Underdog Stories / Sports analysis & explainers

How does non-conference scheduling affect a team's NET ranking?

How does non-conference scheduling affect NET ranking in college basketball? Bye games aren't the hack people think; Q3 teams you can beat big are.

By Brett Hanson · Underdog Podcasts · July 10, 2024

Share on X

Across November 2023, Iowa State beat Green Bay 85-44, Lindenwood 102-47, Idaho State 86-55 and Grambling 92-37. Those scores became evidence in a college basketball argument that was still running when Kevin and Brad recorded the last CBB Central episode, "Final Pod and Bad Gets," on July 9, 2024.

The short version of how non-conference scheduling affects NET ranking in college basketball: opponent strength, game location and your efficiency in each game all feed the number. Bye games give you no free boost. They help only if you beat those teams by far more than expected. The hosts' preferred target is a Q3 team you can blow out.

How does non-conference scheduling affect NET ranking for a high major?

Since the 2020-21 season, the NET has had two components: the Team Value Index, which rewards beating quality opponents, particularly away from home, and an adjusted net efficiency rating. The efficiency piece matters most for scheduling. The same efficiency against stronger opposition rates higher than against lesser opponents, and road efficiency rates higher than home efficiency.

So a 50-point win over a 340th-ranked team is worth less than a 50-point win over a 140th-ranked team would be. Weak opponents set a low ceiling.

Nobody outside the NCAA can say exactly how low. One explainer notes that the NCAA does not publicly share the exact formula it uses to calculate NET rankings. Any claim about "the optimal bye game" is an educated guess, and the hosts guessed too.

Did the Iowa State bye-game model actually work?

Brad's position, which he said he'd held since the NET first came out, is the line at the top of this piece. The problem, in his telling, is that people read Iowa State's schedule as proof that cupcakes are a cheat code. Kevin thinks that reading is wrong, and he went through the receipts.

Iowa State beat Houston twice, finished second in the Big 12 and won the conference tournament. By Kevin's count it was still only the eighth overall seed. That makes it a No. 2 seed, placed in the East Regional. North Carolina took the fourth No. 1 seed even though it lost the ACC title game. Kevin's view is that the metrics liked the Cyclones more than the committee did, and that they "got screwed."

CBS Sports saw the same tension from the other side. Its read was that Iowa State's top non-conference win came against a non-tournament team, Iowa, and it lost to Virginia Tech and Texas A&M before league play. The same piece said seven of the Big 12's 14 teams had non-conference schedules ranked 300th or lower, which prompted grumbling from other coaches about "rigging."

The hosts then asked who else benefited. Oklahoma played a pile of bye games and was left out on Selection Sunday. Cincinnati "rigged the metrics," in Kevin's words, and was never close to the field. BYU was top 15 in the NET and still a 6 seed, though Kevin pointed out it had won at Allen Fieldhouse. These are the hosts' recollections, not numbers I re-checked.

The hosts also took a swipe at the December narrative. Iowa State went 1-2 in its multi-team event, losing to Virginia Tech and Texas A&M, and the one win was a 68-64 squeaker over VCU. Plenty of people decided the ratings were broken. Kevin's answer: Iowa State was just good. It had a physical front line and excellent defense, and he thinks evaluators overweight offense and recruit rankings.

Is a Q3 game better than a Q4 game for NET ranking?

This is where Kevin's argument gets practical, and where the Dayton example comes in.

In the NCAA's quadrant system, a home game against a team ranked 76-160 is Quadrant 3, as is a neutral game against 101-200 and a road game against 136-240. Dayton's schedule, per Kevin, had Youngstown at 134 and Troy at 132. If those were home games (the episode doesn't say), both sit comfortably inside the Q3 band. Oakland, at 161, just misses it.

“It's not so much playing bad teams, it's playing the teams that you can beat the shit out of the most compared to expectations.”
— Underdog Podcasts, on CBB Central

Dayton beat Youngstown by 8, Troy by 12 and Oakland by 24. Kevin's point is that these were not awful opponents, and the margins were healthy. He added that Dayton lost its other games narrowly and won road and neutral games against good opponents. He named SMU, Cincinnati, LSU and St. John's. None of those was a tournament team, but the games still carried weight.

"Dayton sort of stumbled into this great schedule," he said. He wasn't claiming Dayton's staff had built it with that in mind.

The other host put a number on the idea. Beating the 130th team by 12, he said, is "good enough" for an aspiring tournament team. It might even be better than beating a 260th-ranked team by 18. His example was DePaul, who was around 300. He recalled that the T-Rank spread for Marquette hosting them was about 25 points, which is a lot to ask of a top-10 team.

Does margin of victory count in the NET?

Here the hosts and the NCAA are talking past each other a little.

The original 2018 NET did include scoring margin, with the point differential capped at 10 points in each game. That component is gone. The NET no longer uses winning percentage, adjusted winning percentage or scoring margin. The NCAA now calls the run-up-the-score idea a misconception: teams can't boost NET by running up the score, because margin of victory isn't factored in a way that encourages it.

Yet efficiency is points per possession, and a lopsided game produces a lopsided efficiency number. That is why the hosts keep talking about margin. Brad's summary near the end of the debate was that margin of victory is weighted by the other team's NET ranking and by your own. That is a fair shorthand for how an opponent-adjusted efficiency rating behaves, even if the NCAA would phrase it differently.

One of the hosts also lumped the KPI in with the NET as "reliant on point margin." The NCAA's own description of the NET is narrower than that, so treat the phrasing as casual.

My read is that the evidence supports the hosts on the main point. Both sides agree opponent quality changes what a margin is worth. A big win over a bad team simply buys less.

What should a team schedule instead of Q4 bye games?

The hosts' answer is to find the right kind of bad opponent. Brad suggested scheduling small teams that can't shoot. Kevin got specific. Florida was set to play Southern Illinois, coached by Scott Nagy. Kevin said Nagy's teams are competitive but "awful defensively," and a dynamic Florida offense might put up 112 and win by 20 to 30. If Southern Illinois then finished around 122nd in KenPom, Florida would have a bye game that counts like a Q3 result.

That was a prediction, and I haven't checked whether it played out.

Kevin's other target is the Mountain West's "Goldilocks" schedule. A resume full of wins over San Francisco on a neutral court, UNLV at home and UC Irvine at home looks respectable on a quadrant sheet. You don't need to bury those teams. You need to beat them well enough that the efficiency numbers hold up.

The hosts also found time to look at the Providence schedule. They counted Stonehill, Hampton and Green Bay as bye games, and wondered whether Kim English had picked teams with first-year coaches because he smelled blood. One of those coaches was Doug Gottlieb at Green Bay. His team would face Providence the year after Iowa State had beaten Green Bay 85-44.

Listen to the full episode on CBB Central for the debate in the hosts' own words, plus roughly an hour of bad portal gets.