Defending the MRI #1
In what I am sure will become a popular segment on both On the Field and
Sitting Courtside, a friend of mine recommended that I begin to respond to some of the emails that I receive that question the MRI. Typical emails range from "How is [INSERT TEAM HERE] not ranked in the Top 25?" to "You seriously need a life." The life emails, I can't do much about, but the others, I can definitely respond to.
To begin the segment, I have an email from Dave B. According to his son, who is a friend of mine, Dave and I watch more college football than anyone he knows. Dave wrote:
Geez Ben, is there something wrong with the program? VA Tech beats VA by 2 touchdowns and are still ranked lower? Purdue, an average team at best, and Wisky who beat them at Purdue isn't even in the top 25? Utah at #2? Granted they have a nice record......but such a cupcake schedule! Texas Tech at 20 and Fresno at 21???? TTU with their typical no "D" and 60 passes a game offense, and Fresno coming out like gangbusters only to become the mediocre team they really are.
I know the stats dictate how the MRI ranks the teams and this is only my humble opinion. It continues to be a screwy season and may prove once and for all that the BCS system is an absolute joke.
Thanks to Dave for an email which offered so much to write about.
I love emails like Dave's because as a long time reader of the MRI, he knows that the statistics from the games drive the rankings and not the personal feelings I have towards the teams. That said, you would be surprised at the number of emails that I get proclaiming that I certain teams, such as Maryland or Notre Dame.
Let me give a little information on the MRI to start. Behind the scenes, I do run a number of statistics on the formula. First, the actual MRI numbers actually fall in a Normal distribution by some fluke of luck on my part. It is skewed a little toward the forward part of the bell curve, but still mostly normal. Second, the football MRI over the past two years has predicted the winner straight up in 71% of all games when home field advantage is factored in. I have breakdowns by differential using the formula that I use to make the picks. Most of these differential bands are over the 75% mark. Where the MRI does not do as well is in toss-up games, which I will explain. Last, in all these discussions where margin of victory comes up, remember that I use a weighted margin of victory and that you can't win by more than 35 in the MRI.
If you look at the ratings, you will notice tiers of teams. In the rankings
this week, obviously USC is ahead of the pack. When you look at the next group of teams, from #2 to #7, there is a differential in MRI of a little over 9. 9 points is pretty much the break-off point for any tier of teams. In my system, when two teams are 9 points or less apart, unless there is a strong home field advantage for one team, the game is a toss-up. We have already seen this happen this year. Look at the Louisville-Miami game, or the Texas-Oklahoma game. When these were played, the two teams were so close it was hard to choose. The same was true of the Cal-USC game the first time around (though I still believe this would be a close one). What this means is that any team from 2-7 playing each other on a neutral field could go either way. When you think about the quality of the teams involved, it is not hard to imagine.
Now, Boise State is just outside this group and for good reason, with their recent close wins against inferior opponents. It will be interesting to see what happens when they play Louisville in the Liberty Bowl (less than 9 point difference, and my pick for the best bowl game of the year).
Because of their loss, Virginia fell from being an outsider from the top tier and being back in one of these groups below them, a group that now includes Virginia Tech. Virginia's loss dropped them almost 4 MRI points, which is a little less than half of the biggest drop this week. At the same time, Virginia Tech's win was good for 7.5 MRI points, or almost 2/3's of the largest gain for the week. Both teams were rewarded/punished for their
performance. If we move this game to a neutral field, we would again have a very close game (I contend the game was closer than the final score).
Now, Utah with a cupcake schedule? I would tend to disagree. While the combined record of their opponents was under .500, they did not play a Division 1-AA opponent which put their schedule strength in the MRI at 58, about right in the middle. Compare this to Auburn at 82. Utah performed better on a comparative basis against the lesser opponents. I do factor in opposing team's strength into the formula, and they are being hurt for the lesser schedule, but they have overcome it by a far superior performance. Meanwhile, Oklahoma in no way separated itself from Oklahoma State, Texas, or A&M when they played this year. They struggled in each game to win and gave up a lot in return to the opponent, though not in points to Texas. The key here was have a dominating performance which Utah had and Oklahoma did not (unlike last year where OK had many, against almost every team they played, and moved so far ahead that two losses at the end of the season still couldn't bring them back to the pack).
Wisconsin suffers from a lackluster offense which came to light in their two losses to end the season. Their defense finally cracked and the offense wasn't there to make up for it. I had pointed out Wisconsin's deficiencies earlier, just before it went on its two game losing streak. At the time there were 6 teams with no losses and I was certain that Wisconsin would lose before the end of the season. I didn't count on Michigan State being the team to do it, but between them and Iowa, it was definitely likely. Meanwhile, Purdue played the 11th toughest schedule and lost its four games by a total of 10 points! Two of those games were decided in the final few moments when a late turnover cost them. What I find hard to believe is that a team like Ohio State could be in the Top 25 rankings with the same record in conference and overall, and a worse performance over the course of the season in total.
I had thought about only ranking 20 teams in football this year. 25 teams is 21% of all the teams in the league, so being in the list between 21 and 25 doesn't really say much about your team. I couldn't just rank the same percentage as in basketball because that would just be 9 teams. Ranking 15 would probably be a better number but who am I to buck tradition. That whole thing leads me to Fresno State and Texas Tech.
Texas Tech has been around all year in the rankings and was finally recognized this weekend by the writers and coaches. Their defense is actually better than average in the league and while their rushing game is terrible as Dave correctly points out (They average less than 100 yards a game on the ground), they generate an amazing amount of offense in the air. I am guessing they end up in a bowl game against a better opponent, someone who will be able to exploit this weakness and stop them.
Fresno State caught some flack from me earlier in the season, but have rebounded from the stretch that saw them lose 3 in a row and won their last 5 games. In those games, they scored 42 or more in each game, and won by at least 30 each time. This was enough to bounce them back up into the top 25. Again, think dominating performances, both on the field, and on the scoreboard. Notice that this is the Top 25, not the top 15. If that is the discussion, we don't even hear about Fresno all year long.
Hope that explains a little bit plus gives you a little more knowledge into the MRI in general. Dave is correct that this season has been screwy and actually hurt the MRI percentage over the past two years. However, I expect to bounce back soon. If anyone else has an email question, please feel free to pass it on and let me attempt to respond. In the meantime, enjoy and keep on reading.
Just a note that tonight is the MAC championship game. The winner doesn't get a BCS bid, or anything particularly special. The MAC championship over the past few years is normally a good game (Discount last year when Miami rolled Bowling Green by 22). If you want to see some future NFL players, watch tonight. You may be surprised to hear their names called on Sunday in the near future.
