The Projected Final Top 25  for 2014
It’s not hard to figure out the thinking behind the final college football rankings.
Rule 1: Teams with the most wins get ranked the highest.
Rule 2: Tie breakers of equal wins go to the teams from the stronger conferences and fewer losses.
Rule 3: On average four mid majors make the final top 25 each year.
Rule 4: Over the last 20 years, the only rankings that mattered were No. 1 and No. 2.
LISTEN IN WITH DAVE: Audio review of the CFBMatrix Top 25 
Now all that matters is the opinion of the 13 people that counts, and even then, only the top four spots are meaningful. Nevertheless, I do believe that the playoff committee will use the AP and USA Today rankings as a way to gauge public sentiment. Unless instructed otherwise, I expect them to want to be very close to public opinion.
For my top 25, I have provided a number of different metrics for you to consider. My rankings are based on the above rules and are in relation to my projected finish for each team. There are no mid-majors in the rankings. (Our previews focus only on the major teams and if the playoff committee holds even slightly to its word on strength of schedule, there will never be a mid-major in the playoffs).
Ohio State is No. 1. The Buckeyes have the best combination of talent, coaching and ease of schedule. Let’s compare OSU and Alabama across the board, for reference. OSU has a projected record of 12-0 and Alabama 11-1. Both teams’ talent are ranked in the top 10 in the CFB Matrix, coming in at No. 6 and No. 1, respectively.
Ohio State’s strength-of-schedule is No. 56, compared to No. 3 for Alabama. This is a big reason why the projected records are a game different. Both Alabama and Ohio State have top 10 Coach Result rankings.
The last metric I used was a combination of talent and coaching: Ohio State is No. 3; Alabama No. 1. The top eight teams in my 2014 ranking have top 10 talent/coaching combination. Use these rankings to help see where teams could be over or under ranked and where surprises lurk for different reasons.