• Oops!
    Something went wrong.
    Please try again later.

How NCAA rankings have tiny Colgate ahead of a handful of powerhouses

Jeff Eisenberg
·6 min read
  • Oops!
    Something went wrong.
    Please try again later.

The top seven teams in the NCAA’s NET rankings include unbeaten Gonzaga, one-loss Baylor and a trio of Big Ten heavyweights with national title aspirations.

Then there’s No. 8: A small-conference afterthought that has never won an NCAA tournament game in program history, nor even once appeared in the AP Top 25.

How is Colgate — the No. 2 seed in this week’s Patriot League tournament — ranked ahead of 2019 national champ Virginia, blue blood Kansas and powerhouses Ohio State and Villanova? Even the coach of the Raiders didn’t see this coming.

“It’s a mathematical outlier,” Matt Langel told Yahoo Sports. “It’s an interesting tidbit in what has been the craziest college basketball season in my lifetime.”

The explanation for Colgate’s top-10 NET ranking begins with the COVID-19 pandemic’s impact on the Patriot League’s schedule. The league did not allow six of its eight teams to start their seasons until Jan. 2. Only Army and Navy were exempt and had permission to schedule games in November and December.

As a result of that decision, the totality of the Patriot League’s non-league schedule this season consists of nine games involving either Army or Navy. Army went 4-1 including a narrow loss to Florida and an impressive win over mid-major power Buffalo. Navy went 3-1 including an upset of Georgetown.

In a normal year, the NET rankings — a metric created to help select and seed NCAA tournament hopefuls — would evaluate the strength of the Patriot League on the results of roughly 100 non-conference games. In this pandemic year, only nine data points are available. Army and Navy outperforming expectations in those nine games is dramatically warping the NET’s perception of the league as a whole.

“Basically, Colgate and the Patriot League are a closed system,” college basketball analytics expert Bart Torvik told Yahoo Sports. “You can't really use results from just this year with no priors of any kind and compare those teams against the rest of D-I because there’s no connectedness.”

(Michael Wagstaffe/Yahoo Sports illustration)
(Michael Wagstaffe/Yahoo Sports illustration)

Colgate has soared higher in the NET rankings than its Patriot League peers because of its 12-1 record and impressive margin of victory. Because the Patriot League broke its teams into three divisions to help limit travel, Colgate has only played three opponents: Army, Boston University and Holy Cross. The Raiders’ lone loss came by two against Army. They won the rest of their games by an average of nearly 19 points apiece.

“When you beat your opponents by such a huge margin, that shows up,” said data scientist Ed Feng, founder of The Power Rank.

While it’s theoretically possible that the Patriot League could produce a team worthy of a high ranking, recent history suggests it’s unlikely. Only twice in the 2000s has the Patriot League’s NCAA tournament entrant received better than a No. 13 seed. The league has not produced an NCAA tournament victory since 15th-seeded Lehigh toppled mighty Duke in 2012.

Other respected college basketball analytics experts have avoided overrating the Patriot League this season by incorporating data from previous years into their rankings. The assumption that the league would have performed at its recent average level in non-conference play makes a huge difference.

Colgate is 80th in Torvik’s tempo-free rankings. Ken Pomeroy has the Raiders 89th.

“The preseason weighting helps anchor the Patriot League in a more realistic way,” Pomeroy told Yahoo Sports. “The weighting only applies to cases where teams have played few games, so it influences the Patriot League while not being a factor for the vast majority of teams at this point.”

While the NCAA’s algorithm does not include data from previous seasons, Colgate’s top-10 NET ranking isn’t likely to dramatically alter the selection committee’s perception of the Raiders. NCAA spokesman David Worlock cautioned that the NET is “just one tool” and that “other metrics and criteria factor into the committee’s decision for selecting and seeding teams.”

Many mock NCAA tournament brackets project Colgate as a No. 13 or 14 seed if it wins this week’s Patriot League tournament. The Raiders received a No. 15 seed two years ago and lost a first-round matchup to Tennessee.

“I don’t think we’re going to get a seed that is similar to where our NET ranking is or even close,” Langel said, “but my hope is that maybe we can bump up a spot or two.”

The most tangible impact of Colgate’s high NET ranking is that Langel’s phone has rung more often than usual the past few weeks. Coaches of teams on the NCAA tournament bubble keep calling to try to hastily arrange a game.

Stoppages and outbreaks caused an unusual number of teams to scramble to fill holes in their schedules. What better way to do that than by scheduling the Raiders in hopes of earning a résumé-boosting Quadrant 1 victory without having to beat a traditional power?

“I fielded a number of calls from bubble teams as the season was winding down,” Langel said. “My angle was to try to get them to come here. I think a testament to our team is that nobody really had a lot of interest in doing that. They wanted to play us, but they wanted to play us in their venue.”

Colgate’s top-10 NET ranking has also garnered attention from coaches eager to assess how the Raiders did it. They all want to know if this is strictly a COVID anomaly or if there’s a lesson to be learned in how to game the NET in the future.

Among those qualified to answer that question is Matt Dover, co-owner of an analytics company that helps coaches optimize their non-league schedules to maximize their chances of making the NCAA tournament. It’s Dover’s belief that this is simply a “weird artifact of a weird season” — not a loophole to be exploited again and again.

“The NET is built to be pretty hard, if not impossible to game, under normal circumstances,” Dover told Yahoo Sports. “This is just a unique situation caused by the Patriot League playing so few non-conference games. That is leaving the entire league vulnerable to huge statistical swings that there won’t be in future years.”

Could another one-bid league gamble, only play a handful of strategically selected non-league games and hope to outperform expectations the way Army and Navy did this season? In theory, maybe. In reality, too many small-conference teams rely on the revenue generated by playing road non-conference games in return for tens of thousands of dollars.

“What are they going to do?” Dover said. “Totally forgo the option of playing those? Because that’s what you’d have to do. You’d have to limit it to a tiny, tiny sample size. Once you get up to a slightly larger sample size — every team playing a few non-conference games for example — it would be much less likely you’d get something super skewed.”

More from Yahoo Sports: