Many people have expressed surprise at the Week One grades calculated by the RPG system. Here are a couple of comparisons to make the grades more clear and logical.
Experts and playoff committee members are fond of saying they consider strength of schedule/opponent when ranking teams but they don’t actually include that consideration in their weekly rankings. RPG does. Very precisely. An opponent value of 100 points is equivalent to a home game against a team ranked at the midpoint of the FBS division, or around No. 65. Games against higher ranked teams have more value; games against lower ranked teams have less value.
Alabama played a top three ranked Florida State team on a neutral site (away game for both teams). The potential value of that victory for Alabama was 130.8 points (if they played perfectly). Alabama did not play perfectly. Alabama’s performance received a grade of 76.25 based upon the eleven factors that determine the outcome of every game. However, they earned that grade against the best possible opposition on a neutral site so their final grade was 99.73.
Compare that to the performance of the Washington Huskies. Washington played the weakest possible opponent–Rutgers–albeit on Rutgers’ home turf. A perfectly played victory would have been worth 80.12 points for Washington. Washington did not play perfectly. Their performance received a grade of 83, better actually than Alabama’s grade, but when adjusted for competition (value of victory), that grade became a 66.50. In other words, Washington should have played far better against a weak opponent.
This adjustment for strength of opponent is virtually impossible for a human being. We watch the game and see Washington play decently and think they should get a decent grade. We cannot judge the difference in opponent strength between Florida State and Rutgers. So, a computer algorithm has to do it for us.
Another useful comparison is between Penn State and Clemson. Based upon the eyeball test, Clemson was elevated to the No. 3 position in the Week Two polls. However, the numbers tell us something different. Penn State and Clemson played opponents equivalent to that played by Washington–the weakest possible members of the FBS division. Moreover, both played at home which is less difficult than playing on the road. The potential value of victories by Penn State and Clemson was just 73.5 points, less than the potential value of Washington’s road win.
To their credit, Clemson and Penn State played the way elite teams should play against weak opposition. Their play pleased the eyeballs of poll voters. Clemson graded out at 104.8, better than perfect, for a competition-adjusted grade of 77.05, far less than the final grade for Alabama. Penn State played even better. The Nittany Lions received a grade of 114.5 for a competition adjusted grade of 84.14.
Penn State and Clemson played well but against weak competition at home. They were expected to play this well. They did nothing exceptional to merit a promotion in the rankings. They exceeded the play of Washington, which underachieved against equivalent competition, so they are ranked ahead of Washington.
So why did Penn State receive a better grade than Clemson? Clemson played slightly better defense than did Penn State, generating more drive stoppers (sacks, major penalties and turnovers) and limiting first downs and third down conversions. However, Penn State played better on offense and special teams, generating more big plays and short field opportunities. So, Penn State’s performance was slightly better than Clemson’s although both looked good to the naked eye.
The RPG system dispassionately grades performance consistently and meticulously. Humans are incapable of the same judgments.