Jimmystats: A New Bowl Eligibility Plan

The 2013 Motor City Bowl between Pitt and Bowling Green, via StadiumJourney blog.

In our roundtable yesterday I suggested a new way of calculating bowl eligibility. It struck a chord, and it's offseason, so I thought I'd do a follow-up.

The Problem: With 40 (plus the NC) bowls, the bowl field has now expanded to 80 teams, or 62.5% of what's currently 128 FBS schools. However the old six-win provision for bowl eligibility remains mostly intact, disqualifying mostly mediocre schools who played much harder schedules in favor of bad, barely eligible, barely FBS teams.

This system doesn't just create less watchable bowls. It incentivizes schools to pad their non-conference schedules with noncompetitive opponents and FCS programs, and incentivizes conferences to play fewer conference games lest they disqualify more of their teams from bowl play. The result is a less competitive, and thus less interesting, football season.

My Proposal: A simple points system:

• 3 points for a win over any team in the final CFP Top 25
• 2 points for a win over any Power 5 school not included above
• 1 point for a win over any FBS school
• -1 points for a loss to any FCS school

I initially proposed 7 points as the cutoff for eligibility, but as one reader correctly predicted, this is still too exclusive. So I amend that to the highest bowl points level you need to fill the available bowl games is your bubble region.

[After the jump: I try this out with the 2015-'16  bowl field]

Foreseeable problems:

1. Wins are not equal and not counting losses and score and stuff removes useful data. That's true, but as I outlined in yesterday's thread I don't want advanced stats to play into this. The point of football is to win, not to look like you would win. Just counting wins is too problematic for the reasons I outlined above, but wins should still be heavily favored. Late-season rivalry games with a bowl game on the line are fun and cool.

2. Sometimes Power 5 schools—ahem Kansas—are worse than most mid-majors. It depends on the mid-major—some are worth 3 points because they're Top 25. This is admittedly unfair, but it's crucial for providing the right scheduling incentives.

Power 5 schools are for the most part those with larger stadiums, bigger followings, and greater histories, i.e. more interesting matchups for fans. There should be extra incentive to play them and play each other. That means scheduling each other in the non-conference season, and increasing the number of conference games, two actions which under the current system lead to a lot of halfway decent Power 5 teams getting disqualified.

3. The Top 25 isn't a great measure.Nick Jervey noted the fickle nature of the Top 25 might cause some controversy. I could see this happening where a team on the bubble beat a mid-major just outside the Top 25—for example Indiana's win over Western Kentucky is worth just 1 point, not 3, because the Hilltoppers barely missed being ranked in the CFP Top 25.

I'd argue that you have to make the cutoff somewhere, and a bubble team with an upset over a top 10 team won't worry about this, while beating a team that's ultimately ranked between 20 and 30 shouldn't be a sure bet. Plus for the first time in college football history the Top 25 will mean something.

Let's try it out

Here are the non-bowl teams who would have qualified under a 7-point threshold:

• 5-7 Texas (11 points)

Derf. Texas had wins over Oklahoma (3 points), Baylor (3), KSU (2), Kansas (2) and Rice (1). They're also a massive, rich, well-traveling fanbase that almost any bowl would want even on a down year. They're a case example of a 5-7 team not going to a bowl this year who would have been a better selection than the bottom eighth of bowl teams.

Oh, and they're just one team, which probably isn't going to be enough to offset bowl-bound teams who fell below 7 points. Here are those "bad" bowl teams that wouldn't have qualified under a 7-point threshold:

• 7-5 Central Michigan (6 points)
• 7-5 Middle Tennessee State (6)
• 7-5 New Mexico (6)
• 7-5 Colorado State (6)
• 7-5 Western Michigan (6)
• 7-5 Akron (6)
• 6-6 Tulsa (6)
• 6-6 Georgia State (5)
• 6-6 Nevada (5)
• 6-6 Utah State (5)
• 5-7 San Jose State (4)

If you're looking for a "drop those guys!" team, San Jose State beat only New Mexico, New Hampshire (FCS), UNLV, Hawaii and Fresno State, and got in because they were 3rd in APR (you can't make this stuff up!).

Replacing the obviously best 5-7 team for the obviously worst doesn't require a whole system. However I've created a problem by having a cutoff that's 10 spots short of filling the bowl lineup.

So let's look at all the teams with six points:

Currently in: CMU, MTSU, UNM, CSU, WMU, Akron, Tulsa

Currently out: Illinois (5-7), Mizzou (5-7), Kentucky (5-7) and Virginia (4-8)

We are left with 10 spots to fill, and a field of 11 teams:

* = currently attending a bowl   //  bold = head-to-head

Virginia played a tough schedule but has just three wins over FBS teams, only one of which—Duke—over a team that's bowl-bound. They're an obvious choice to be left out.

So how'd it go?

Ultimately I replaced Georgia State (6-6 with a loss to an FCS school), Nevada (6-6), Utah State (6-6) and San Jose State (5-7) with four Power 5 teams that went 5-7: Texas, Illinois, Kentucky, and Missouri. None of the teams left on the bubble had much of a case to be a bowl team, and none of those who got in were awful.

I tried it again for last year, when there were 38 bowl games, and it ended right at 7 points so the field was clear. In: Northwestern (5-7), Cal (5-7), Georgia Southern (9-3), Michigan (5-7), Virginia (5-7), and Oregon State (5-7). Out: Fresno State (6-7), South Alabama (6-6), SD State (7-5), Navy (7-5), Houston (7-5), and Arkansas State (7-5).

Lessons

Not too long ago there was only room for about the top third of FBS teams, so the arbitrary line of "need six wins" was close enough to "should be among the better half of teams in the country" for its purpose. Now that the field has expanded to the top 62.5% of the league, having a cut-off that's conceptually equivalent to 50% is creating an unnecessary crunch.

The teams disproportionately affected by that are mediocre Power 5 programs with tougher schedules. It has created a situation where mid-major teams are systematically favored for bowl eligibility, and programs are rewarded for not scheduling interesting games during the season.

It's also forced bowls to scrape the bottom of the barrel of what constitutes "eligible." Even before they acquiesced to Nebraska and let SJ State in, Georgia State was in the bowl field. Georgia State lost to an FCS team, beat just one team (Ball State) that's been FBS more than a few years, and maxed out home attendance at 11,500. But they're technically 6-6. Meanwhile Texas, at 5-7 with wins over Baylor and Oklahoma, is staying home because they opened the season at Notre Dame rather than an FCS snack.

A common response to this is hurr durr "losing season don't deserve." That comes from growing up thinking bowls are only for good teams. Two thirds of a massively bloated FBS go to bowl games now.

An arbitrary distinction that consistently disqualifies mediocre Power 5 schools and those extra practices and recruiting trips to barely-out-of-the-FCS Sun Belt teams doesn't serve college football. It doesn't serve the bowls who miss out on Texas money, or the fans who want to see better football over Christmas Break, or the Sun Belt schools who can barely afford the bowl swindle. If the NCAA would cap the bowl games at 30, or better yet fix the dang economics of them by forcing the bowl organizers to cover their own overhead, "hurr durr win at least half your games" at least would make some sense. Until they get rid of eight or ten bowl games, this response should get an eye roll.

My solution is far from perfect, but it can be calculated simply, and emphasizes the two things football teams should be doing: scheduling interesting matchups, and winning football games.