Are preseason polls any goods (part 3): overrated college football teams

In the final part of this 3 part series looking at preseason polls, we examine the movements of specific teams and conferences over the past 20 seasons. In particular, we look for the most overrated college football teams.

Conference Strength

We calculate by ‘rank points’, such that being #1 in the polls is worth 25 points, #2 worth 24, etc. This way, a conference with a few strongly ranked teams will compare favorably to one with more, yet poorly ranked teams.

In the following graphic of pre-season polls, we see a few noteworthy things:

  1. The SEC has steadily increased preseason vote share over the last 20 years
  2. This has come at the expense of the Big 12. This is not a shock as the number of teams in the former has grown while the number of teams in the latter has fallen
  3. We see the demise of the Big East and the virtual elimination of any pre-season ranks for non P5 teams by 2017

Here’s where things get curious. If we look at how those conferences performed AFTER the season, every single P5 conference drops in rankings in aggregate. This accounts for teams moving into rankings and dropping out of rankings over the course of a season, for example. It is purely the sum of the raw rank points at the beginning and end of a season:

Some take-aways from this chart:

  • The Big 10 performs the best of the P5 conferences. A net add of 22 in most recent decade is impressive.
  • The SEC is interesting. Between decades, its pre-season numbers increased by 185, but end of season numbers did not match. Evidence of a bit of over-hyping?
  • Comparing on percentages suggests the ACC is the biggest dog in the clubhouse. It looses a whopping 14.7% of its pre-season poll slots.
  • The Big 12 has disappointing 10% loss in aggregate poll slots.
  • The Pac 12¬†is just kind of there. It gathers comparatively fewer ranks and generally does OK at holding on to them.
  • Noteworthy is the upward move of the non-P5 conferences. Whether or not this reflects actual quality of the teams or just voter sympathy to rank a team that won a lot of games (UCF) highly is undetermined.

Team Performance

Using the same metric, we can identify the biggest movers over the past 2 decades, both to the upside and to the downside.

To recap, a team that fell from #1 to unranked in 1 season would suffer an implied score of -25. A team that went from #15 to #10 would show a gain of +5.

The Serially Overranked

The ten teams suffering the biggest poll moves to the downside are the heavy hitters of college football. This is somewhat unsurprising as they are a fixture of almost every pre-season poll.

In a way this list seems oddly familiar of teams that always seem perpetually ranked pre-season in top 5 only to disappoint, so the data certainly supports that finding.

The Serially Underanked

This list is also completely not surprising. Gary Patterson probably deserves a raise for what he’s accomplished.

The total hierarchy of every team by season is at the bottom of this post.

If you enjoyed this post, please follow us @fattailedhappy and/or share it with any other fan of data and college football.

 

Read more about projected win totals by conferences in part 2

Read more about validity of poll projections in part 1

 

Appendix: The total team breakdown of poll movements