The Road through the BCS Top 25 starts with the SEC
The Hidden BCS Conference Ranking System
Part 3 of 6 in the BCS Top 25 Unplugged
Part 1 Link: Winning is Everything in Today’s BCS Top25
Part 2 Link: BCS or old AP Poll. Same Trends, Different Day
In the last few days I have shown that the current BCS top 25 ranking system is nearly a perfect trend line of total wins by ranking position and that the AP Poll used prior to the BCS was on a near exact parallel behavior of ranking teams.
Today, I explore the trends of potential bias in the rankings when confronted by teams of equal records. With no good way of evaluating or understanding the evaluation of any given team’s strength of schedule (SOS), I assume that the voters and people programming the computer systems to respond would so do based on a common perception of conference strength. It is much easier to justify a ranking based on a few conference power rankings versus every individual team.
Breakdown by Total Wins and Conference Ranking
0 Loss Teams – There appears to be little bias in this small group. Undefeated seasons are rare but the SEC has had 3 of them since 2006 and all those teams were ranked #1 at the end of the season by the BCS Top 25 composite.
1 Loss – The first real group of 26 total teams produced a fairly clear order. Again, the SEC has the best average ranking for their one loss teams. The PAC12 is over a full rank below them on average followed closely by the Big 12. The Big Ten is a full ranking spot behind the Big 12. The Big East, the Non AQ 1′s and Non AQ 2′s are well behind. The ACC nor Notre Dame has not had a one loss team in the last 6 years, so they have no ranking average.
2 Loses – Finally we see the ACC and Notre Dame jumping into the averages. And just like the 1 loss teams average by conference, the order is the same. SEC #1, PAC12 #2, Big 12 #3 and so forth to the lowest 2 loss average ranking position in the Non AQs #2 teams at an average of #21. Although many conferences are very close to one another, the conference strength perception is holding true.
3 Loses – The SEC remains #1, the BIG 12 #3 and the bottom 3 stay inside the conference trend bias. Not surprisingly, the ACC jumps ahead of the Big Ten for 9-3 teams. The PAC12 takes a bias hit in dropping from #2 to #4 for this segment of win/loss rankings.
4 & 5 Loses – Right back to the conference trend bias. The average rankings are getting thin with just teams ranked #22-#25 but the top 3 ranked conferences with 1 to 3 loses have 16 of the 19 total top 25 rankings from teams finishing with 4 or 5 loses.
Average Yearly Top 25 Rankings
After seeing the very clear conference strength/power rankings by the computers and voters in the BCS Top 25 you would expect the conferences with the total number of top 25 appearances in the final ranking to be in the same order.
The big exception…. the PAC12. The only conference with just 3 out of conference (OoC) games each of the last 6 years.. the PAC12. In a total win based system where 4 easy wins in a OoC schedule puts you in a much better position to get to 9 wins and a top 25 slot. My PAC12 2012 predictions, I forecast 4 PAC12 teams at 8-4. You know what a 4th easy OoC win give the PAC12? That’s right, the potential for 6+ teams to finish the regular season in the BCS Top 25. Instead 4 of the last 6 years the PAC12 has seen 2 teams in the final BCS regular season Top 25 rankings.
Non AQ teams have had 24 appearances in the top 25 final BCS rankings since 2006. That is over triple the number in the AP from 1993-1997. It is 1 more than the Big Ten, 7 more than the ACC and PAC12 and 10 more than the Big East all while winning 24 total games against AQ opponents. Strength of schedule component means something? Riiiigghhttt.
My Read on the Stats
It is abundantly clear to me that win totals will always be king when human voters and computer programmers are involved. Even more so that the past human voter only systems. It tells me several things:
- The SEC is viewed as the toughest and best talented conference and rightfully so
- The Total Wins by 11-1 and 10-2 non-AQ teams trump 9-3 and 8-4 AQ teams nearly every time
- The PAC12 has a very high reputation as a tough conference but is killing itself in the rankings by requiring 9 conference games.
- The ACC and especially Big Ten and Big East teams need to win out or go 11-1 and get help from other conferences beating each other up to get into the top 10.
- The ACC, based on the 9-3 rankings, could be held in voters minds with much higher esteem than most think. If they can get Florida State, North Carolina and Miami to get their heads on straight, it will be interesting.
- The idea of the Big Ten/PAC12 conference OoC match ups is dumb for the teams of both conferences if it is not done by other conferences and non AQs can play 1 or no AQ teams to get into the top 25.
- 3 to 4 non-AQs that get to 10+ wins in 2012 will bump 3 to 4 teams that are 8-4.
- Creating your best odds pre-season Top 25 to be right at the end of the year is a simple two step process
- Dave Bartoo, The CFBMatrix & your Anti-Homer Follow @cfbmatrix
Next: The Best Odds for Non AQs to Crash the Top 25 and how my CFBMatrix predictions create a Top 25