College Football Playoff 4 vs 8
One of several completely pointless and never ending debates in the world of sports, yet for some reason it still captures conversations. I’ve been a big fan of the 4 team college football playoff so that “big regular season games matter”, but I wanted some numerical evidence to back it up. So, I parsed the AP Poll website from the BCS era through today (~20 years) and pulled out the win, loss, and AP point values of the top 25 teams. I then averaged the results for the Top 2, Top 4, Top 6, Top 8, Top 10, and Top 16 teams.
Scraped Data (CSV’s)
AP Poll Points Averages
| Year | Top2 | Top4 | Top6 | Top8 | Top10 | Top12 | Top14 | Top16 |
|---|---|---|---|---|---|---|---|---|
| 1998 | 1710.5 | 1624.75 | 1551.5 | 1494.5 | 1419.9 | 1336.08 | 1261.79 | 1196.69 |
| 1999 | 1714.5 | 1629.25 | 1563.17 | 1492.62 | 1422.9 | 1350.92 | 1266.36 | 1198.62 |
| 2000 | 1730.5 | 1668 | 1590.17 | 1512.12 | 1434.3 | 1364 | 1288.93 | 1226.5 |
etc, etc
| Average | 1566.12 | 1499.23 | 1432.87 | 1370.62 | 1305.12 | 1240.15 | 1173.23 | 1107.67 |
Well, in all honestly, a table just isn’t very appealing to look at so its time to pull out the old D3 javascript and get this thing graphed.
Average losses makes the Top 2 teams stand out, averaging slightly than one loss between them in any given year. My mental benchmark coming in was each teaming having a one loss maximum, depending on the year cough 2007 cough. Top 4 definitely accomplishes this task while Top 8 respectibly fails, but interestingly enough Top 6 is still a candidate. My rationale with Top 6 is that the Top 2 teams would get a bye while the next 4 play, and that seems to be what this graph best argues.
Average wins primarily highlights the increasing number of games per season played in college football, which makes the Top 2 going undefeated more often than not even more impressive. Aside from that, the Top 4, Top 6, and Top 8 all squeak by averaging more than 11 wins, my other benchmark for a generic playoff candidate.
Average AP poll points fascinates me, mostly because I have no idea how they generate the point values, which is probably the actual explanation behind why the differences in the averages are so standard. This one doesn’t say much in my opinion, other than the AP thinks the top teams are better than the next ones. Oh well, if you have the data, you might as well graph it.
Conclusion
I still think Top 4 is better than Top 8 with an interesting case to be made for Top 6 as well as Top 2 frankly. Games like LSU vs Alabama or Ohio State vs Penn State just wouldn’t matter much when both teams are going to get in anyway. There is still a lot of data to be had scraped; Strength of Schedule, Strength of Victory, Quality Wins, Quality Losses, FPI all come to mind to provide more insight, but alas I shall save that for my next bout of curiousity. Until then, may the pointless discussions never end…