Consider this: According to Elias, out of the 52 best-of-seven series in baseball history to be tied 2-2 after four games, as this one was, the winner of Game 5 went on to win the series 36 times, or 69 percent of the time…
The author of this insight got paid for making it. A while back if you had someone on a newspaper’s staff and he was unqualified for just about anything you made him the music critic. Today he goes to the sports desk. I’ll come back to the above tautology in a moment.
Baseball is an unusual game in that the best team loses about 40% of its games. Thus a seven game series can easily be won by a much weaker team. Look what happened to the NY Yankees in a five game series. They lost three games to two against the Detroit Tigers. The Tigers had a similar record. If the American League had operated as it did before the playoff scheme was hatched the Tigers would have finished 2 games behind the New Yorkers. It took 162 games to distinguish the almost imperceptible difference between the two teams. A five game series was no better than a coin toss which would have averaged a 3:2 ratio if performed often enough.
Back to the winner of the fifth game of a seven game series. Three things can happen to the winner of the fifth game. They win the sixth and thus the series. They lose the sixth but win the seventh and win the series. The last possibility is that they lose both the sixth and seventh games and thus lose the series. Notice that two of the possibilities have the winner of the fifth game winning the series, while only one has them losing. Thus the likelihood that the winner of the fifth game winning the series is 67% – just about what Elias showed. You don’t need a sports bureau or a degree in probability analysis to know that the winner of the fifth game of a seven game series will win that series two thirds of the time.
Since only the good teams get into the playoffs the differences in ability among them are very small. Thus the winner of each game and each series follows the rules of tossing a coin. No matter how good the players, how skillful the coaching, and no matter all the planning that goes into each game and each series the outcome is virtually pure luck or rather pure probability – that of a coin toss. Toss a coin seven times and any possibility might come up. To get the expected 4:3 ratio with 95% certainty you’d have to toss the coin many multiples of seven.
I understand that the playoff system gives the fans something to root for and keeps them interested in a weak team for far longer than would the old system where there were no playoffs, just a World Series. But the owners and general managers should be expected to realize that this set up is an exercise in randomness and not make important personnel decisions on the basis of what happen in these chance encounters. There’s nothing they can do that will meaningfully affect the outcomes of these brief liaisons. But then it’s just a game.
As for the sports writers, they have to produce something even if they’re clueless about what determines the outcome of a short playoff series. In other words, they don’t realize that the outcome is pure chance. Sending them back to the music desk is not an option.
Once it wasn’t so, when just 4 teams made the playoffs, or even pre-1969 when it was a straight 2 team showdown. I don’t buy into all the “luck” put forward by today’s analysts. In the 1970’s the best team probably won every World Series, in the 80’s a Kansas City would sneak in, or the Dodgers in 1988. Now with eight teams luck starts to factor in. But the whole thing is not random, just as Rod Carew’s hits didn’t just drop in all the time by accident.
In a game where no team wins 60% of its games the outcome of a short series is random. If there’s a 10% difference between two teams which almost never is the case in the playoffs you would have to play 216 games to be 95% sure that the better team would win – 95% is where statistical significance is usually set. Thus even a whole season often doesn’t reveal the best team. I’m not saying that there aren’t differences among teams or players – Babe Ruth was obviously a much better hitter than Roger Maris – I’m just saying that a short series can’t, because of the nature of baseball, reliably go to the better team and that its result is no better than a coin toss. Like it or no, it’s a statistical reality.
[…] annual encounter with randomness is well underway. I’ve already written about this – here and here. But the baseball playoffs allow revisiting the topic. Baseball is a sport where the best […]