Dan Olds interviewed me for NEMO5’s use in the Student Cluster Competition at Supercomputing 13
More info about the competition is here:
I was interviewed about our group’s work for the fall issue of the National Center for Supercomputing Applications’ Access Magazine
Story featuring our group’s work from a couple months ago about some cool new storage at Purdue
The group at Advanced NFL Stats have created a lot of metrics derived from NFL statistics. One of these is the measures is the Excitement Index (EI), that is supposed to indicate whether there were only 1,000 fans in the stands at the end of the game, or whether you were asking people the next day if they had seen gridiron battle the previous night. You can read about the EI here, but basically the higher score, the more exciting the game.
After analyzing games from 2010, we found that the average EI was 4.03, but for Monday Night games, it was only 3.51.
While this is not a large difference, it does support the authors’ feelings that Monday Night Football are too often a letdown. It will be interesting to see if this trend holds for other years.
One argument for less exciting Monday Night Games is due to parity in the NFL. Monday Night schedules are made months in advance and it is difficult to ensure that there will be good matchups. In fact, if the game choice is overanalyzed, a boring game may be the result.
For instance the Detroit Lions had the highest average EI (4.9) in 2010 but no MNF games. This may have been due to their winless ‘08 season and 2-14 record in 2009. They finished 6-10 in 2010.
Another point to note is that two teams (Jets and Vikings) had three appearance, but several teams had none. It may make sense to give the Jets three games because of their market, but the Vikings three appearances are surprising. The Bills, Panthers, Browns, Lions, Raiders, Seahawks, Rams, and Bucs had zero appearances.
While it’s difficult to whittle down excitement to a single number, it is surprising that the games chosen for a national audience are less exciting than the average matchup. Maybe there is some way to predict, months in advance, the most exciting games of the week. Or, maybe the league can implement a system that would allow them to choose a time much closer to the matchup. For instance, the league could say: matchups A and B are the possible MFN matchups for week N. Then the league could choose the MNF matchup a week or two in advance—the other game would be played on Sunday. That would give the the league leeway to choose the more exciting matchup without causing undue scheduling issues. Certainly when you take in close to $10 bn a year, you can have some leeway in things like that.
So is everyone excited to see the NFC West powerhouse matchup when the abysmal Rams (2-10) take on the “last year we won our division at 7-9″ Seahawks (5-7) on Dec 12th? If the game keeps me interested through the 3rd quarter, I truthfully am.