Sports are full of conspiracy theories, some of them very strange. The most popular fringe theory these days is probably the (absurd) notion that the Super Bowl has been fixed for at least one of the last four seasons. But I have a sports conspiracy theory so full of holes it makes that one look positively sane.
This whole project developed because of my birthday. You start thinking about getting older and what that means, maybe not a full-blown "crisis," as we call them, but at least a somber contemplation of the state of things. One of the things I did around my birthday was go over my mental checklist of things I want to get around to one day, hopefully while I'm still somewhat young. One of the things on my list is to attend a World Series Game 7.
But here's the thing: not every Series goes seven games.
If two teams are evenly matched — if each has a 50% chance of winning any given game — there is only a 31.25% chance that the Series will go the full seven. Flip a coin until you get four heads or four tails, and it's the same thing. But shift the balance even a little — say, 55-45 — and that percentage drops to 30.3. Given some of the blowouts in the last few years, it seems reasonable to guess that a Series could be as unbalanced as 60-40. In that scenario, the chances of seeing Game Seven drop to 27.65%, or about one in four.
Those aren't bad odds. Assuming that most World Series opponents are within that 50-50 to 60-40 range, with the majority around the slight advantage of 55-45, any given Series would have about a 30% chance of lasting seven games. Wait a decade, and you'll see three Game Sevens.
After I figured out all these odds, I decided to check the actual record. The math is absolute, so it should be more accurate, but it's worth looking at the box scores. You could flip heads 10 times in a row (it happens about one in a thousand times), but if you do, something is probably up. When you look at the statistics above, and then at World Series results, it looks like something is probably up.
I'm really interested only in MLB's modern era. Historians may have a different cutoff date, but I used the end of the color barrier. Since Jackie Robinson's debut with the Dodgers in 1947, there have been 57 World Series. Twenty-five of them — 43.9% — have gone the distance. The margin of error here is only +/- 13.2%, which would mean it is almost impossible that the Series is being played straight.
The World Series is fixed! Shout it from the rooftops!
Some fans may take Boston's unlikely sweep of the Cardinals last year as sufficient evidence that the World Series is crooked, but why would games be fixed to extend the Series? Revenue and fan interest. A terrific Series could be especially desirable when the league could use a little help.
Immediately following the second world war (1945-47), three consecutive series went the distance. Coincidence? Could be. But it would also be a great time for a legendary Series or two to get the country thinking about baseball instead of war crimes and Communism and Nagasaki.
In the '50s, fully half the series required seventh games, including four in a row, all involving the Yankees, from 1955-58. In the '60s, six World Series Game 7s were played. In the '70s, five. Things calmed down a little in the '80s, but there were still four Game 7s (and the average should be three), including three in a row from '85-'87. In all, from 1945-87, there were 43 World Series, and 23 of them (53.5%) went to a seventh game.
The juiced Oakland teams of the late-'80s finally overcame this apparent corruption by playing in three straight blowout series (1-4, 4-0, 0-4). In the '90s, two of the nine WS went seven games. Two of the five this century have required a seventh game. The '90s were the only post-color barrier decade in which 30% or fewer of the October Classics lasted seven games.
Why hasn't anyone realized and publicized this before? Because my methodology is badly flawed.
Some of you have probably noticed by now that I haven't mentioned home field advantage. In all but the most unbalanced matchups, the home team wins more often than it loses. That means when the favorite (which might have a 55% chance to win any given game on a neutral field) plays at home, it's almost a lock. But it also means that the underdog gets a few games in which it becomes the favorite. Alternating home fields produces more upsets and longer series. It explains the unusual number of seven-game World Series in the last 60 years.
But riddle me this: without adjusting for home field, an evenly-matched series (with neither team having an advantage larger than 55% versus 45%) should end in a sweep about 13% of the time. Since 1947, there have been 10 4-0 World Series sweeps — 17.5%. That could be a statistical aberration, but it's unlikely. Either the Series is fixed to produce sweeps (shout it from the rooftops!), or the advantage of the stronger team frequently runs higher than 55-45 or even 60-40. If the latter is true, then our seven-game numbers are skewed again, even accounting for home field advantage.
All in all, 35 of the 57 series since Jackie Robinson's first game in the majors (61.4%) have produced either sweeps or seven-game series. Statistically, the majority of World Series should last five or six games. Sweeps have their own magic, of course, so it's conceivable that arranging sweeps could benefit the league. There have been three in the last seven years: two establishing the contemporary Yankee dynasty in the late-'90s, and last year's Red Sox victory that kept the focus on Boston's amazing comeback in the ALCS and one of the most overblown feel-good stories of the last half-century.
I don't really believe the World Series is fixed. I don't see how it could be done, and the risks outweigh the benefits. It's such an unlikely theory that it can be reasonably dismissed out of hand. That said, the numbers do not add up. I'm not a statistician, but I do understand basic statistical concepts, and unless I've made a major error in my calculations, it is statistically impossible or close to it for the World Series to produce so many four- or seven-game series.
Shout it from the rooftops.
June 2, 2005
eric mcclung:
i think a lot of a sweep or a 7 game series is logical- either the two teams are match each other well [most likely causing 7 games] or one team is better [most likely causing a sweep]
also, the world series was fixed in 1919 and only a few of the players were in on the fix- so it is possiable to fix a series even w/ just some of the roster knowing about it.