SMQ: How does the AP Top 25 correlate to College Football Playoff rankings?

With the first College Football Playoff Top 25 coming out in Week 10, let’s look at how the AP Poll has correlated to the first set of rankings in the past.

Ever since the shift to the College Football Playoff in 2014, the AP Top 25 and the Coaches Poll have seen diminishing impact on the sport. Once the CFP selection committee releases its first set of rankings, usually in late October or early November, the other human polls effectively become afterthoughts.

While they are fun parlor games, and while the NCAA still recognizes both polls as legitimate selectors of the national champion, the majority of fans focus exclusively on the CFP Top 25. That leaves legitimate questions about whether there is any correlation that can be found between the oldest poll and the most important poll of college football’s current era.

Can we find any linkages between where a team lands in the AP Top 25 and where they might end up on the first CFP rankings of the season. In this week’s Sunday Morning Quarterback, we backcast where teams place in the first CFP selection committee rankings and the last AP Top 25 before the committee’s release against their final position in the race for a Playoff spot.

Keep in mind that these are small sample sizes, given that we are only five years into the College Football Playoff era. But with that caveat on the table, let’s look at what general trends are starting to appear from a half-decade of rankings from a dozen selectors crafting their rankings behind closed doors and how they correlate to the more transparent poll process managed by the Associated Press.

Characteristics of top-four programs

The ultimate goal in the College Football Playoff Top 25 is to land in one of the top four spots in the final poll of the year. Landing one of those positions correlates to one of the four semifinal spots in the playoff bracket, and ultimately it is rarely shocking when we see a team land one of those positions.

As much as we manufacture drama from season to season in the quest for those four berths in the bracket, it is relatively simple to see what kind of team actually makes it into one of those spots by the end of the season:

• 11 of the 20 semifinalists opened in a top-four position in the first CFP Top 25 of the season.
• 16 of the 20 semifinalists were ranked in the top six of the opening CFP Top 25.
• Only two semifinalists (2014 Ohio State, 2015 Oklahoma) were ranked lower than No. 7 in the first set of CFP rankings.

Top-four teams in October and early November, then, tend to be top-four teams in December once the dust settles on conference championship games. Here is the full list as sorted by final position in the CFP Top 25:

In terms of the individual top-four positions, here are some quick thoughts on each of them.

1. Over the first five years of the College Football Playoff, Clemson and Alabama have been the only two teams to land the No. 1 seed in the semifinals. Clemson tended to get a bump from the CFP selection committee relative to where they sit in the previous week’s AP Top 25 (+2.5 spot average), while Alabama has been ranked No. 1 in two of the three seasons where they landed the top seed. In three of the first five years of the Playoff bracket, the No. 1 team in the opening CFP Top 25 is also the No. 1 team at the end.
2. All five teams that have earned the No. 2 seed in the CFP bracket were ranked in the top five of the opening CFP Top 25. Only two of those five, however, were ranked in the top five of the previous week’s AP Top 25, showing a greater discrepancy between the two ranking systems. Teams that finish the regular season at No. 2 tend to be rated two full spots higher by the CFP selection committee than where the AP voters rank them the previous weekend.
3. The No. 3 team in the final CFP rankings tends to fall somewhere in the top seven spots of the initial CFP Top 25 as well as the previous week’s AP poll. Among the various programs that land that No. 3 ranking, there is less variance than you tend to see between where the AP puts at team at the end of October and where the selection committee puts a team in their opening list.
4. Among the teams that finish the season in the No. 4 spot, there is the widest variance in where teams open in the CFP rankings. In 2017, Alabama was No. 1 in the AP Top 25 and No. 2 in the opening CFP list before an Iron Bowl loss opened the door for the Crimson Tide to end the year at No. 4 despite failing to win their division. Oklahoma was No. 15 in the opening CFP rankings in 2015, and Ohio State was No. 16 a year earlier when they emerged as the inaugural champions.

What differentiates top-four teams from the first five out?

The big argument when the BCS system was still in place was always whether the No. 3 (and often No. 4) team deserved a shot at the national title. That was a big spark behind the eventual development of the current plus-one playoff model, except now we are left to wonder about whether the No. 5 or No. 6 team might be deserving of a top-four position.

In any playoff system, these questions will always be inevitable. What we must look to discover, then, is the point at which there is a clear disparity in quality where expanding further would be a case of diminishing returns. Is that really the case for teams in those next two spots just on the cusp of inclusion?

Short answer? No.

A longer answer? There is no real drop in quality for the next two teams on the outside looking in at the College Football Playoff. You can actually find evidence that says the teams that finish the regular season ranked No. 4 through No. 7 are essentially indistinguishable both in terms of where they are ranked in the opening CFP rankings and in the AP Top 25.

The average opening ranking in the College Football Playoff for a team that lands at No. 4 is 9.0, and they land an average ranking of 8.0 in the AP Top 25. The selection committee basically sees No. 5 through No. 7 teams as roughly equivalent, with an average position of 9.2, 8.0, and 7.4 respectively in where they open in the first CFP rankings of the year.

Ultimately, any questions for an expanded playoff would fall on the debate over the No. 8 and No. 9 positions. On average, eventual No. 8 teams are ranked by the AP Top 25 inside the top 12 while the CFP committee puts those teams at No. 10 in their opening list. No. 9 teams are listed about 1.8 spots lower in the AP Top 25, but one spot higher in the College Football Playoff hierarchy at No. 9.

An eight-team playoff, then, would be completely logical given the way these teams tend to parse out and the fact that we have seen teams like Ohio State win as No. 4 seeds. When access is expanded, we find that relative quality is more expansive than an artificially deflated quantity of opportunity allows us to witness.

Would there be a rationale for expanding beyond eight?

This is where things get trickier. Given the current composition of conferences at the FBS level, and the dividing line built in between Power Five leagues and Group of Five leagues, any expansion would likely need to include at least one automatic spot into the bracket for the top mid-major in a given year lest the CFP court congressional ire like the BCS did with its own restrictions on BCS Busters.

What the late October AP Top 25 rankings and the initial position of teams in the first CFP rankings of each season show us is that there are clearly different divisions of teams that fall right near (but not directly on) logical playoff breaking points.

Teams ranked No. 1 through No. 3 have very similar characteristics. Teams ranked No. 4 through No. 7 fall into a similar level of quality, and the same can be said about teams from No. 8 to No. 11. From No. 12 through No. 17, there is a corresponding level of quality that is evident before the final month of the regular season.

In the end, more inclusion tends to reveal more teams that can make a play at the national championship. As soon as the top-two system of the BCS was expanded into the top-four bracket of the CFP era, No. 4 teams proved capable of winning. And as the numbers show, that could just as easily be a No. 5 or a No. 7 team breaking through and winning the national title as well.

What does this say for 2019?

Well, the other aspect of looking into these numbers is an attempt to get back into the minds of the dozen selectors who populate the CFP committee ahead of next week’s inaugural release for 2019. The committee is releasing their rankings a week later this season than they have in the past five years, but based on the trends we have seen in the recent past — and barring Week 10 upsets that topple teams from their perch — we will almost certainly see the following teams in their top four spots:

• Ohio State is going to be the third team ranked No. 3 in the AP Top 25 that will be ranked No. 1 by the selection committee.
• Alabama and Clemson will both be right behind the Buckeyes in the rankings.
• The winner of the Florida-Georgia showdown in Jacksonville will leapfrog Penn State to take the No. 4 spot.

Of course, with upsets in the top-10 ranks of the AP Top 25 each of the past few weeks, there is no guarantee that all of the top teams will remain undefeated. There are also outliers like Minnesota and Baylor who are currently undefeated but ranked lower by the AP voters than they likely will be by the selection committee.

We will have more definitive projections next Saturday night in our weekly look at the Top 25 projections here at Saturday Blitz. The data show that there is some ability to predict placement from the AP Top 25, though it is as yet an inexact science.