“I’m sorry, Mike. I’m afraid I can’t do that.”

One reason I question how seriously the CFP selection committee takes all the trappings surrounding its mission is because I’m not sure the committee itself will take that stuff all too seriously.

Take, for example, the role statistical analysis will play in seeding the semi-final pool.

“The real difference is we are going to pick four teams and it is going to be done by human beings and not by computers,” said Tranghese.

The hope is to avoid some of the controversies that plagued the BCS during its reign over college football, some of which were caused by those pesky computers. For example, the split national titles for LSU and USC in 2003 and an undefeated Auburn team left out of the title picture a year later. And let’s not forget the 2007 LSU squad that won a championship despite suffering two losses during the season.

Blaming the final selections in those controversial years on computers is so much happy horse shit, of course.  (And it’s not like anyone’s quick to give the computers the credit for the years the BCS nailed the final two teams.) Computers are there to calculate the data that humans feed them.  It’s up to the latter to prevent garbage in, garbage out situations.

But anyway, it’s not like they’re locking the computers out of the selection room now.  Quite the contrary, there’s going to be a dizzying amount of data made available to committee members.  The issue is how much understanding of statistical analysis do they have to make useful decisions based on the information they will access.  Or, from the original Geek,

“The custom platform we built for selection committee was to their specifications and has approximately 100 million pieces of data, including those at the season, game, possession and play-by-play level,” SportSource co-founder Scott Prather said. “We put that into different tools that make it easy for the selection committee to compare and contrast teams.”

“We have statistics at every level of granularity,” Prather continued. “Using custom filters each committee member can determine the information important to them.”

Tranghese’s response manages to be both predictable and hilarious at the same time.

“I don’t think this is rocket science,” said Tranghese.

Yeah, there’s no reason to think it can’t work.

About these ads

18 Comments

Filed under BCS/Playoffs, Stats Geek!

18 responses to ““I’m sorry, Mike. I’m afraid I can’t do that.”

  1. Dog in Fla

    “The real difference is we are going to pick four teams and it is going to be done by human beings and not by computers,” said Tranghese.

    I’d feel much better if Mike would just say it’s going to be done by the Family

    “We have statistics at every level of granularity,” Prather continued. “Using custom filters each committee member can determine the information important to them.”

    I bet these filtered information meetings are going to be sweet

  2. Dog in Fla

    “I don’t think this is rocket science,” said Tranghese.

    It’s going to be more like the six faked Apollo Lunar Missions

  3. Hogbody Spradlin

    Seriously, this crap gives me comfort that they (whoever they are) will never completely eradicate one of the great things about college football: arguments over who shoulda and should’na been higher on the totem pole.

  4. Mayor

    I don’t understand what the linked author is saying about 2007 and LSU winning the BCSNC with 2 losses. Only 1 major college team went undefeated that season (Hawaii) and only 1 major college team (Ohio State) only had 1 loss. Is he saying Hawaii should have played tOSU in the BCSNCG? We now know exactly how good Hawaii was ’cause the Dawgs exposed them. We also know just how good tOSU was ’cause the Tigers exposed them. Who is he saying should have won the BCSNC that season, exactly? If he’s saying Georgia, I agree–but Georgia had 2 losses, too. What is this guy really saying? I don’t think he knows himself.

  5. Dog in Fla

    “going to be done by human beings and not by computers,” said Tranghese.

    Can you read me now?

  6. What bothers me the most is, the best data source may not be used at all by a majority of committee members. There may not be more than 3 or 4 there who know how to watch film. Further, even of there were a few more, it may not carry the weight that it should.

    If it’s utilized properly by people who know what they’re doing, and applied fairly, then we have nothing to worry about. The 4 best teams will be selected. But I have no confidence that will be the case.
    ~~~

  7. Reipar

    Please can we move to right teams now.

  8. DawgPhan

    Unless UGA is playing in the game I will most certainly view any playoff games as I do the Compass Bowl or the BCS Title game.

    Just things that are on that I might be able to watch if I have 3 hours to kill.

    UGA has a clear path to the title game. They just have to win most of their games.

  9. Scorpio Jones, III

    Capitalism…it’s a great thing when you see it at work.

  10. Michael

    This use of data reminds me of how Scalia uses his law clerks to comb through the historical record to support whatever position he wants to reach from the outset. (Richard Posner wrote a good article on this. I’ll try to find it.) The Committee members are going to have a massive amount of data and they’ll cherry-pick to reach their preferred conclusions.