by Matt Barr

The most exciting sport

The rules changes in hockey have been fodder for discussion about what makes a sport, or a particular game, "exciting," in terms of what will make fans with lots of other entertainment or leisure choices tune in or go to a game. Hockey had a real problem in recent years (beyond its real real problem last year) in that games were essentially over after two periods. With less obstruction, more power plays, no two-line pass rule, goalies wandering less and wearing smaller equipment and so on, hockey has become a game you need to watch in the third period, because two and three goal leads can be overcome. It had gotten to the point where one goal leads going into the third couldn't.

Call this the more for your money theory. "Money" is shorthand for whatever investment you make in watching the game. I can be more confident now that if I go to or tune into an NHL game the third period will be worth watching. Previously, given a choice between paying to watch 40 minutes of a hockey game or 60 minutes of something else, I'll probably pick the 60 minutes because I'm getting more from my investment.

Scientists Eli Ben-Naim, Sidney Redner and Federico Vazquez at the Los Alamos National Laboratory in New Mexico decided to step back and look at this "unpredictability" over many seasons' worth of games for five different leagues: the NHL, NFL, NBA, Major League Baseball and the English Premier League. Their tack was looking for "upsets," defined, sensibly, as games in which the team with the poorer record going in won. "If there are no upsets, then every game is predictable and hence boring," says Ben-Naim. The men found that soccer was the least predictable and therefore most exciting, followed by baseball. NFL football came up the rear.

Reading their paper (PDF) I think something important is missing. It's not what Tyler Cowen implies, which is that you have to account for injuries and trades in considering whether a game result was really an "upset" or not. (And the paper controls for varying season lengths by "set[ting] up mock sports leagues where teams, paired at random, play a fixed number of games.")

As the paper says, "game location was ignored." But game location is a concrete, knowable factor in advance of a decision whether to view a game. And teams almost always have different winning percentages at home than on the road, where they have to travel to get there, live out of a hotel room and face a generally hostile playing venue.

Scroll down to "League Miscellaneous Stats" on this page to see 2005 AL home and road records. If Minnesota (record: 83-79) played the Yankees (95-67) it would be an upset if the Twins won. But not at the Metrodome, where the Twins' record was 45-36 to the Yankees' road record of 42-39. The converse can be true: The White Sox finished six games better than the Indians, and were ahead of the in the standings all year, but if the teams played in Chicago, their home-road records suggest a White Sox win would have been an upset. But it's far and away more common for a team's home winning percentage to be better than its overall winning percentage, sometimes by a lot.

Cowen notes that a logical conclusion to be drawn from the scientists' data is that "you have to know about many players and teams to figure out what is going on, who is likely to win, and why" in baseball and soccer, as opposed to football and basketball, which will have more casually informed fans. But again, you don't have to be any more than casually informed to know where a game is being played, particularly if you're talking about whether to attend or not.

The American League as a whole in 2005 had a .539 winning percentage at home and .472 on the road. There are many different combinations of teams where accounting for the venue of the game would make a different team out to be the favorite. Those League Miscellaneous Stats linked above order the teams by actual 2005 standings. Count the number of times a home record is better than a road record of a team above it on the list. I count 39, accounting for I don't know how many games (with the unbalanced schedule), where the Los Alamos' scientists would have had the wrong team as the favorite.

Location should be accounted for before you decide which league produces the most upsets.

UPDATE: Spelling of Cowen's name corrected, thanks to commenter Nathan.

Browse books from Amazon.com:
sports