Sorting Signal From Noise as Conference Play Begins

One of the hardest parts of evaluating college basketball teams in January is separating signal from noise.

Some teams are benefiting from schedule strength. Others are being rewarded by efficiency models despite limited tests. And a few are simply so volatile that they defy reliable projection altogether.

Using data from TeamRankings.com, this week’s Data Dive focuses on three categories of teams worth scrutinizing as we move deeper into conference play:

• Teams that appear overvalued relative to their underlying performance
• Teams whose results remain unusually inconsistent
• Teams with strong profiles built on particularly soft non-conference schedules

Overvalued Relative to the Market

Poll rankings are not designed to be predictive. They are a snapshot of perception, shaped by timing, visibility, and recent results.

When we compare AP Poll rankings to TeamRankings predictive power ratings, several teams show meaningful gaps.

Below are the five most overvalued teams in this week’s AP Top 25 based on where TeamRankings currently rates them:

• Nebraska: AP #10 / TeamRankings #29
• Texas Tech: AP #14 / TeamRankings #23
• North Carolina: AP #17 / TeamRankings #25
• SMU: AP #24 / TeamRankings #31
• UCF: AP #25 / TeamRankings #50

Bonus: Villanova is ranked 24th in the Coaches Poll, while TeamRankings currently has them 33rd.

This does not mean these teams are “bad” or undeserving of attention. It means their current market perception is stronger than their predictive profile.

North Carolina is a clear example. Their results have exceeded their underlying efficiency metrics for much of the season, and TeamRankings has consistently rated them closer to the mid-20s than the top tier. That gap does not guarantee regression, but it does suggest a thinner margin for error than the ranking implies.

Nebraska deserves special context. Over the past month, the Cornhuskers have posted legitimate wins over Wisconsin, Illinois, Michigan State, Ohio State, and others. While their predictive rating still lags their ranking, the quality of recent results suggests the model may be slow to fully catch up. Some regression would not be surprising, but the résumé is no longer easily dismissed.

Across this group, the common theme is not fraud or failure. It is that expectations may be running ahead of sustainable performance.

Inconsistent and Difficult to Project

Some teams are not overvalued or undervalued. They are simply unstable.

TeamRankings includes a consistency metric that measures how reliably a team’s performance predicts future results. Low rankings here indicate wide swings from game to game.

The most confusing teams by this measure include:

• TCU
• Boise State
• Florida State
• Wake Forest
• Troy

Bonus: Gonzaga and Delaware also rate extremely low in predictability, though for different reasons.

Delaware’s volatility is unsurprising given extensive roster turnover. Gonzaga’s case is more nuanced. Within a four-game span, the Zags beat Alabama, Maryland, and Kentucky by massive margins, then lost to Michigan by 40. The extremes are real, but context matters, including travel and compressed scheduling.

Michigan presents a different kind of puzzle. They have been dominant at home and on neutral courts, with TeamRankings’ top home and neutral-site ratings nationally. On the road, however, their rating drops sharply, placing them outside the top 20.

That split matters. Michigan’s résumé includes historically impressive stretches, but also several narrow escapes and uneven road performances. The takeaway is not that Michigan is weak, but that their profile is more human and more environment-dependent than raw efficiency numbers suggest.

Troy’s season has been defined by chaos. Eight overtime games, four one-point decisions, and results that swing wildly week to week. Florida State and Wake Forest show similar traits, pairing near-upsets of elite teams with puzzling losses elsewhere. Boise State remains difficult to evaluate after an early loss to a Division II opponent, followed by multiple quality wins.

The common thread across all of these teams is volatility. When variance is this high, projections carry more uncertainty, and matchup context becomes especially important.

Schedule-Inflated Profiles

Finally, several teams are being rewarded by efficiency metrics despite playing particularly soft non-conference schedules.

Teams with strong records and bottom-tier non-conference strength of schedule include:

• High Point (15–3) – 355th NCSOS
• George Mason (15–1) – 347th
• Iowa (12–3) – 342nd
• Miami (13–2) – 341st
• Cal (13–2) – 338th
• Texas A&M (12–3) – 337th
• Buffalo (13–2) – 336th
• Georgia (13–2) – 335th
• Saint Louis (14–1) – 328th

These profiles are not invalid. They are incomplete.

Without stronger competition, it becomes difficult to separate true quality from favorable circumstances. As conference play progresses, these teams will either validate their metrics or see them normalize.

Several notable teams sit just above this range and are worth monitoring closely, including Seton Hall, Nebraska, Ohio State, Virginia, LSU, Indiana, Oklahoma State, and Iowa State.

Closing Thought

January is where models begin to matter more than narratives, but uncertainty is still high. Teams with inflated expectations, volatile profiles, or untested résumés often reveal themselves over the next four to six weeks.

We will have some Bracketology thoughts on Monday, with another Data Dive coming next Thursday.

The Best Way to Follow College Hoops

College basketball is noisy. Even if you watch every game, perception and media narratives rarely line up with what the numbers are showing.