I’m aware that I run the risk of being called a bad loser with this post but I’m going to do it anyway 🙂
It’s always struck me as a bit odd that in many games, soccer, rugby, tennis and so on, the outcome of a championship is often decided by a single game – albeit usually at the end of one or more rounds. While the ‘single game pressure’ can add to the excitement of the event, playing only one game also runs the risk of overemphasising the role of chance in determining the outcome.
There are a number of codes where the final playoff is a multi-game affair: The America’s Cup challenge (best of 9), the baseball “World Series” (best of 7), Stanley Cup ice hockey (best of 7) and NBA basketball (best of 7). While some may argue that these tournaments are too long the higher number of games does intuitively seem to have the effect of minimising the role of chance and help to ensure that the best team really does win the title. Presumably the same is true of tournament stuctures that include more games rather than fewer.
Of course our intuitions can be misleading. With this in mind, researchers at Los Alamos have recently conducted a study aimed at predicting the number of games needed in a sports *season* to ensure that the best team finishes first. It turns out that a lot of games are required.
The paper doesn’t specifically address playoff structures – so may be of limited comfort to All Black fans – but reading it did distract me enough to temporarily forget about this morning’s loss…