The Alma Mater Bowl
From the Print Edition:
Don Johnson, Mar/Apr 02
There isn't enough Mylanta in the world to cure me of college bowl sickness. This malady creeps up in my stomach once a year when I rediscover that there are more bowls than cereals to put them in. The past bowl season offered up no fewer than 25 bowl games in 17 days. 'Twas the season to be inundated. Worse yet, the Bowl Championship Series, while ultimately deciding a national champion, could have left the whole college scene in a mess. Fortunately, the one game that really counted among those 25 -- the Rose Bowl contest between Miami and Nebraska on January 3 -- ended not only with a clear winner, but a clear-cut national champion. We were lucky this time.
The solution to this madness? We need a playoff. A genuine playoff between the top teams would decide the winners. Only, in my series I want to change things. Since I regard performance in the pros as more significant than college performance, I want to know which colleges would win an All-Star Game using only those players they've put in the pros.
To that end, we set up an eight-team championship series in which each of the seven games counts. Notice that in the NCAA Bowl Championship Series, the top eight teams don't actually play one another. That is, after Miami won the Rose Bowl, it was declared No.1 without having had to play the other six contestants. Imagine if this happened in golf or basketball seedings! Should Tiger Woods be declared No.1 in the Masters for just beating the second-ranked player in the tournament? Then why play the rest of the Bowl Championship games? Just to sell tickets -- and products -- and make some money for television?
Here we will play the games using a computer simulation to run an eight-team, single-elimination tournament. The teams will comprise current National Football League players, and the simulation will be based on the players' performance in the pros during the 2000 season, because we had full-season stats to rely on. In cases of some outstanding rookies for the 2001 season, such as Santana Moss of the New York Jets, we fudged a little bit and used 2000 college stats. The other benefit of a computer simulation? You don't have to sit through names like The Crucial.com Humanitarian Bowl, the Tostitos Fiesta Bowl or the Fed Ex Orange Bowl. Who invented these titles? An ad writer on a binge?
We had wanted to kick this off with a field comprised of the top-12-ranked colleges by the number of players they put in the NFL. But four of these 12 teams were eliminated before the first coin flip because of roster holes. Those eliminated are Nebraska, Ohio State, Florida and Georgia. I'm not doing this to be mean-spirited or because I want to aggravate the fans of those teams. Nebraska has to go because it doesn't have a single quarterback or wide receiver on any of the 31 NFL rosters. Teams have won games, even Super Bowls, with poor quarterbacks. But you'd be hard pressed to find a team that has won with no quarterback. Ohio State? It's out for less obvious reasons: it lacks a center, a guard and a offensive tackle in a pro uniform. Florida can't make the tournament either, as it would be playing without a linebacker, strong safety and free safety. Georgia gets the boot since it lacks a strong and free safety.
That leaves a strong field of eight, seeded by the number of players each college put in the pros: Florida State (41), Notre Dame (41), Miami, FL (37), Michigan (37), Texas A & M (34), Penn State (31), Tennessee (29) and the University of Southern California (28). Hence the pairings are as follows: in one bracket, the winner of USC versus Florida State will play the winner of Texas A & M versus Michigan. In the other bracket, the victor of Tennessee versus Notre Dame will take on the winner of Penn State versus Miami.
In case you were wondering, there's nothing mysterious about a computer simulation. Since the process is grounded in statistics that measure performance, it is more objective than polling. You start by entering the rosters and the relevant numbers for each player. These statistics include a mix of the obvious and the complex -- like yards per attempt for runners and quarterbacks, but also blocking, tackling, break-away ability and other player tendencies. If the passing statistics of quarterbacks like Kurt Warner and Kordell Stewart were similar, the computer will still make distinctions between a pocket passer like Warner and an effective scrambler like Stewart. Likewise, the simulation will treat differently punishing backs that run between the tackles, such as Jerome Bettis, and slashing runners like Edgerrin James, who can run wide. It is this numerical subtlety that makes the results so true to reality, or at least our manufactured reality.
Since these are one-time All-Star games, the computer couldn't analyze what a team's historic tendencies are, say, on third-and-2 -- is it running, passing, changing zone coverage, making personnel substitutions? -- or how often it employs gimmick plays, or whether first-and-10 is a passing or running down. Rather, the computer had to analyze the talent of each All-Star squad and come up with lineups and strategies based on comparisons to teams of comparable skill throughout NFL history.
Once the button is pushed, there is no digital equivalent of thermonuclear war. Nothing like the whirring of machines and blinking of lights that accompanied Frankenstein's birth, either. The simulation takes a matter of minutes to go through all seven games. It generates a complete play-by-play report -- usually 150 to 175 plays per game -- and a box score. All games are played on a neutral site.
Let the games begin.
You must be logged in to post a comment.