2005/10/06 | Major Indoor Soccer League | My Favorite Sport That You've Never Heard Of

Soccer is often characterized as "boring", which is a bit unfair (especially if the criticism is coming from a baseball fan), but not entirely unfair (unless the criticism is coming from a baseball fan). There are beautiful moves and combinations, but there's a lot of time spent in between the highlights just moving the ball down the field. If you aren't interested in seeing how the combinations are set up, most of the game won't interest you.

What if you shrunk the field down? That would increase the action, but the ball would go out of bounds a lot. So try enclosing the field in a rink, like hockey. If you add a few more wrinkles, you get Indoor Soccer, as played in the MISL starting about a month from now. It's a lot like hockey, except that I enjoy watching it, perhaps because the shot-on-goal per possession ratio is more to my taste and because there's no icing.

My data for the MISL is from my favorite website for American sports leagues you've never heard of, OurSportsCentral.com. They don't cover Major League Baseball, but they cover the minor leagues and several independent leagues. They have at least four American soccer leagues (I'm composing this off-line, so I can't check the precise number). They cover lacrosse. Even I don't pay attention to lacrosse.

Anyway, if you have a taste for the obscure, or you want to cheer on athletes who have to keep real jobs, even while they're active, or you care about lacrosse, check them out.

2005/09/26 | general | Point Spreads

Due to popular demand for point spreads, I ran some regressions of the probabilities generated by my program against the difference in score in games actually played. You can estimate a point-spread from a probability by selecting the appropriate sport, subtracting .5 from the probability, and multiplying by the multiplier. If you don't know what R2 is, you can think of it as a number from 0 to 1 indicating how good the prediction has been (higher is better).

NFL 200446.29
most of CFL 200569.37
most of MLB 200512.8.04
most of MLS 20053.3.28
WNBA 2007 (added 2008/05/27)45.11

2005/09/11 | College Football |

Iowa State 23, Iowa 3 (I'm an Iowa State fan.)

2005/09/08 | Soccer World Cup Qualifying | Man vs. Machine

Late in the US v. Guatemala match last night, the commentators noted that the US was ranked #6 in the FIFA rankings, one spot below Mexico. They speculated that America's 2-0 victory over Mexico last Saturday could push the US past Mexico. I don't know how FIFA's standings work, but this would seem reasonable to me. In general, I think the evidence, such as our recent record against Mexico, suggests that we are at least their equal, though my heart won't really believe that until we can get at least a draw out of Mexico City.

Unfortunately, Mexico beat Panama 5-0 last night, while the US got a scoreless draw hosting Guatemala. Guatemala is a better team than Panama, but not five goals better. This analysis misses a couple of important points, though, which I'm sure FIFA's rankings (like mine) will miss.

For one thing, the United States clinched a spot in the world cup last weekend, while Guatemala is on the bubble. Not only did this lessen game's importance to the American players and increase it to the Guatemalans, it lessened it to Bruce Arena, who started 11 players who hadn't started any of the previous games in the series. Since Arena wants to fill out his World Cup roster more than he wants to win games, this was primarily the American second string that got the draw last night (though Eddie Johnson and Landon Donovan did come in during the second half).

It can be argued that the American players last night were quite motivated to play well, since they were playing for a spot on the roster. But the fact that they were second-string players won't be reflected in any of the automated rankings. The weakness of automated rankings is that they don't account for imprecise, subjective factors like how important the game is to the coach and the players. The weakness of human-generated rankings is that they include a lot of imprecise, subjective factors and unacknowledged biases.

UPDATE: My rankings, such as they are, still put the US above Mexico.