College Football Playoff: Mock selection exercise raises issues real committee will face.
GRAPEVINE, Texas -- A moment after I was elected the chair of the fake College Football Playoff selection committee Wednesday night, I turned to the chair of the real selection committee and asked a question. What do I wear?
“I’m a tie guy,” Arkansas athletic director Jeff Long replied. He said the dress code for him, former Secretary of State Condoleezza Rice, ex-Big East commissioner Mike Tranghese and the other 10 members of the real committee is business casual, but Long likes to gussy up a little more so he can exude an air of authority. This apparently helps when one is trying to shush chatty types such as West Virginia AD Oliver Luck and all-time Ole Miss great Archie Manning, two former quarterbacks who passed on their genes spectacularly well.
When Long saw me Thursday morning as we prepared to enter the room where the actual committee will decide who plays in the inaugural playoff, he clucked his tongue. I had chosen one of my more understated ensembles -- a red sportcoat and a navy bowtie. “Too bright,” Long said. “It needs to be gray, navy or black.”
Hours later, after we had selected our four playoff teams, rounded out the Top 25 and populated the big-money bowls using data from the 2008 season, I understood why Long would want to blend into the scenery. When the time comes for the chair to explain why each team was ranked where it was ranked, the choices will be loud enough. No need to draw more attention.
The mock selection exercise confirmed one thing I suspected: This will be a lot like the NCAA men’s basketball tournament selection process. Multiple votes involving overlapping sets of teams allow members to parse teams more thoroughly, and they tend to suck conference superiority arguments from the room and force members to examine each individual team’s résumé more closely. The mock selection exercise also surprised me. I thought 13 people charged with finding the four best teams would have a fairly similar idea what “best” meant. As it turns out, those definitions vary wildly.
The employees of the College Football Playoff deliberately chose 2008 for this exercise since it would challenge us in a number of ways. After Championship Saturday in that season, eight FBS teams (Alabama, Ball State, Florida, Oklahoma, Penn State, Texas, Texas Tech and USC) had only one loss. Two FBS teams (Boise State and Utah) were undefeated. Selecting and seeding the playoff teams would be agonizing. (For the purposes of the exercise, we used conference alignments circa 2008, but today’s contractual arrangements with the bowls.)
That season Florida beat Oklahoma in the BCS title game after the Gators defeated 12-0 Alabama in the SEC Championship Game and the Sooners beat Missouri in the Big 12 title game. Oklahoma secured a league championship berth only after winning a tiebreaker with fellow 11-1 Big 12 South teams Texas and Texas Tech, giving the Sooners the division title by virtue of having the highest BCS ranking. Undefeated Utah would go on to thrash Bama in the Sugar Bowl, but we had to pretend we didn’t know that. For our purposes Crimson Tide offensive tackle Andre Smith remained happily eligible.
The night before arriving, I put together my own Top 25 for the 2008 campaign after studying the records and stats from that year. I fully expected Oklahoma, Texas and Florida to make the playoff. I wasn’t sure about the fourth spot, but I figured it would go to Alabama or Utah. I got the first part correct. I got the second part completely wrong.
Chairmanship doesn’t offer as much power as you’d think. He or she can’t influence the discussion any more than anyone else. About the only time the chair can exert any authority is when ending a discussion and moving on, but I tried to allow all of the members to make their points. Long noted that in the two practice selections performed by the actual committee, his greatest challenge was encouraging quieter personalities to speak up and getting stronger personalities to let the others get a word in edgewise. I didn’t have to worry about the first issue. With a room full of sportswriters and TV personalities, there were no timid participants. There was no shortage of opinion. The chair is more of a traffic cop.
The selection process begins with each committee member inputting his or her Top 25 in no particular order. This becomes the pool from which the committee will select. Because of conflicts of interest, certain members are not allowed to vote for certain teams. As the mock committee’s version of Long, I could not vote for Arkansas. As mock representatives of USC AD Pat Haden, the team of Chris Dufresne (Los Angeles Times) and Kirk Bohls (Austin American-Statesman) couldn’t vote for the Trojans. Our two-headed Haden would later need to be recused from nearly every discussion involving the potential playoff teams. Our panel wound up selecting 33 different teams for the pool.
Next each member lists his or her top six teams, in no particular order. On a Dell laptop provided by the College Football Playoff staff, committee members lock in their choices. The results are tabulated and the top six vote-getters are shown on the big board. These six -- presented in alphabetical order -- got the honors.
So much for Utah. But wait, what about Utah? I asked that, and so did ESPN’s Holly Rowe, who, alongside ESPN colleague Rod Gilmore, was playing the role of current Big East advisor and former NCAA vice president Tom Jernstedt. The Utes had won every game that season. They had an excellent quarterback (Brian Johnson) and serious speed for a Mountain West team. Playoff director Bill Hancock informed us that if we weren’t satisfied with the six after ranking the teams, we could call for a re-vote. Still, we discussed Utah anyway. Other fake committee members pointed out that beyond a narrow win over a quality TCU team -- but not nearly as good as Oklahoma’s win over TCU -- and a win over a 10-2 BYU team that had zero good wins, the Utes didn’t have the résumé to compete with the others.
We also discussed the differences between Alabama, Penn State and USC. The crowd that values a conference championship more than anything else was down on the Crimson Tide. I don’t worry as much about conference titles, because I don’t feel the choice of a few like-minded schools to come together in the 1930s should affect a football decision the following century. But I did weigh other factors. USC played Ohio State and Notre Dame in its nonconference slate, as well as a full Pac-10 round-robin. Alabama had opened against Clemson, which was supposed to be good but turned out to be average. Of that group, Ohio State -- which USC beat handily -- was the best nonconference opponent. Alabama had the better loss. The Tide fell to 11-1 Florida in the SEC title game. The Trojans lost at Oregon State -- which finished 8-4 -- on a Thursday night in late September. Further complicating matters, Big Ten champion Penn State had crushed Oregon State in State College in early September. (Remember, we don’t know that USC pounds the Nittany Lions in the Rose Bowl.)
After going through some of the stats provided by SportSource Analytics, I found myself leaning toward the Trojans for the fourth spot. The Relative Scoring Offense and Relative Scoring Defense stats were especially interesting. They measure how many points a team scores or allows relative to the opponents’ averages. USC scored 29 percent more points than its opponents’ opponents averaged. Alabama scored 42.7 percent more. USC’s superiority came on defense. Teams scored about half their average (48.9 percent) against the Tide. Teams scored only about a quarter of their average (28.1 percent) against the Trojans. This was far and away the best Relative Scoring Defense in the country. (It’s important to remember that the SEC wasn’t THE SEC in 2008. It didn’t pull away from the pack until ‘09.)
So, when we voted, my rankings looked like this:
6. Penn State
After this step, playoff staffers don’t show the committee’s composite top six. They show the committee’s composite top three. The other three schools are thrown back into the pool. Here were the top three. Needless to say, I was outvoted.
Florida had lost at home to Ole Miss, which was easily the worst loss among the one-loss contenders. That’s why I dropped the Gators to No. 3. Oklahoma had compiled the best overall résumé with quality nonconference wins over 10-2 TCU and 10-2 Cincinnati, as well as an annihilation of 11-1 Texas Tech on its league slate. But Texas had beaten the Sooners by 10 points on a neutral field. That’s why I had Texas No. 1. However, with the exception of the Sporting News’ Matt Hayes (Tranghese), my fellow mock committee members disagreed. They cited Florida’s offensive and defensive dominance in every game except the loss to Ole Miss and the Gators’ win over Alabama in the SEC title game.
We would get another crack at picking the No. 4 spot by design. Since that’s the cutoff point for the playoff, the voting process intentionally forces the committee to essentially vote on that spot twice before it can move on. But we weren’t ready to move on. Those still clinging to Utah wanted a re-vote of the top six. We voted again and wound up with the same pool of six and the same top three. So, we moved on to the toughest choice.
With USC, Alabama and Penn State back in the pool, we were asked to select the best six of the remaining 30 teams. When our picks were tallied, those three teams were joined by Boise State, Texas Tech and Utah. We were then asked to rank those six. When the votes were tallied the top three were added to the ranking sheet and the bottom three were thrown back in the pool. These were the next three we chose.
6. Penn State
We chatted and were about to move on to the next step when I asked Hancock a question. “Did we just pick the playoff teams?” Basically, he said, we had.
None of us was ready to be done with that part of the process. It had happened too quickly. We needed to discuss it further. Stewart Mandel of Fox Sports -- you may remember him from the occasional SI.com column -- suggested we assign a person who felt strongly about each of the three teams ranked four through six to make the case for their team. Because USC was involved, Dufresne and Bohls (Haden) were not allowed to speak. In a real committee meeting Haden would leave the room.
Mandel (playing Tom Osborne) made the case for USC, emphasizing the defensive dominance and conference title. Yahoo! Sports’ Pat Forde (Tyrone Willingham) made the case for Alabama. Forde started by mentioning that Bama was ranked No. 1 for much of the season. I reminded him we shouldn’t care what a bunch of dumb sportswriters think. He then moved on to the dominance of the Tide’s win during Georgia’s Blackout night in Athens and of how close Alabama came to beating Florida for the SEC title. Rowe (Jernstedt) argued on behalf of Penn State. She cited the Nittany Lions’ common opponents with USC (Oregon State and Ohio State) and reminded everyone that Terrelle Pryor did not start for the Buckeyes at USC. Also, tailback Beanie Wells missed the USC game with a foot injury.
After this, we voted again. This time the result changed.
5. Penn State
Were I the real committee chair, this is the part where I’d call my wife and tell her to buy a bomb shelter. I would have to go on national television and explain that not only had we left 12-1 Bama out of the playoff, but we also had dropped the Crimson Tide behind a Big Ten team. College Football Playoff chief financial officer Reid Sigmon, who tallied our votes, told us the margin between No. 4 and No. 5 wasn’t close. Most of us were convinced USC was the fourth-best team.
We took a break, still not quite ready to admit we had decided the playoff. During the break ESPN’s Gilmore had an excellent idea. All of the members should explain what matters most to them when picking the best team. Then we would discuss the playoff teams once more and re-rank No. 1 through No. 4.
For me, conference titles matter a lot less than quality wins. I also prefer demanding nonconference schedules, and, when all else seems equal, head-to-head matters a lot. For ESPN’s Heather Dinich (Rice), league titles hold more weight. “Conference championships should matter,” she said. “I don’t think they should be a guarantee or a lock, but I do think you should be rewarded.”
Dufresne (Haden) agreed. “If you’re going to write into the mission statements that conference championships are important,” he said, “then you have to consider that.” Jerry Palm of CBS (Barry Alvarez), the most mathematical of the group, values margin of victory even though the selection guidelines state margin of victory should not be incentivized. “At the end of the day,” Palm said, “the scoreboard has points and not yards.” Ralph Russo of The Associated Press, half of a team playing Clemson AD Dan Radakovich, agreed with Palm. “It can be misleading over one game,” Russo said. “It’s not misleading over 12 games.”
Every person’s answer differed. They probably don’t match the opinions of the committee members they were representing, but the diversity of opinion in that group is likely just as broad. “It’s 13 individuals with different perspectives and different weight on different things,” Long said. “That’s the human element -- whether we like it or not.”
More familiar with each other’s human elements, we re-ranked the top four teams. Once again, I voted this way:
Once again, the committee voted this way:
We had voted three times, and the order hadn’t changed. We could have voted five more times and nothing would have changed. So, we decided we were at peace with it. Those were our playoff teams and seeds. Florida would meet USC in the Sugar Bowl, while Oklahoma and Texas would have a rematch in the Rose Bowl. The real committee also mocked the 2008 season, but it came up with a different four and a different order. Of course, the playoff staff said we remembered the details of the ‘08 season much more vividly because many of us had covered the games we discussed.
While the playoff is the part most people care about, we still had work to do. The real committee must keep ranking teams to determine who plays in the other four big-money bowls. This has a serious impact on conference coffers and team prestige, so the committee must take it as seriously as picking the playoff field.
At this point we had a top six, so we repeated the process described above to wind up with Nos. 7-9. They became Utah, Texas Tech and Ohio State. After that the voting changes only a little. Instead of picking and ranking six teams to get three, we picked and ranked eight to get four. That is how we determined Nos. 10-25.
By the time we reached this juncture, one task had already been completed. The highest-ranked Group of Five team (The American, Conference USA, MAC, Mountain West and Sun Belt) automatically gets a berth into a high-dollar bowl. Utah was already locked in. All that remained was to determine the rest. In 2008, we could have stopped at No. 11. Some years will require the committee to dip much deeper into the rankings for that top-ranked Group of Five team, but the committee is always charged with creating a Top 25, because Top 25s are good for television and ESPN will produce a weekly rankings reveal beginning Oct. 28.
We worried that if we didn’t take the bottom of the Top 25 seriously and screwed something up, we might damage the credibility of the top of our rankings. That’s why, after finishing the entire Top 25, we took a break and returned to perform a “sanity check.” And we found a potential problem.
We worried we had ranked Cal too low. Cal didn’t even make the final BCS Top 25 in 2008, but we had diverged from the Coaches’ and Harris poll voters and computers on several occasions. For example, we didn’t reward 12-1 Ball State for staying unbeaten playing lackluster competition until losing to Buffalo in the MAC title game. Ball State finished No. 22 in the final ‘08 BCS standings even though many of the teams ranked below it would have beaten the Cardinals. Our problem was we had Cal below Michigan State. The Spartans were 9-3 and the Bears 8-4, but Cal had played a tougher schedule and had beaten Michigan State in the season opener. So, we grabbed three-team section and re-ranked it.
20. Michigan State
21. Ole Miss
… became this …
21. Michigan State
22. Ole Miss
The Top 25 done, we began populating the bowls. The playoff was done, and so was the Orange Bowl. ACC champion Virginia Tech (our No. 13) was guaranteed a spot. So was the highest-rated nonconference champ from the Big Ten, SEC or Notre Dame. For 2008, that was Alabama.
After that, it’s just the top-ranked remaining teams. They get placed in bowls to generate good matchups and sellouts, but the committee can’t choose a lower-ranked team for the sake of a matchup. Utah, then a Mountain West team, was already locked into a big bowl. TCU, then in the Mountain West, and Boise State, then in the WAC, made it as well. It’s safe to say this year’s committee will not put three Group of Five teams in the big-money bowls. We didn’t want to match Boise State and TCU, which were actually matched in the Poinsettia Bowl following the 2008 season and the Fiesta Bowl following the ‘09 season. Not having Alabama to place in the Peach or Cotton Bowl reduced our flexibility, but we did guarantee a Cotton Bowl sellout by matching TCU and Texas Tech, which weren’t in the same league at the time. Boise State returned to the Fiesta Bowl to play Ohio State. We wound up putting Penn State and Utah in the Peach Bowl because Penn State travels well and because we figured Utah fans would have little trouble flying from one Delta hub to another.
After that, I asked my fellow committee members to become their reporter selves again and pepper me with questions, just as we will pepper Long with questions. Hopefully, Long will fare better than I did.
He has been an AD for a long time, so he has mastered the art of talking without saying anything. I tend to say exactly what is on my mind, and that got me in trouble during our practice press conference. Someone asked the $64,000 question -- why USC over Penn State or Alabama? -- and I tried to hammer the talking points about USC’s defense and nonconference schedule, but I struggled explaining away the Oregon State issue.
Then, someone asked why Texas was ranked behind Oklahoma despite beating the Sooners by 10 points on a neutral field and having the same regular-season record. I answered honestly. I pointed out that I had voted Texas No. 1 on multiple occasions. This is a no-no, because I basically said, “I had it correct, but these bozos messed it up.” That’s no way to make friends or extend one’s time as committee chair.
Long took copious notes throughout the process, looking for ideas he can use with the real committee. Gilmore’s suggestion of having every member explain what matters to him or her is something Long should definitely use. Members should answer that question at the beginning of their first real meeting later this month. That way, everyone will know where everyone else is coming from when the discussions begin.
No matter what the real committee decides, it is going to make a lot of people mad. But if it knows exactly why it chose the teams it did, it will be able to feed Long an explanation that might allow him to avoid buying that bomb shelter.