Fairness is such a complicated and compelling topic that it seems to have a way of taking over tourneygeek. But let’s leave it to one side, at least for a post or two. Recall that fairness is only one element in the FEPS framework for the goals of tourney design: Fairness, Efficiency, Participation, and Spectacle. Let us, in honor of March Madness – that great annual spectacle of a tournament – shift attention to the fourth element, spectator appeal, and see whether we can reach into tourneygeek’s bag of tricks and find something that will help us design tournaments that are compelling to watch.
What can we do to make our tourneys produce close games? In particular, can anything be done about the NCAA basketball tournaments ridiculous tendency to produce blowouts in its early rounds?
There are any number of aspects of watchability, and those who make money when other people watch games, and especially sports, study the matter deeply. Understanding why the Federer/Nadal final at this year’s Australian Open was so much more compelling than something more expected, like Djokovic/Murray, is a topic that could be (and has) been discussed for hours.
For our purposes, let’s focus on one plausible conjecture about spectator appeal: Close contests are more interesting to watch than ones that are uncompetitive. Presumably, tournament formats that are good at producing close contests will be more watchable than others.
For my money, the champion format for producing close matches is the Swiss system. I’ll discuss the format in more detail some time in future, but for now, it’s enough to think of a Swiss system format as working like an elimination tournament, but one in which no one ever gets eliminated. Winners play other winners, losers play losers, and players in between play others with in-between records. At the beginning, there mat be lopsided pairings – indeed, the initial rounds might be seeded in such a way as to make that more likely. But as the tournament progresses, especially in events with a large element of skill, the tournament will tend to find opponents who are well matched, and likely to produce close games.
Having said this much, I have to admit that this is, at best, plausible, based on what I understand about how the Swiss system operates, and my experience playing in a number of Swiss tourneys in chess, bridge, and backgammon. I don’t really know this to be true, and come to think of it, there’s good reason to believe that it’s really not particularly true of backgammon, with its high degree of luck and low level of skill progression.
There are lots of plausible ways to reform March Madness so that we won’t have to endure so many virtual byes like Kansas v. Austin Peay. But first we need a way of measuring the extent to which a tourney tends to produce close games. In the next post, I’ll outline some of the possibilities.