WPS and the postseason, part three

This is the concluding installment of a three-part series on measuring the most exciting postseason series, using the Win Percentage Sum (WPS) Index I devised a few months ago in these articles. Part one covered the LDS (and tiebreakers) here, and part two handled the LCS here.

Now, just as in real life, it’s time for the World Series. I have plenty of ground to cover, so I will dispense with the preliminaries. No full-team introductions, no fly-overs, not even the National Anthem. (You may stand at home and hum if you like.) I will remind readers of my original articles that I’m using a slightly different method for counting up WPA events, so games I reviewed then may have slightly different numbers when I look back at them now.

Finding a happy median

In previous installments, along with the WPS Recap I’ve been doing for the 2012 playoffs, I estimated a median WPS Index for baseball games at between 305 and 310 points. (Median means half have more, half have less.) This was based on the first two rounds of playoffs, and was subject to further data. The World Series provided that data, and it has bent the curve.

Taking all 1,348 postseason and tiebreaker games for which I can derive WPS values (not counting this year’s contests), I now have the historical median for the WPS Index at 297.45 points. In practical terms, this means 300 works fine as a round estimate. This drop in the median implies that World Series games, strictly isolated as games rather than in a larger context of being the World Series, have historically been less exciting than games in earlier rounds. We can blame the deadball eras for that.

This isn’t really surprising. I’ve noted before that the WPS system leans in favor of high scores, which is fair enough since the fans I’m trying to emulate tend to lean the same way. The other playoff rounds postdate 1968, the end of baseball’s second deadball era. The World Series, though, carries the weight of those low-offense periods with it.

During the first deadball era, which I’ll date from 1901-1919, there were 16 World Series. Three of them had more games above a 300 WPS score than under; 11 had more games below 300 than above. Two of the three high-excitement series came during the 1911-12 “oasis” in that era, when the original “rabbit ball” boosted scoring until the powers that be got together and returned things close to normal. Low offense led to low WPS scores.

A similar thing happened in the deadball era of the 1960s. From 1963 to 1968, no World Series featured a positive ratio of above-median games to below-median games. To anticipate an upcoming table, two of the three least-exciting World Series ever came from that six-year stretch.

I am not going to attempt a corrective for these differing numbers. If WPS says that deadball-era games are duller, that’s what a lot of fans believed too. It’s not just chicks who dig the longball. Mirroring this subjective reaction is arguably a strength of WPS rather than a flaw. In any case, it stays, freely admitted to.

The most exciting games

I will fulfill that subhead’s promise, but first we need to acknowledge the most boredom-inducing World Series games. Fret not: the gratification won’t be deferred for long.

    Year/Game     Teams     WPS Index
5.  1985 Gm.7  KC  over StL   124.3 [Clincher]
4.  1923 Gm.5  NYY over NYG   124.1
3.  1959 Gm.1  ChA over LAD   101.8
2.  1967 Gm.4  StL over Bos    98.9
1.  1960 Gm.3  NYY over Pit    98.8

1985 was the day after the Don Denkinger game, when the Cardinals just fell apart and Whitey Herzog and Joaquin Andujar got full value (emotional for them; entertainment for us) out of their ejections. The rest follow the same scoring pattern: big leads early that were never challenged. The Yankees-Pirates game was one of the three blowouts New York famously engineered in the 1960 Series, one of which took until the fifth inning to get out of hand and wasn’t actually that bad a game. One of Pittsburgh’s wins just barely beat it out.

My most-exciting game list was going to be 10 items long, but there are two games among them that arguably merit asterisks. I’ll go to 12, and explain the problems with the two peculiar games when they come up. It is notable that all top 12 games went to extra innings. The highest WPS Index for a nine-inning World Series contest belongs—no big surprise— to Kirk Gibson and the 1988 Dodgers-A’s, Game One.

No. 12: 10/24/1992, Game 6, Blue Jays over Braves—666.4 points

Game       1   2   3   4   5   6   7   8   9  10  11    F
Blue Jays  1   0   0   1   0   0   0   0   0   0   2    4
Braves     0   0   1   0   0   0   0   0   1   0   1    3

WPS        1   2   3   4   5   6   7   8   9  10  11
Blue Jays 24  11  13  27  14   7  10  13   7  25  76
Braves     7  26  16   6  30  17  16  21  99  13  61
WPS Base: 541.4  Best Plays: 110  Last Play: 15  Grand Total: 666.4

It’s the second title-clincher for Toronto, with Joe Carter‘s walk-off homer, that has stuck in fans’ memories, but it was 1992 that was better, for Series and final game alike. The game was close, low-scoring but fairly active for most of the regulation nine. Down one, Atlanta mounted a rally in the bottom of the ninth, getting the tying and winning runs on base but finding themselves one out, then one strike, from defeat. Otis Nixon came through, his single scoring Jeff Blauser and forcing a 10th inning.

The 11th proved decisive. Toronto got two aboard against Charlie Leibrandt, and after Joe Carter flied out (he’d do better in his next big Game Six situation), it was Dave Winfield‘s double that put the Jays up 4-2. The Braves rebounded, getting the first two batters onto the corners against Jimmy Key. A dubious non-squeeze sacrifice by Rafael Belliard moved the tying run (a pinch-running John Smoltz!) to second, and he reached third on Brian Hunter‘s RBI groundout.

Otis Nixon, Atlanta’s hero of the hour, came to the plate against new reliever Mike Timlin, and tried a surprise bunt to bring Smoltz home. It’s brilliant if it works—but it didn’t. Timlin got to the ball in time, and the championship of America’s Pastime went to Canada.

No. 11: 10/4/1924, Game 1, Giants over Nationals/Senators—670.9 points

Game       1   2   3   4   5   6   7   8   9  10  11  12    F
Giants     0   1   0   1   0   0   0   0   0   0   0   2    4
Nationals  0   0   0   0   0   1   0   0   1   0   0   1    3

WPS        1   2   3   4   5   6   7   8   9  10  11  12
Giants     5  27   9  16   8  10  25  18  10  35  15  69
Nationals  5  12  18  10   6  19  25  29  87  41  15  52
WPS Base: 564.9  Best Plays: 89  Last Play: 17  Grand Total: 670.9

N.B. The official nickname of the Washington franchise at this time was “Nationals,” but their old name of “Senators” stuck around so tenaciously that it was restored in the 1950s. Being a tiresome showoff, I use “Nationals” here. (Ask me about the Philadelphia Blue Jays sometime.)

A remarkable parallel to the No. 12 game: the home team gets a run in the last of the ninth to tie it 2-2, then plates a comeback run in extras only to fall short 4-3. This time the game was saved two outs from the end, as Washington’s Roger Peckinpaugh doubled home Ossie Bluege in the ninth, but perished on third as the possible winning run.

Walter Johnson and Art Nehf dueled on into the 12th. Nehf was even part of the Giants’ 12th-inning charge, his single plus error putting men on second and third with no outs. (Nehf might not have been pinch-hit for even today: he slugged over .500 in 1924.) Both runners came home on Ross Youngs‘ single and George Kelly‘s flyout.

The Nationals did not quit. Bucky Harris‘ one-out single halved the gap, followed by a Sam Rice single. Rice stretched for two, so one more hit could win the game, but future managing great Billy Southworth cut him down from center, and Goose Goslin‘s groundout ended the affair. Walter Johnson suffered the loss, but his role in this series was not nearly over.

No. 10: 10/22/1991, Game 3, Braves over Twins—673.4points

Game       1   2   3   4   5   6   7   8   9  10  11  12    F
Twins      1   0   0   0   0   0   1   2   0   0   0   0    4
Braves     0   1   0   1   2   0   0   0   0   0   0   1    5

WPS        1   2   3   4   5   6   7   8   9  10  11  12
Twins     12   4   6   6   6  27  14  86  27  31  13  69
Braves     5  27   5  30  30   2   3  11  28  28  13  61
WPS Base: 546.4  Best Plays: 88  Last Play: 39  Grand Total: 673.4

This is not the Twins-Braves game you were expecting. Game Seven was great, Game Six was better, but this was the best of that memorable series. It nearly forfeited that primacy when Atlanta pulled ahead 4-1 in the fifth, but Minnesota threatened in the sixth, got one back in the seventh, then tied it the next inning when Chili Davis scorched one over the left-field wall.

Both squads had chances in the ninth and 10th, but couldn’t convert. In the Twins’ 12th, a Mark Lemke error helped get the go-ahead run to third with one out, but Braves relievers Kent Mercker and Jim Clancy extinguished the blaze. Twins closer Rick Aguilera almost got out of the home 12th, before Dave Justice stole second with two outs and two strikes. Aguilera then lost Greg Olson, and Mark Lemke lined a single through into left that brought Justice home with the game-winner.

No. 9: 10/25/1986, Game 6, Mets over Red Sox—675.6 points

Game       1   2   3   4   5   6   7   8   9  10    F
Red Sox    1   1   0   0   0   0   1   0   0   2    5
Mets       0   0   0   0   2   0   0   1   0   3    6

WPS        1   2   3   4   5   6   7   8   9  10
Red Sox   29  21   4   8  11  14  28  16  25  49
Mets       5  10   6   6  41  34  10  42  49 106
WPS Base: 519.6  Best Plays: 116  Last Play: 40  Grand Total: 675.6

How much more is left to say about this one? It was a good game through nine, with some early action and no dull innings after the fourth. Then the pendulum swung about as far toward the Red Sox as it could, before scything back like something out of Edgar Allan Poe. I’ll only add in conclusion that there has never been a greater sporting event that was interrupted by a rogue parachutist, and that list is longer than you might think.

No. 8*: 10/8/1945, Game 6, Cubs over Tigers—727.4 points

Game       1   2   3   4   5   6   7   8   9  10  11  12    F
Tigers     0   1   0   0   0   0   2   4   0   0   0   0    7
Cubs       0   0   0   0   4   1   2   0   0   0   0   1    8

WPS        1   2   3   4   5   6   7   8   9  10  11  12
Tigers     5  38   5  20   5  17  26  46  63  26  13  19
Cubs      12  14  10  13  60   7  15  23  48  28  13  62

Baseball had the famed “green light” to play through the Second World War, but that doesn’t mean it was the same game. Even a month after the final surrender, rosters were a hodgepodge of draft rejectees and deferees, veterans back from active duty, and whoever else could be found who could play. If you don’t care to consider this game fully major-league, you have some justification.

Viewed solely for itself, however, this game is a barn-burner. The numbers piled up nicely early, even as Chicago ran away to a 5-1 lead. Detroit recouped half that margin, lost it again, then wiped away the whole lead in an eighth inning capped by Hank Greenberg‘s homer. Both sides traded threats through the 10th inning, but relievers Dizzy Trout for Detroit and Hank Borowy for Chicago held firm, and the game went into the 12th.

It ended up Trout who got landed, without getting hooked. (Sorry.) (No, I’m not.) Dizzy had almost quelled a small Cubs uprising in the home 12th when leadoff man Stan Hack doubled to left. Pinch-runner Bill Schuster came around from first with the winning run—in the last major league game he would ever play. He was, after all, not nearly the only ballplayer coming home.

No. 7: 10/10/1924, Game 7, Nationals/Senators over Giants—734.5 points

Game       1   2   3   4   5   6   7   8   9  10  11  12    F
Giants     0   0   0   0   0   3   0   0   0   0   0   0    3
Nationals  0   0   0   1   0   0   0   2   0   0   0   1    4

WPS        1   2   3   4   5   6   7   8   9  10  11  12
Giants     9  11  12   6  13  65   8   6  64  31  34  31
Nationals  5   5   6  19   5   9  27  90  64  15  30  47
WPS Base: 610.5  Best Plays: 96  Last Play: 28  Grand Total: 734.5

Walter Johnson had already taken two losses in this, the first World Series he and his Nationals had ever played. When the Giants erupted for three runs in the sixth to take a 3-1 lead against George Mogridge, it appeared the greatest pitcher in the game was going to be the October goat. But fate was about to take some funny hops.

Washington got the bases jammed in the eighth, but Virgil Barnes coaxed a short fly and was about to escape untouched when Bucky Harris grounded one to third. The ball hit a pebble in the infield dirt and bounced over Freddie Lindstrom‘s head, bringing the tying runs around. The Nats had pulled their third pitcher for a pinch bat to sustain that rally, and in came Walter Johnson to hold the line.

He almost gave it right away on Frankie Frisch‘s one-out triple in the top of the ninth, but he bore down for a strikeout and groundout. Washington got its own runner to third with one gone in the ninth, but a double play snuffed that hope. Johnson let the leadoff man get on in the next three innings, but fought back to put up three goose eggs. In the home 12th, Johnson, apparently pitching for the duration, reached on an error to make it first and second, one gone.

Up came Earl McNeely, who grounded one toward third. The ball hit a pebble—we will never know, but it may have been the exact same pebble—caromed over Fred Lindstrom’s head, and Muddy Ruel raced home with the run that made Washington first in war, first in peace, and finally first in baseball.

And made Walter Johnson the hero of the 1924 World Series.

No. 6*: 10/9/1912, Game 2, Giants at Red Sox—743.7 points

Game       1   2   3   4   5   6   7   8   9  10  11    F
Giants     0   1   0   1   0   0   0   3   0   1   0    6
Red Sox    3   0   0   0   1   0   0   1   0   1   0    6

WPS        1   2   3   4   5   6   7   8   9  10  11
Giants    17  17   6  21   6  14  22 114  42  55  37
Red Sox   37   7   4   5  25   7   3  66  15  88  15
WPS Base: 619.7  Best Plays: 120  Last Play: 4  Grand Total: 743.7

WPS measures, in a sense, the length of the road a game takes to its resolution. But what if there is no resolution? Three World Series games have ended in ties; this was the best of them. But how exciting is a game that’s supposed to produce a winner, but doesn’t? How you answer that question decides whether you think this game belongs on the list.

Boston broke ahead early, and though New York inched back in, the Red Sox had it 4-2 after seven. In a see-saw eighth, New York struck for three, taking advantage of two ground-rule doubles into the crowd standing in left field. (It was a different era.) Boston got their own ground-rule double, then plated the tying run on shortstop Art Fletcher‘s third error of the day.

Fred Merkle opened the 10th, his long fly eluding the standees and ending up a triple: he came home on Moose McCormick‘s sacrifice fly. Tris Speaker would respond with his own big hit in the home half, a ball boxed around in center field, on the relay, and at the plate. Speaker scored on what was ruled a triple and an E-2. Two Giants reached first in the 11th, and both were cut down stealing second. Christy Mathewson set down Boston in order on three grounders to close the 11th.

And that was it. Game called on account of darkness.

(Tangential interjection: a Twitter message circulated from ESPN on Monday said that the Giants were 0-5 all-time in Game Sevens. Not so. The Giants won Game Seven of the 1912 World Series … which tied it at three. They also took Game Seven in 1921 against the Yankees, in a scheduled five-of-nine World Series. ESPN should have said “deciding Game Sevens,” in which case the Giants were 0-4, or “winner-take-all games ending seven-game series” or such, where the Giants were 0-5.

Of course, the Giants have since won a deciding Game Seven, hours after the ill-conceived Tweet. So that bungled factoid need never bother us again.)

No. 5: 10/21/1975, Game 6, Red Sox over Reds—755.2 points

Game       1   2   3   4   5   6   7   8   9  10  11  12    F
Reds       0   0   0   0   3   0   2   1   0   0   0   0    6
Red Sox    3   0   0   0   0   0   0   3   0   0   0   1    7

WPS        1   2   3   4   5   6   7   8   9  10  11  12
Reds       9   4   6  13  53  17  62   9  15  38  30  43
Red Sox    32  3  10  11  16  15   9  75  72  15  15  36
WPS Base: 607.2  Best Plays: 112  Last Play: 36  Grand Total: 755.2

And if there was nothing new to say about Mookie/Buckner, what can I possibly add about this one? I’d spend most of my time justifying why it isn’t numero uno, which I can do better in the following comments. Just remember: being the fifth most-exciting World Series game ever is not an insult.

No. 4: 10/21/2000, Game 1, Yankees over Mets—766 points

Game       1   2   3   4   5   6   7   8   9  10  11  12    F
Mets       0   0   0   0   0   0   3   0   0   0   0   0    3
Yankees    0   0   0   0   0   2   0   0   1   0   0   1    4

WPS        1   2   3   4   5   6   7   8   9  10  11  12
Mets       5  13  14  14  23  18  70   7  23  16  16  16
Yankees    5  16  12   6   7  43  22  33  72  70  41  84
WPS Base: 645.0  Best Plays: 88  Last Play: 33  Grand Total: 766

This first stanza of the most recent Subway Series took its time working up to tempo, but it ended fortissimo. The Yankees broke the drought in the sixth on Dave Justice’s two-run double, but the Mets rebounded with three in the seventh, driven in by Bubba Trammell and an Edgardo Alfonzo infield single. Al Leiter and John Franco held firm in the seventh and eighth, but the Yanks got to closer Armando Benitez, Chuck Knoblauch‘s sacrifice fly knotting the score at three.

From then on, Yankees pitchers stymied Mets bats, while their hitters battered the Mets bullpen like a rising tide. The 10th saw them on second and third with no outs, but Glendon Rusch survived with a popout and double play. The 11th saw second and third again, now with two gone, but Turk Wendell coaxed the final fly-out. The story repeated itself in the 12th: second and third, one down. Wendell got the foul popout, but Jose Vizcaino‘s single breached the seawall, and put the Bombers up 1-0 in the Series.

No. 3: 10/14/1973, Game 2, Mets over A’s—784.3 points

Game       1   2   3   4   5   6   7   8   9  10  11  12    F
Mets       0   1   1   0   0   4   0   0   0   0   0   4   10
A's        2   1   0   0   0   0   1   0   2   0   0   1    7

WPS        1   2   3   4   5   6   7   8   9  10  11  12
Mets       4  15  15  18  10  72   8   3   5  55  35 112
A's       41  17  25   4   9   5  36   8  92  12  12  50
WPS Base: 675.3  Best Plays: 108  Last Play: 9  Grand Total: 784.3

A super game, with a painful ending and a worse aftermath. There was good see-saw action early, then a big rally in the sixth to put New York ahead. Oakland chipped away in the seventh, then last-chance hits by Reggie Jackson and Gene Tenace got them tied in the ninth. After that, it became a mirror-image of the Subway Series above, this time the Mets mounting rallies that Rollie Fingers quelled while the A’s went down quietly, until the 12th.

New York had already scored one in that frame when Mike Andrews, one of the last men off Oakland’s bench, made a pair of two-out errors at second base that let three more Mets cross the plate. The A’s seemed cooked, but they got the earned run back and had the tying run on base with one out, only for their comeback to sputter out. A’s owner Charlie Finley infamously coerced Andrews to sign a statement saying he was injured so Oakland could replace him on the World Series roster, but Commissioner Bowie Kuhn snapped Finley back. Andrews would bat once more in the Series, then play out his career in Japan.

No. 2: 10/25/2005. Game 3, White Sox over Astros—890.9 points

Game       1   2   3   4   5   6   7   8   9  10  11  12  13  14    F
White Sox  0   0   0   0   5   0   0   0   0   0   0   0   0   2    7
Astros     1   0   2   1   0   0   0   1   0   0   0   0   0   0    5

WPS        1   2   3   4   5   6   7   8   9  10  11  12  13  14
White Sox  9  30   6   6  84   9  14  14  26  14  50  14  30  82
Astros    22   9  25   7   7   9  26  64  68  38  40  15  28  32
WPS Base: 779.9  Best Plays: 101  Last Play: 10  Grand Total: 890.9

This game seemed a likely yawner when Jason Lane‘s fourth-inning homer stretched Houston’s lead to 4-0. Chicago didn’t let that illusion linger: they surged for five in the fifth, capped by A.J. Pierzynski‘s two-run double, and had the bases loaded when Roy Oswalt finally stopped the landslide. The score stayed there until the home eighth, when Jason Lane struck again with a two-out double. They got Chris Burke to third with one down in the ninth, 90 feet from victory, but Orlando Hernandez barely held the line.

Into extras they went, and deep into them. Threats arose, and receded. It seemed to be happening again when Paul Konerko hit into a 5-4-3 double play in the 14th for two away. Backup second baseman Geoff Blum might have seemed an easy third out, but appearances again were deceiving. His homer put the White Sox ahead, and they tacked on another via a bases-loaded walk. The Astros did not quit, and got the winning run to the plate in their half. Mark Buehrle, coming on in emergency relief, produced the pop-up from Adam Everett, and the dominating 3-0 Series lead was Chicago’s.

No. 1: 10/27/2011, Game 6, Cardinals over Rangers—898.8 points

Game       1   2   3   4   5   6   7   8   9  10  11    F
Rangers    1   1   0   1   1   0   3   0   0   2   0    9
Cardinals  2   0   0   1   0   1   0   1   2   2   1   10

WPS        1   2   3   4   5   6   7   8   9  10  11
Rangers   38  37  11  31  39   7  47   1   4  54  24
Cardinals 32   5   5  23   7  70   4  38  95 103  38
WPS Base: 716.8  Best Plays: 144  Last Play: 38  Grand Total: 898.8

There is no Platonic ideal of an exciting baseball game, but this one comes close. The lead swung back and forth early, always staying within one run. This allows maximum range for exciting innings, and the teams took advantage. Then Texas had its big inning, launched by back-to-back longballs, and went into the stretch up three. St. Louis pecked away in the eighth on Allen Craig‘s solo shot, but needed two more runs in the final inning just to stay alive. They got those runs on base, but Craig struck out looking and David Freese went down to his final strike.

Hollywood would have had Freese hit the game-winning home run here. Real life did better.

Freese’s long fly just got over the head of Nelson Cruz, good for a triple and the tie. In the visitors 10th, Josh Hamilton gave his audition for the role of Game Six hero, slamming a two-run homer. However, as Hamilton would say later in a much-mocked declaration, God told him he’d hit the home run, but didn’t tell him he’d win. St. Louis’s first two got on, but its next two got out, and Lance Berkman went down to his final strike.

Again, real lie spun out the tale, Berkman singling to tie but the rally ending there. One inning later, David Freese got his second chance at immortality, and hit it out to dead center, winning what the WPS Index (barely) calls the most exciting World Series game ever.

The most exciting series

Onward we go, now looking at the World Series with the highest cumulative WPS numbers. And, briefly, the lowest.

   Year     Teams (Games) WPS Index
5. 1928  NYY over SLN (4)   1123.8
4. 2007  Bos over Col (4)   1034.9
3. 1966  Bal over LAD (4)    968.4
2. 1963  LAD over NYY (4)    910.7
1. 1989  Oak over SF  (4)    805.1

Not too much to say about dull series. 1989, of course, was the year the Loma Prieta earthquake killed scores in the Bay Area and postponed Game Three for over a week. Given the combination of WPS disinterest and real-life misery, I have no hesitation calling this the worst World Series of all time. Yes, even ahead of 1919.

All of the laggard-board Series were sweeps. The worst five-game World Series was the 1937 Yankees beating the Giants; the worst to go six was in 1930 with the A’s and Cardinals. The worst Series to go all the way was the 1965 Dodgers-Twins matchup, with just one game barely beating the median in that dreary deadball era. Four World Series have gone to an eighth game, and the worst by WPS was the first of them all, 1903 between Pittsburgh and the Boston Americans. (I will say here that by any non-statistical measure, 1919 should get the booby prize among the eights.)

As for the best Series, only No. 10 on the list went fewer than seven games. The best five-game set was the 2000 Subway Series, the first game of which hit No. 4 on the single-game list. The best sweep, by far, was the White Sox beating the Astros in 2005. So consistently excellent were all its games that it finished 18th out of all 107 World Series, easily beating the best five-game Series. Had Houston extended it just one more day, an average Game Five would have put them into the top ten.

I’ll give capsules for the lower five, and something more for the top five. Again, there is an asterisk case—and it’s an awkwardly placed one, as you’ll see.

    Year     Teams (Games) WPS Index
10. 1992  Tor over Atl (6)   2527
9.  1952  NYY over Bkn (7)   2528
8.  1925  Pit over Was (7)   2636.1
7.  1947  NYY over Bkn (7)   2655
6.  1972  Oak over Cin (7)   2723.5

Again, Toronto’s 1992 victory had more excitement than the better-remembered 1993 repeat. The 1952 Yankees-Dodgers set had six tightly-fought games decided by one or two runs, and deserves Bill James’ accolade as the Series of the 1950s. 1925′s Series had only one poor game, and ended with Walter Johnson failing under the twin burdens of injury and Roger Peckinpaugh’s errors, giving the Pirates the crown. The peak of the 1947 Subway Series was Game Four, where Yankees hurler Bill Bevens went from a no-hitter to defeat on the final pitch of the game. Six of the Oakland-Cincinnati games in 1972 ended with one-run margins, a World Series record.

No. 5: St. Louis over Texas, 2011—2840.1 points

The first two games were one-run affairs, including the Rangers taking the second with a ninth-inning comeback. They snagged Game Five with two in the bottom of the eighth. And then there was the Game Six of the millennium: if we see a better one in our lifetimes, we will be very fortunate. A pedestrian Game Seven barely damped the final result.

No. 4: Minnesota over Atlanta, 1991—2946.2 points

In one sense, a series where the home team wins every game bears a structural lack of suspense: the “favorite” wins every time. In another sense, it’s good for game excitement in giving maximum range for walk-off victories. The ’91 World Series had four of those, three in extra innings. Atlanta’s came first in Games Three and Four, Mark Lemke involved first as the RBI and second as the run scored. Kirby Puckett‘s game-winning homer in Game Six gave us an announcer’s call that will be with us as long as there is a Buck in the booth, and the 10th-inning win in Game Seven sealed what some consider the greatest World Series pitching performance ever, by Jack Morris.

No. 3: Washington over New York(N), 1924—3018.6 points

An old showbiz saw holds that the key to success is a great opening and a great finish. This series followed that advice, bookended with top-12 games (top-10 if you’re ruthless with the asterisks). Game Two ended with a Washington walk-off, and Game Six was a one-run squeaker also. The external narrative of Walter Johnson’s quest for a title adds even more spice to this Series, but that’s something the numbers just cannot measure.

No. 2: Cincinnati over Boston, 1975—3069.3 points

Five of the games were one-run affairs, and all five finished with WPS scores above 400. (Plus, one of the blowouts was a scoreless tie through the seventh-inning stretch, and beats the WPS average of 300.) Such sustained excellence would have made this World Series memorable, and Game Six made it historic. The one thing keeping it off the top of this list is a fluke: a Series that went one game longer.

No. 1*: Boston(A) over New York(N), 1912—3097.8 points

The tied Game Two I covered above made possible the incongruous result of a best-of-seven series ending in Game Eight. It had a compelling finish: a decision deferred to the 10th inning, and New York pulling ahead, only for Fred Snodgrass‘ muff of a routine fly to center to trigger the Boston comeback. His spectacular catch of Harry Hooper‘s subsequent blast to center could have redeemed his failure, but soon Tris Speaker’s foul pop would fall between three Giants fielders—including Fred Merkle, whose RBI single in the top of the 10th could have been the game-winner and erased the “Bonehead” tag that had plagued his life since 1908.

Speaker exploited his second chance. His single to right drove home Clyde Engle from third, and Josh Devore‘s unwise throw home let the winning run go first to third, Speaker close behind at second. An intentional walk for the force at every base ended up meaningless, as Larry Gardner‘s deep fly brought home the championship run, ending an undeniably exciting Series.

Series Percentage Sum, mark two

But it’s also undeniable that the 1912 Series finishes first here because of the Game Two tie: remove it, and the series falls well out of the top 10. If you don’t believe a tie game should count as much as any other, you’re looking at a different most exciting Series.

There are also other factors that go into an exciting series aside from counting up scores for each game. The flow of the series itself matters: whether and how long it stays closely fought, and whether the most exciting games are the ones that matter most in determining a winner. A great Game Six or Seven should count more than a great Game One, or Game Four in a sweep.

I’ve argued this before, and tried to solve the problem before. The LDS installment of this trilogy offered a multiplier to produce something called the Series Percentage Sum (SPS), based on the change in probability of teams winning the series caused by resolution of one specific game. (E.g., in a five-game series, Game One would have a multiplier of 0.1875, Game Four would be 0.25, and Game Five 0.5.)

This proved unsatisfactory, as winner-take-all games overwhelmed all the others, which does not happen with our perception of the games. We remember World Series Game Sixes as all-time greats as often as Game Sevens, or more so. A good SPS system needs a less severe curve, giving some value to games as games and not just stepping-stones to a series win.

My latest formula does that much better. Take the probability of each team winning the series should it win the game in question, and add them together: that is the SPS multiplier for that game. For example, in Game One of a seven-game series, the probability of series victory for each team will be .65625 should it win, giving a multiplier of .65625 + .65625 = 1.3125. (An alternative way of calculating this is to add the series Win Probability Added numbers for each team, and add 1 to that.)

For the first and second games of a seven-game World Series, the SPS multiplier is 1.3125. Game Three of an even series will be at 1.375, while a 2-0 series advantage will make it 1.25. Game Four of a potential sweep will be 1.125, while a 2-1 series will make it 1.375. A 3-1 series will produce a 1.25 multiplier for Game Five, while a 2-2 series will come in at 1.5, not only for Game Five but Game Six as well. Game Seven, for all the marbles, has an SPS multiplier of 2. I made similar calculations for the few best-of-nine World Series.

As for the tie games, I treat them as though no change in series winning probabilities was going to happen (as ended up the case), and give them multipliers of 1. This seems a fair penalty for the lack of resolution, making it less meaningful than any game with a winner but not punishing it too severely.

This method gives an advantage to series that go the distance or close, and which have their best games late. That’s how it should be, as that’s how fans regard the games, but hopefully it doesn’t statistically bury a series that doesn’t measure up exactly.

A second look at the best

Re-evaluating the best World Series by this new SPS method gives us moderately different results. All of the top 10 series now went the distance, or more, while the 2005 super-sweep gets dropped from 18th to 32nd out of 107. The SPS top 10, with their ranks by the WPS method, are as follows.

    Year     Teams     SPS Index  Prev. Rank
10. 1952  NYY over Bkn   3803.9       9th
9.  1997  Flo over Cle   3855.7      12th
8.  1972  Oak over Cin   3863.7       6th
7.  1947  NYY over Bkn   3867.9       7th
6.  1925  Pit over Was   3877.9       8th
5.  2011  StL over Tex   4191.1       5th
4.  1912* BoA over NYG   4282.9       1st
3.  1991  Min over Atl   4339.2       4th
2.  1975  Cin over Bos   4580.5       2nd
1.  1924  Was over NYG   4619.6       3rd

Most of the differences are series getting shuffled but staying in the top ten. The 1992 Series does fall out (to 14th), replaced by the 1997 version. The Marlins and Indians had some snoozers in their seven-game set, but are redeemed by a Game Seven that crashes the best-game party (see below). The 1912 Series slips to fourth after its tie gets re-evaluated, which feels less problematic while still giving credit for great games.

Note the separation of the top five, and the top two, from the rest of the pack. Thirty-nine points is too small a margin to call decisive between first and second, though 240 points is probably enough to make them clear winners over the number three series.

The results for the most exciting single game are pretty satisfying. (Due to a lingering asterisk game, I’ll stretch this list to 11 games.) There is a sprinkling of early-series contests, but the great majority of occupants come from Game Six or later, and Game Sevens don’t totally dominate, which feels right. We also have more new entries, including two whose previous absence likely irked some readers.

    Year/Game     Teams (Innings) SPS Index  Prev. Rank
11. 1991 Gm.7  Min over Atl (10)   1009.4       63rd
10. 1986 Gm.6  NYM over Bos (10)   1013.4        9th
9.  1973 Gm.2  NYM over Oak (12)   1029.4        3rd
8.  1945 Gm.6* ChN over Det (12)   1091.1        8th
7.  2005 Gm.3  ChA over Hou (14)   1113.6        2nd
6.  1975 Gm.6  Bos over Cin (12)   1132.8        5th
5.  1912 Gm.8  BoA over NYG (10)   1162.2       33rd
4.  1997 Gm.7  Flo over Cle (11)   1174.8       28th
3.  1960 Gm.7  Pit over NYY (9)    1250.6       19th
2.  2011 Gm.6  StL over Tex (11)   1348.2        1st
1.  1924 Gm.7  Was over NYG (12)   1469          7th

Jack Morris can breathe a little easier: his most famous game makes the slightly extended list. The 1912 clincher I already described. Game Seven in 1997 had the Marlins save their season with a two-out ninth-inning run to tie Cleveland. They loaded the bases with one out in the home 11th, lost the lead runner to a force at home, then pulled it out with Edgar Renteria‘s grounder threading through the middle.

The reason 1960′s Game Seven wasn’t on the original list was a few early frames with a four-run Pittsburgh lead making it feel like the game could already be over. It wasn’t. The Yankees clawed to a 5-4 lead in the sixth, and extended it in the eighth. Then Pittsburgh made its comeback, capped by Hal Smith donning the hero’s mantle with a three-run bomb to make it 9-7. New York promptly ripped away that mantle (with some help from Mantle) with two runs to tie in the top of the ninth. Then Bill Mazeroski hit one the Yankees couldn’t come back from.

The most noteworthy change on the list is at the very top, as David Freese gives way to Walter Johnson and the Pebble of Doom. Much as I love and cherish last year’s Game Six, there is some justice to this ranking. A winner-take-all game going 12 innings, and ending with a split-second turn of fortune that must have sent spectators’ hearts leaping into their throats, makes a compelling case. Even more so when the main beneficiary is perhaps the game’s greatest pitcher, getting the glory after a long career toiling on lesser teams. Imagine if Ernie Banks had finally gotten to the World Series, and been pivotal in the game that won it all for the Cubs. That’s as close as I can come to equating it for a modern audience.

Yet there’s also a disappointment in this game taking the top spot. With many games on this list, we have the opportunity to relive the excitement directly, as recordings of the TV broadcasts can show us Freese and Mookie and Fisk and Maz in their immortal moments. But we cannot watch Walter Johnson straining on the mound to redeem himself, or Freddie Lindstrom’s head snapping upward as a baseball ricochets over his head and into history. If television has made it easier to recapture some games, it has made others even more distant from us, drawing a dividing line between eras that we can be reluctant to cross.

If these rankings and the other ones I’ve provided over the course of this month have a point, other than my eccentric desire to quantify so subjective a concept as game excitement, it is to guide people toward games that deserve to be remembered, that fulfill a fan’s desire for exciting contests. A lot of them are easy to summon up, and some are not. If in the final analysis the greatest World Series, and its greatest game, come from the obscure depths of 1924, maybe that can sow a little better appreciation for those long-gone days, and show the rewards of digging a little deeper into baseball’s history to find its gems.

References & Resources
Baseball-Reference provided the WPA data without which this enterprise would have been impossible, along with some historical narratives for past games that put some meat on the bones.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Should we ever use a complex ERA estimator?
Next: Getting out of the injury zone, part two »

Comments

  1. Greg Simons said...

    This was a terrific article, Shane.  You mentioned what are probably my favorite and least favorite games.  The worst is 1985 World Series Game Seven when the Cardinals imploded.  The best?  Last year’s Game Six, of course.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

Current day month ye@r *