(From BCS Guru)
The Guru takes his work very seriously. And unlike our politicians, I'm all about accountability. So lately, I've had readers asking, just how accurate are you?
Well, good question. And you know what? I'd like to know that myself. So the Guru dug up the archives and put together a little chart (yes, spreadsheets are addictive, so don't try this at home), and voila! I've got myself a scorecard.
There's no need to get too much into the methodology here. We all know what a report card looks like. It seems that the Guru is an A-student, but when he's not, he got a royal spanking - from the readers - just as he did from his teachers when he was in school. So here it is:
|Accuracy||Perfect Top 5||Note|
|2006|| || || |
|Final||84%|| ||Only Top 10 Projected|
| || || || |
|2007|| || || |
|Week 1||93%||YES|| |
|Week 2||83%|| ||Worst Performance Ever; Missed #2|
|Week 3||90%||YES|| |
|Week 4||93%||YES|| |
|Week 5||94%||YES|| |
|Week 6||99%||YES||Best Performance Ever; Perfect Top 10|
|Week 7||90%|| || |
|Final||95%|| ||Perfect (5-for-5) on BCS Bowl Matchups|
| || || || |
|2008|| || || |
|Week 1||90%|| || |
|Week 2||97%||YES|| |
|Week 3||90%|| ||Missed #2 - Second Time Ever|
Now, you see that in the Guru's entire tenure, I've had one disastrous week - last year's Week 2: When I was out of town attending a wedding. I had a terrible cold and my laptop didn't have all the data I needed, yadi yada ... anyway, I ended up projecting LSU as No. 2 and it ended up to be Boston College.
History repeated itself last week - for the second time since I started the projections, I failed to nail the top 2. Texas Tech got enough of a boost to sneak past Penn State. While I had no doubt that Tech would've eventually ascended to No. 2 - could've been as soon as this week - I didn't think the Red Raiders would get by Penn State immediately after their first meaningful victory of the year.
I bring this up because this is where the BCS will have huge problems. One of the reasons that Tech got past Penn State is that it went from being 13th in the Billingsley Report to No. 1. Imagine that! A 12-spot jump over one game! The first week of the BCS standings, Billingsley had Tech 19th. That sort of quantum leap defies logic.
Now, Billingsley, like all but one of the six computers (Colley Matrix is the exception) does not make available the exact methodology of his system. You can read it here what he wants you to know. But his "computer rating" is without a doubt the most biased, with many of his personal preferences deliberately installed in the "formula." Even he will tell you that his "rating" is more like a one-man poll (the last one on the page, but you should read the whole thing, it's quite entertaining, actually).
Each week, out of every team's six computer scores, the highest and the lowest are discarded. Billingsley consistently has about 40-50% of his numbers thrown out. Why? Because his system, simply put, is not a pure computer formula. He relies heavily on a preseason ranking that is admittedly biased and traditional football powers always get preferential treatment. Check out where he has USC ranked throughout this season, in relation to where the Trojans were on the other computers.
So what's the purpose of this Billingsley rant? What does it have to do with my own scorecard?
The point is this: accountability. It's high time for the BCS to require transparency from the computers - as well as the polls. If the computer ratings are proprietary, then at least they should be audited. And the poll results should be available to the public, every week, just like the AP. It's a joke that the Harris voters and coaches don't need to be accountable until the end of the season.
Since this is the system we have to live with and there are lots at stake, it's beyond comprehension that the BCS is run like a mom-and-pop shop ... wait, that's an insult to mom-and-pop shops that are responsive to their customers' needs and reactions. It cannot - and should not - continue to tolerate and allow numbers that seemingly come off a dart board to decide who gets to play for the national championship.