Wednesday, February 03, 2010

Stunner: Champ Isn't Always 'Best Team'

As you may have heard, Sunday is the Super Bowl.

(You'll notice that I have avoided saying much since the Packers were eliminated. I also didn't say a word about the Vikings being eliminated. It's partially by design. I figure that I'm not in Miami, and you're getting plenty of Super Bowl coverage elsewhere. I'll break down the matchup and throw out a prediction while I'm in Houghton this weekend.)

As the Colts and Saints prepare to lock horns, fans are left to marvel at the sight of No. 1 seeds advancing out of the respective conferences. It doesn't happen often.

Parity in sports is a hot-button subject. In the NFL, the objective these days is more to just get in the playoffs than it is to have a 14-2 season. Other sports, while still leaning toward the favorites, have had their moments.

Of course, it helps that other pro sports feature best-of-seven series in the playoffs. The format tends to tilt in favor of the superior team, since it's pretty hard to be an inferior team and win four of seven games.

If you do, it's probably because you deserve it, no matter your status in the regular season. After all, you just won four out of seven (or fewer) games against a good team. It's not a one-and-done fluke. It's more of the real deal.

For the NFL, a study confirms what most of us have known for many, many years.

The best team doesn't always win.

To quantify this, consider an alternate universe in which the NBA adopted the NFL’s format for the regular season and playoffs. Neil Paine of Basketball Reference constructed an approximation of such an NBA season, aligning teams to match as closely as possible the NBA’s 30-team league with the NFL’s 32-team league. With 16 games to go around, there aren’t enough games for every team to play every other one, like in the NFL — so, for instance, the Lakers don’t play the Cleveland Cavaliers in this alternate reality. Then Paine simulated 10,000 such NBA seasons, using teams’ stats through last Tuesday’s games to estimate their strength and therefore their likelihood of winning each game.

In these seasons, the usual suspects tend to finish on top. The Cavs win the “Super Bowl” 19.6% of the time, followed by the Lakers at 19.4% and the Celtics at 12.1%. The Hawks, Nuggets, Spurs, Jazz and Magic all also have at least a 6% chance of winning it all. However, some losing teams win in more than one of these simulated seasons, including the 76ers, the Clippers and the Kings.

Before this NBA season, Paine ran similar simulations, using the actual NBA season structure, and found that the best team has about a 48% chance of winning the title.

I've said this before, and I'll say it again.

Playoffs aren't designed to figure out which team is the best. They're designed for drama, television ratings, and revenue.

The only way for the best team to win the championship every year is to adopt the "everyone plays each other twice" system used by -- among others -- the English Premier League.

It will never happen. Instead, fans in the States have always enjoyed playoff competition, and they're not at all concerned about proving who the absolute best team is. That's why college football gets so much crap from American sports fans, after all.

No comments: