated this week on its 20th anniversary -- has been endlessly picked apart, GIF'd, and described with hyperbolic language that still seems to come up short. Johnson got 36 swinging strikes in his start, and his highlight reel is a series of bats flying, batters falling over, and the catcher chasing down third strikes he simply couldn't catch. Watching Clemens' first 20-strikeout game is like watching a movie in which the sound and the visuals aren't quite synced up; the batters' swings are all awkwardly late, like a bad edit.

Clemens, upon punching out his 20th, got a long ovation, got mobbed by his teammates after the final out, and was well known to hold the record throughout his career. It's a good record. But it has arguably been a second-tier record, trailing the more common and arguably less impressive no-hitter and perfect games for tension and headline size.

Historically, this has been justifiable: A pitcher needed to be dominant to set the strikeout record, but he didn't necessarily have to be good. Dean allowed a couple of runs when he struck out 17; Feller walked seven batters, allowed four runs and took the loss when he struck out 18; Carlton allowed four runs and took the loss when he struck out 19, likely throwing around 160 pitches. (Feller likely threw around 180.) It was certainly difficult to strike out 19 batters in a game, but without any real pitch count limits, the limiting factor -- the 27 available outs -- couldn't tell you anything about how many hits, how many walks, how many runs scored in between those outs.

But the limiting factor today is pitch counts; Paxton and Scherzer didn't run out of outs, but ran out of pitches. To strike out 21, a pitcher has to (A) strike everybody out, (B) allow few baserunners, and probably even (C) strike everybody out quickly. Feller's 18 strikeouts were 43 percent of the batters he faced, while Paxton last week struck out 59 percent of the batters he faced.

We've reached a point where, almost by definition, a 21-strikeout game also will be an overall masterpiece of dominance and effectiveness. Over the past 20 years the six 18-plus-K starts we've seen have featured a combined four runs and a combined one walk. Scherzer's start over the weekend wasn't really a threat to 21, because he wasn't throwing that masterpiece. Paxton's was -- no walks, only four hits through six innings -- until the seventh inning, when it took him 25 pitches to strike out the side. Just like that, a bid for 21 was unrealistic, on pitch counts alone, based on one "long" inning. That's how narrow the margin is, and that's why a 21-strikeout game will almost certainly rival Wood's start, and any no-hitter or perfect game you can think of, as the best start you've ever seen. In an era of GIFs, full games on YouTube, and retrospectives about GIFs and full games on YouTube, it will be an extremely durable record with endlessly appreciated highlights. It will live forever, or until somebody strikes out 22.

The opportunity

Here's where we all come in.

A 21-strikeout game will almost certainly be one of the highest pitch-count games of the year, even if it's a perfect game. The average strikeout takes a little more than 4.8 pitches, so 21 K's would take a pitcher to 101 pitches, with (presumably) six other outs mixed in. At three pitches per plate appearance, that's another 18 pitches. A single walk, or a couple of hits, or a few extra full counts, will push pitchers into the 120s, maybe the 130s.

A few weeks ago, we prophesied the coming decline in no-hitters by noting that managers almost never let pitchers go past 115 pitches anymore. The lone exception has, recently, been in no-hitter bids. There have been 11 no-hitters this decade in which a pitcher was allowed to throw at least 120 pitches, eight over 125, four over 130 and a pair over 145. The achievement of throwing a no-hitter isn't just about the pitcher, but about the manager, who can either will a no-hit bid or quash it. It's also about us, because we freak out about no-hitters, making this an achievement players and managers chase.

The same allowances have generally not been granted of 20- or 21-K pursuits. In Kluber's 18-K game, he was pulled after just 113 pitches -- less than the median no-hitter. It's easy to imagine he could have struck out the side in the ninth in 15 pitches, but 21 K's wasn't worth taking that chance or extending that leash. Paxton was pulled after 105. If they were chasing no-hitters, history tells us, they would have been allowed to go up to 130 pitches. But chasing 21 K's doesn't trigger the same rules.

That's because we don't make a big enough deal about 21 K's. A no-hitter gets an alert on our phones, sets off that superstitious routine of mentioning/not mentioning, compels the pitcher's teammates to leave him alone in the dugout between innings, and generally ratchets up the tension throughout the pursuit. That's appropriate: Paxton's completed no-hitter on Tuesday was spectacular, one of the most beautiful things we'll see this year, a historic achievement. It's just that striking out 21 batters a week ago would have been even more spectacular, even more beautiful, and undoubtedly an achievement that would be remembered by far more people for far longer.

We should be just as tense during a 21-K pursuit. Our desire to witness one should be a social force, just as it is for no-hitters, even more than it is for no-hitters.

So, here are the rules: We have to stay relatively cool. We can't turn on the 21-K siren in the second inning, and we can't expect (nor should we want) managers to let pitchers throw 150 pitches in a pursuit. We just want an extra 10 pitches to chase history, maybe 15. So if a starting pitcher has gone at least four innings, struck out more than two batters per inning, and his pitch count is no more than six times his strikeout total, shout it out. Show that you care. Make it obvious that this is the sort of performance that we'll promise to remember forever, that we'll put on plaques. Sometimes we forget how much baseball players are playing for our admiration, but they don't.

* Unless Mike Trout really does have a 14.8-WAR season.