Posted October 08, 2013

On anniversary of Don Larsen’s perfect game, where does it rank all-time?

New York Yankees, No-hitters, World Series
Don Larsen, Yankees

Don Larsen needed just 97 pitches to blank Brooklyn. (National Baseball Hall of Fame/SI)

Tuesday marks the 57th anniversary of Don Larsen’s perfect game for the Yankees against the Dodgers in Game 5 of the 1956 World Series. It was and remains the first such game in World Series history and until 2010, it was the only postseason no-hitter. Building off what I wrote earlier today about the best postseason starts among young pitchers as ranked by Game Score, it’s interesting and somewhat surprising to note that Larsen’s gem does not top the list as the highest-rated postseason start in baseball history.

That may seem to be a flaw in Bill James’ metric, which credits or debits points for every inning, hit, run (earned and unearned), walk and strikeout en route to a number roughly indicative of quality for the purposes of ranking each outing. That’s because the formula rewards strikeouts and innings pitched, so a high-K outing or one that goes beyond nine innings may trump even a hitless one. In fact, when James introduced what he conceded was “a garbage stat” in The Bill James Baseball Abstract 1988, that’s what he had in mind. After all, there have only been four nine-inning, 20-strikeout games — two for Roger Clemens and one each for Kerry Wood and Randy Johnson — (and one start, by the Senators’ Tom Cheney in 1962 in which he went 16 innings and struck out 21; Game Score: 115) but 23 perfect games.

Via the Baseball-Reference.com glossary, here’s how the formula works:

1. Start with 50 points.
2. Add 1 point for each out recorded, so 3 points for every complete inning pitched.
3. Add 2 points for each inning completed after the 4th.
4. Add 1 point for each strikeout.
5. Subtract 2 points for each hit allowed.
6. Subtract 4 points for each earned run allowed.
7. Subtract 2 points for each unearned run allowed.
8. Subtract 1 point for each walk.

According to that methodology, Larsen’s perfect game winds up tied for fifth among alongside Roy Halladay’s no-hitter against the Reds in Game 1 of the 2010 NLDS among all 2,768 postseason starts through Monday:

Player  Series, Gm IP  ER  BB  SO  GSc
Roger Clemens 2000 ALCS, G4 9 1 0 0 2 15 98
Dave McNally 1969 ALCS, G2 11 3 0 0 5 11 97
Babe Ruth 1916 WS, G2 14 6 1 1 3 4 97
Tim Lincecum 2010 NLDS, G1 9 2 0 0 1 14 96
Roy Halladay 2010 NLDS, G1 9 0 0 0 1 8 94
Don Larsen 1956 WS, G5 9 0 0 0 0 7 94
Ed Walsh 1906 WS, G3 9 2 0 0 1 12 94
Josh Beckett 2003 NLCS, G5 9 2 0 0 1 11 93
Ken Holtzman 1973 ALCS, G3 11 3 1 1 1 7 93
Bob Gibson 1968, WS G1 9 5 0 0 1 17 93

Of the four games that outrank Larsen’s perfecto, three had double-digit strikeout totals and two went at least 11 innings. Those who read my previous piece will be familiar with the 21-year-old Ruth’s 14-inning start — still the longest in postseason history — for the Red Sox against the Dodgers as well as Beckett’s NLCS two-hitter for the Marlins against the Cubs, the second-best postseason outing among pitchers age 23-and-under. Gibson’s start for the Cardinals against the Tigers above represents the single-game postseason record for strikeouts.

In any event, none of the starts that outrank Larsen’s by Game Score get to call themselves perfect games. So here’s to Larsen’s remarkable feat, which took place at the original Yankee Stadium in front of 64,519 fans, was won by New York 2-0 and wrapped up in a tidy two hours and six minutes. Check out this pair of videos from MLB.com, the first of which includes the reminiscences of Larsen, catcher Yogi Berra and other participants, and the second of which includes a brief appearance by Dodgers broadcaster Vin Scully, who took the baton from Yankees broadcaster Mel Allen and called the second half of the game:

13 comments
ThomasW.Tilert
ThomasW.Tilert

This is one of the stupidest articles I've ever read.  I can tell you the date of Don Larsen't perfecto, but I don't even know who Clemens was pitching against, never mind the date.  Here's my list of the greatest playoff pitching performances of all time:  1. Larsen.  2.  Halladay.  3.  Jack Morris 10 inning shutout to WIN the World Series.  4. Gibson (only because I'm a Cards fan.)  The rest are, "Hmmm...really, I didn't know that."  And hardly even creep into memory, much less ingrained there.

Tim72
Tim72

I don't see Jack Morris's 10-inning shutout of the Braves for the Twins in the 1991 World Series here.  There's something wrong with this formula.

According to this utterly arbitrary formula, it scores a 91.  But how about, "Add 10 points if it's a complete-game extra-inning shutout in Game Seven of the frickin' World Series" -- especially if it's one of the closest and most evenly matched World Series ever played?

DumbDadNU87
DumbDadNU87

in reading how this "statistic" is calculated it sounds much more like a fantasy sports game than a meaningful statistic.

Kolo
Kolo

Wasn't in a post-season game, but Harvey Haddix's almost perfect game in 1959 was the best game ever pitched.

Michael10
Michael10

Where does it rank all-time? It ranks as perfect, genius.

I'm almost as tired of watching Jaffe play with his toys (Game Score, JAWS, age-based charts generated at Baseball-Reference, etc.) as I am of Cliff Corcoran mancrushing on Mike Trout...almost.

GeraldRosenstingel1
GeraldRosenstingel1

One of the worst articles I've ever seen. A perfect game is a perfect game. Roger Clemens game was better because he had more strikeouts? You gotta be kidding me! He wasn't perfect, so there is no way he did better than Larsen. Using a random formula to calculate how good a pitcher was? Had Clemes allowed an earned run, he would have ended up with the same score as Larsen. That is pathetic!

Vinny Cordoba
Vinny Cordoba

This is a classic example of why the old adage holds true that there are three types of lies: Lies, damn lies, and statistics. It is impossible to do better than 27 batters up, 27 batters down. You can equal it, but you can't top it. You could throw 81 straight strikes and you would have 27 strikeouts and a perfect game, but it's not any more perfect than Larsen's perfect game. You could throw only 27 pitches and throw a perfect game and it's still no better than Larsen's. You could take every batter to a 3-0 count and still throw a perfect game, and it's no worse than Larsen's. That's why it's called "perfect" -- because it can't be improved on.

hjwall
hjwall

Why on earth are strikeouts included? An out is an out. It might be flashy or show dominance, but it is only for style points, not actual quality. A metric that allows for a game with hits or runs be "better" than a no hitter is simply a nonsensical metric.

rpearlston
rpearlston

@ThomasW.Tilert The article used a widely-known and approved of OBJECTIVE metric (game score) to rank post-season pitching performances.  K's add points to the total, and Larsen only struck out 7 batters, so his game score isn't that high, by comparison.  

But game score does not, and never can, take into account the circumstances of any given game.  So, while Roy Halladay's no-no in post-season makes the top-10 post-season list, his perfect game earlier in 2010 resulted in a higher game score.  That said, when Brandon Morrow K'ed 17 in a regular-season almost (clean single with 2 out, top 9, in a home game) , his game score was higher than was that of Halladay's perfect game, and one of the top 3 game scores ever.

It's judged by an objective metric, nothing more and nothing less.

rpearlston
rpearlston

@Tim72 As an objective metric, Game Score cannot take context into account.  But doing that, taking context into account, results in arguments about this one vs this one vs that one, and there is no proof to arguments based on opinions.

misterdeltoid
misterdeltoid

@Vinny Cordoba Wrong Vinnie.  in 1990, Terry Mullholland threw a no-hitter.  The only runner reached base on an error, and was then erased on a double play.  Mulholland got 28 outs in 27 hitters.  Unfortunately, MLB (and, the rest of the world) doesn't see it that way.

rpearlston
rpearlston

@hjwall A strikeout is a pure pitching stat, one that has nothing whatsoever to do with any other member of the team.  The same is true of walks. Strikeouts are also a way out of a tight spot (say, 2nd and 3rd, none out) without having the ball put into play, as the latter always carries with it the risk of errors and of plays not made.  That's why Game Score rewards pitchers strikeouts but punishes their walks, and to the same extent (ie 2 points up for a K and down for a BB).  That makes it an effective way to objectively compare performances.

If you still thing that it's a " nonsensical metric", then you probably feel the same way about ERA, another metric to compare pitchers' performances.

John NoLastName
John NoLastName

@hjwall What you say is true, but I think that's exactly what they're going for ... style points. And why penalize for unearned runs? Isn't it the sabremetrics folks who scoff at pitcher wins because they're influenced by the pitcher's teammates?

For that matter, why penalize for hits or walks? Neither matters ... what matters is runs. But if you don't, then what you're left with simply reduces to ERA, which isn't what they're going for. They're going for style.