Since Stewart Mandel Asked

Stewart Mandel ‏@slmandel 22h22 hours ago
@ChadHansen9972 As I always say, by all means let me know how one whittles five major conferences into 4 spots with no subjectivity involved
9:37 AM - 7 Oct 2014 · Details


Stewart Mandel tweeted about how Condoleezza Rice monitors 40 teams and how Steve Weiberg has watched 92 games as examples of how committee members are prepping for the task of selecting the four playoff teams at the end of the season. I replied to share that I found the whole thing hilariously unnecessary when objective rules could be used. To my surprise, Mr. Mandel offered the response above. Here is my answer.

To start, its best if all FBS teams play equal maximum game regular season schedules versus FBS competition only. For example, all play 12 games plus a possible conference championship game. If necessary, FCS games can be accommodated with weighted values if these games continue. That said, I would place teams most to least:

1-Power Points (Games Won Opponents' Wins - Games Lost Opponents' Losses)
2-Net Wins (Wins - Losses)
3-Schedule Strength (Opponents' Power Points)
4-Net Points (Points For - Points Against)

Once the season ends, the top four teams or, if preferred, the top four conference champions advance to the playoffs. That's it. So simple a caveman could do it. Of course, I'm sure there will be plenty of people to offer their welcomed criticisms of my idea. I'd be surprised if anyone could tell me how a selection committee is better than my idea much less necessary. Here I offer rules that define best/most accomplished season and leave it up to the teams to beat each other based on those rules. With a committee, teams will never know what separates the winners from the losers. Even after the selections are made and the committee explains its reasoning behind its choices, those reasons are not guaranteed to favor teams that own those arguments in subsequent years.

So what kind of results does my point system yield? I have 36 seasons worth of results based on FBS games only. Without the benefit of teams playing equal maximum game schedules or knowingly competing based on its rules, the point system matched the AP poll, Coaches poll, and/or BCS on 108 of 144 top four teams. That is an exact three for four average. So the difference between objective and subjective is one team even without equal maximum game schedules and knowledge of the rules. That said, let's just compare the point system to the AP poll. Since 1978, the two methods have agreed on 103 of 144 top four teams which is slightly below a three for average. More than half of the 41 teams that missed the rival method's top four were that method's number four team. The 41 disputed top four finishes make for 67 disputed pairs. In those pairs, teams favored by the point system own a 34-19-14 advantage in most wins versus AP ranked teams. The advantage is 31-8-14 just among major conference/independent teams. The AP owns an 11-3 advantage when the point system favors a mid-major team. Otherwise, where disputed top four teams are concerned, the point system dominates the AP poll at favoring teams that beat more AP ranked teams.

So far we have a point system that matches the AP on nearly three top four teams, most disputes involve #4, and the teams its favors average more wins versus teams ranked by the AP poll. Again, that is without the benefit of equal maximum game regular season schedules and knowledge of the rules in play. But suppose the goal were to match the AP, Coaches, or BCS on all top four teams. How close is the point system to that "goal"? The answer is 44 games. Over 22,000 FBS games played in 36 seasons and I would only need to change the result of 44 games to match the top four teams in one of the major ranking systems employed by college football. In 25 of 36 seasons, I need one game or less.

Want more? 155 different teams have finished in the AP, CP, or BCS top fours. 143 of those teams controlled their point system top four destiny based on how the season played out. Among the 12 that did not, eight would have needed help in a single game if undefeated and eight played less than a full FBS schedule. Only two of 12 played a full FBS schedule and needed help in two or more games.

Did I mention the point system only needs wins and losses and opponents' wins and losses to separate any pair of teams better than 99% of the time? What does a committee have on this point system?