Who am I? My name is Eugene Potemkin. I have not been an ardent college football fan for many years. I have graduated from the Moscow Physical Engineering Institute in 1970 yr. So I've seen American football one time only in Moscow in 1993 yr. More information about me You can read in the Wall Street Journal.

Just now I've read very interesting article by Mr. Hartman. All next text is his article is his with my remarks. In this way I try to solve language problem. :)

Now hold on! I can almost hear you: "Another one of those math geeks! There's no way I'm going to understand this guy!" Well, here's a promise I'll make to you now: You'll be amazed at how incredibly easy, and how black and white, nearly all of this is! So please, hang in there with me. If you enjoy college football and pride yourself on being able to do a little critical thinking, then you'll relish the ride!

What are my goals? Well, they do not include trying to make a case for one specific team or another. Rather, they are simply to educate and to reform. The more people who know of the problems with the BCS, the more likely we are to see this process improved (or removed). If you find yourself flabbergasted as you become aware of the problems in the system, please vent! Share this website with any friend who would have an interest. I'd encourage you in particular to call, write, or e-mail any member of the local or national media (of any form) that you can. And in particular, if you can get in touch with any of the members of the BCS committee, namely the commissioners of the six "power conferences", go for it!!

Keep this in mind throughout: There doesn't exist "the correct method" for determining who should play in the championship game. We'll never achieve total agreement on just which components should make up the process, nor how they should best be combined. These are qualitative, subjective concerns. So, although I'm going to shortly pick apart various procedures that the BCS committee has used, I do want to at least applaud their initial effort. Theirs is a difficult, thankless task to be sure.

Having said all that, however, I must tell you that there are any number of critically important procedures within the BCS components which are indefensibly inappropriate! And in almost all instances, detecting these errors is a very simple matter! It takes only a willingness to do some basic "critical thinking". Recall the promise I made to you at the outset: understanding this presentation will be a "piece of cake" for you, and when you have finished reading and digesting it, you will likely be thinking, "This guy's really onto something!" (as opposed to "This guy's really on something!")

OK, let's get cracking. Here's a quick outline of what I'm going to present: first, I'll briefly summarize the components of the BCS rating formula. If you already are perfectly conversant in this area, skip directly to the next phase, which is the presentation of the specific procedures which I earlier labeled as indefensibly inappropriate. Nearly every component of the process has about it an aspect that will leave you scratching your head, wondering just how in the world it could have been put in place. At each step I'll point out what the committee should have done. And lastly, I would be remiss if I didn't make a final recommendation or two. So, without further adieu, let's look at ...

The BCS system started in 1998. For the first 3 years, 4 components comprised the system. For each component a numerical value, or rating, was obtained for each of the Division 1A college football teams. The lower the number, the better. For each team, these four numbers were added together to yield a final number, which I'll refer to as a team's BCS rating. Four sites (the Rose, Orange, Sugar, and Fiesta) hosted the four "BCS bowl games" with the championship game rotated among them. The two teams with the lowest (best) BCS ratings earned a berth in that championship game. The champion of each of the "power" conferences --- SEC, ACC, PAC-10, Big East, Big 12, Big 10 --- is guaranteed a berth in a BCS bowl, with teams within the top 12 of the BCS standings being eligible for the remaining 2 berths. There is another detail or two, but they are not an issue to me here.

Back to the components. The first involves determining a "human poll average". This involves utilizing the ballots cast by the 72 sportswriters who make up the AP poll and the 60 coaches who comprise the USA Today poll. Again, the goal is to arrive at a numerical rating for each top team by using this data. I will discuss this component in more detail shortly. [Editor's note: In 2003 only 65 sportswriters were used.]

The second component involves determining a "computer poll average" The first year 3 computers were employed. Since then 8 have been used. As each human pollster is to do, so does each computer system turn in rankings (1,2,3,...) of the division 1A teams. Currently, for any given team, its lowest (best) and highest (worst) computer ratings are dropped, and the remaining 6 are then averaged. [Editor's note: In 2003 only 7 computers were used, with each team's worst mark being dropped.]

The third component is called the "strength of schedule", or SOS, component. It involves trying to gauge the difficulty of each team's schedule and to assign a numerical rating accordingly. The more difficult a team's schedule is, the lower (better) its SOS rating. Again, more on this later.

The fourth component is quite simple. I'll refer to it as the "loss" component. The numerical rating assigned to this component for each team is simply the number of losses that the team has incurred.

So, each component generated a numerical value, and these were added together to produce the "BCS rating". Generally these computations were made weekly from about the 7th week until the end of the season, with the only one that truly mattered being the final one.

Now in 2001, an additional feature was added to the mix. It is applied after the previous 4 components have been generated and added together. I will refer to it as the "quality win" (QW) component, because it tries to reward teams for having quality wins. It works as follows: after the BCS ratings are initially determined, a team is then allowed to deduct 1.5 points from its total for a win over the BCS #1 team, 1.4 points for a win over #2, and so on down to .1 point for a win over #15. (Example: if a team has wins over the # 7 and #11 teams, it would receive a 1.4 total point reduction --- .9 points for beating #7 and .5 points for beating #11). After the quality win component has been calculated and applied, we arrive at the final BCS ratings. [Editor's note: As of 2003, only wins over teams in the Top 10 are considered "quality wins". Now, 1.0 points are deducted for a win over #1, 0.9 points for a win over #2, down to 0.1 points for a win over #10.]

Click here to read Hartman's questions and my commentaries