There’s an old saying: “Hypocrisy is the tribute that vice pays to virtue.”
Say what you will about the flaws of the RPI, but it was perfect for one thing: allowing the NCAA to talk out both sides of its mouth:
- “The RPI helps the little guy in the non-conference!” This was somewhat true. The RPI weighted road wins more heavily than home wins, which gave power conference teams an incentive to schedule good mid-majors on the road. Or if mid-majors could only get away games against power conference teams, it rewarded them for winning road games.
- But when March rolled around, you got, “They had a good team, but their RPI just wasn’t high enough.” This was definitely true. For teams like Belmont, the dregs of the OVC depressed their RPI so much that it was nearly impossible for them to get an at-large bid.
- And of course, when you had a situation like Missouri State in 2006, you’d hear, “The RPI isn’t the end-all and be-all.” The Selection Committee could simply point to all of the RPI’s flaws as justification for tossing it completely out the window.
There’s another old saying: “The rich get richer.” It’s difficult to enact any change that doesn’t benefit the already-powerful, because if it hurt them, they’d use their power to veto it. This is true even for things that look like they help the little guy.
Take college football. Several years ago, the College Football Playoff came into existence. At the time, the “mid-majors” of college football—those FBS schools in conferences other than the Big Ten, SEC, ACC, Big 12, and PAC-12—were enthusiastic about the change. The rule for the playoff was simply “the four best teams”, so there was nothing locking them out of the system. And they were going to be making more money. A lot more.
After what happened last year, ask UCF if they’re happy with the current system.
It seems unlikely to me that replacing the RPI with the NET will be a net win for the non-power-conference schools. The hypocrisy that went along with the RPI is gone, but I suspect it’s not because virtue won but because vice stopped paying virtue tribute altogether. Speaking of which...
When the NET was announced, it was shocking to me how little was said about scheduling incentives. When college football implemented the CFP, all you heard from the NCAA was how strength of schedule and conference championships would be important considerations. Your mileage may vary on whether that was true or false, but they at least felt compelled to say something.
So, the NCAA is keeping quiet about what they want to incentivize. Let’s see if we can figure it out for ourselves.
The NET is based on similar principles to KenPom or Sagarin—efficiency is what matters. Efficiency is roughly defined as points per possession. A good offense scores more points per possession, and a good defense gives up fewer points per possession. But you also have to pay attention to how good the other team is. Scoring 1.02 points per possession against Virginia is pretty good. Scoring 1.02 points per possession against VMI is pretty terrible.
You can use math to adjust for the strength of your opponents and develop an efficiency score that shows how many points per possession you’d be expected to score against or give up to the average D-I team. Last year North Carolina scored 1.13 points per possession, but they played a schedule full of good defensive teams. KenPom adjusted their offensive rating to 1.20 points per possession to compensate.
That means that it doesn’t matter who you play. Assuming the formulas are calibrated correctly—and they can be finicky early in the year, though by March there have been enough games that they are going to do a pretty good job—you can adjust for any strength of schedule.
Let’s say you’ve got a team that’s 30th best in the country or so. That’s good enough for an at-large bid. If that team plays in the Big Ten or another strong league, their raw numbers are going to look worse than if that team played in the OVC.
Or to come at it from the other end, if you play in the OVC, you’re going to have to put up much higher raw numbers than a Big Ten school to end up looking the same in the metrics.
The way you put up higher raw numbers is to score more points per possession.
In other words, teams in bad leagues need to run up the score against their awful conference opponents to prove themselves.
EXCEPT... the NET doesn’t factor in margins of victory greater than 10 points.
So if mid-majors don’t have the option of running up the score to improve their NET, that means they’re going to have to schedule tougher opponents out-of-confernece. Same as now.
OK. So do the big boys now have any added incentive to go on the road to play good mid-majors? Under the RPI, they ostensibly did. And under that system, all they had to do was get the W to get full credit. Under this system, they have to win by 10 to get full credit.
So who are the big boys likely to schedule?
Under this system, here’s what you want to do:
- Win by 10+;
- Against the best opponent possible; and
- While minimizing your chances of losing the game.
Let’s look at a Big Ten team last year that was pretty good, but not elite. The type of team a good mid-major would love to schedule. Who’s the best team the likes of Ohio State can schedule that they’d still expect to win by 10 against? Anybody worse, and OSU wouldn’t get credit for winning by more than 10. Anybody better, and while their strength of schedule would improve, so would their changes of taking an L, and wins and losses are still king. So who fits into that sweet spot?
I built a little tool that uses the same underlying methodology as KenPom to predict scores of games. Based on last year’s data, here are the best teams Ohio State could play and still expect to win by 10:
- In Columbus: Fresno State (KenPom No. 79)
- At a neutral site: Northern Colorado (KenPom No. 104)
- On the road: Drake (Kenpom No. 155)
That’s actually pretty good considering last year they played home games against Nos. 315, 253, 243, and 208. There’s room for the Buckeyes to improve their schedule to optimize their NET.
Of course, there’s no reason their improved schedule would have to include mid-majors. If the Buckeyes wanted to schedule someone around No. 80 on KenPom, they could have played South Carolina. If they wanted to schedule someone around No. 105 they could have scheduled Ole Miss.
At the end of the day, economics and perception and pride outweigh any additional benefit from squeezing up another spot or two in the NET rankings. When it comes to mid-majors, we’re in the exact same spot we were with the RPI—power conference schools can improve their metrics by playing them, but they probably won’t, and because of the 10-point cap the mid-majors are going to continue to get punished for playing a weak bottom half of their league. But at least the NCAA isn’t sputtering and spluttering about how this change “helps the little guy”. That’s something.