Agent 1: The Statistician

Agent 1: The Statistician

AGENT 1: THE STATISTICIAN

Advanced Metrics & Efficiency Analysis — 2026 NCAA Tournament

Data sources: KenPom, Barttorvik, ESPN BPI, NCAA NET Rankings Updated through conference tournaments, March 17, 2026


Executive Summary

The 2026 tournament field features an unusually clear top tier. Only 9 teams in the entire field rank in the top 25 of BOTH adjusted offensive and defensive efficiency — and historically, 23 of the last 24 national champions have come from the top 21 in offense and top 31 in defense. This year’s championship contenders are statistically identifiable.

The field also has a significant number of mispriced teams — squads whose KenPom ranking disagrees sharply with their seed. Vanderbilt (KenPom #11, seeded 5th) is the most underseeded team in the tournament. Georgia (KenPom #32, seeded 8th) is among the most overseeded.


Stat-Based Power Rankings (Top 25)

These rankings are based on KenPom adjusted efficiency margin, weighted alongside offensive/defensive balance, strength of schedule, and Q1/Q2 records. Teams are ranked by analytical strength, NOT seed.

Power RankTeamSeedKenPomOff. RankDef. RankKey Stat Outlier
1Duke1 (East)#142+40 net rating, among highest all-time. Elite two-way balance.
2Michigan1 (MW)#281#1 defense in the nation. Opponents shoot 30.2% from 3, 44.3% from 2.
3Arizona1 (West)#353Top-5 on both sides. Won Big 12 tournament. Net rating above +35.
4Florida1 (South)#496Defending champ. Top-10 both sides. Dropped slightly after SEC semi loss.
5Houston2 (South)#5145Defensive powerhouse. Returns core of last year’s Final Four team.
6Iowa State2 (MW)#6214#4 defense drives ranking. Offense just outside top 20 — ceiling concern.
7Illinois3 (South)#7128Best offense in America (131.2 adj. points/100 poss). 5 players in double figures. Defense is the variable.
8Purdue2 (West)#8236#2 offense. Won Big Ten tournament. But #36 defense is a red flag for title contention.
9Michigan State3 (East)#92413Balanced profile. Tom Izzo factor. Comfortably in top 25 both sides.
10Gonzaga3 (West)#10299Strong defense (#9) carries them. BUT Braden Huff OUT (knee) — was 17.8 PPG. Significantly weaker without him.
11Vanderbilt5 (South)#11729MOST UNDERSEEDED TEAM. KenPom #11 as a 5-seed. Beat Florida by 17 in SEC semis. Tyler Tanner (19.1 PPG) is an elite guard.
12UConn2 (East)#123011Defense-first (#11). Back-to-back titles in 2023-24. Dan Hurley’s tournament DNA is elite.
13Virginia3 (MW)#132716Trademark defense (#16). Ryan Odom’s first year has been impressive. Methodical offense.
14Nebraska4 (South)#14557Defensive surprise of the year (#7 defense). But #55 offense is a major ceiling limiter. Can they score enough to win 4+ games?
15Tennessee6 (MW)#153715All defense, no offense. #173 in eFG%. Classic Rick Barnes tournament disappointment profile.
16St. John’s5 (East)#164412Big East champions under Pitino. Elite defense (#12) but offense (#44) struggles to score in the half court. Lost R2 last year scoring 66 points.
17Alabama4 (MW)#17367#3 offense (91.7 PPG, 55.4% eFG) but #67 defense is a massive problem. Give up 83.5 PPG. Live-by-the-sword, die-by-the-sword.
18Arkansas4 (West)#18652#6 offense, conference tournament champions. But #52 defense limits ceiling. Similar one-dimensional profile to Alabama.
19Louisville6 (East)#192025Solid but not elite on either end. Brown injury (back) is the wild card — he’s 18.2 PPG when healthy.
20Texas Tech5 (MW)#201233#12 offense still potent, but JT Toppin (ACL) was their best player. 3 straight losses since. This ranking OVERSTATES their current strength.
21Kansas4 (East)#215710Elite defense (#10) but #57 offense is a real problem. Peterson saga has killed offensive consistency. Scored 47 in Big 12 tourney loss.
22Wisconsin5 (West)#221151#11 offense is quietly excellent. But #51 defense is a vulnerability.
23BYU6 (West)#231057#10 offense with Dybantsa. BUT Saunders OUT (knee) — they’re 2-4 since. Current form way worse than season-long stats suggest.
24Saint Mary’s7 (South)#244319#18 defense is excellent. Historically struggles in the tournament. 300-1 title odds seems steep but WCC teams rarely go deep.
25Iowa9 (South)#253131Perfectly balanced (#31/#31). Solid in every area, elite in none. Classic dangerous 9-seed.

Championship Contender Filter

Historically, the national champion has ALWAYS (23 of 24 years in the KenPom era) met both of these thresholds:

  • Top 40 adjusted offensive efficiency
  • Top 25 adjusted defensive efficiency

Teams that pass BOTH filters (2026):

TeamSeedOff. RankDef. RankTitle Odds
Duke142+325
Michigan181+350
Arizona153+380
Florida196+700
Houston2145+1000
Iowa State2214+2200
Michigan State32413+2500
Gonzaga3299+3500
UConn23011+1700

Notable teams that FAIL the filter:

  • Illinois (#1 offense but #28 defense — just outside the defensive threshold)
  • Purdue (#2 offense but #36 defense — clear defensive hole)
  • Vanderbilt (#7 offense but #29 defense — barely misses on defense)
  • Alabama (#3 offense but #67 defense — disqualified)
  • Nebraska (#55 offense but #7 defense — disqualified on offense)
  • St. John’s (#44 offense — borderline miss)

Statistical recommendation: Your champion should come from the 9 teams that pass both filters. Michigan, Duke, and Arizona are the clear top 3 analytically.


Statistical Outliers & Red Flags

Teams Where Stats Disagree With Seed (Underseeded)

  • Vanderbilt (KenPom #11, Seed #5): 6 seed lines too low. Their efficiency profile looks like a 2-seed.
  • Iowa State (KenPom #6, Seed #2): Strong profile but seeded appropriately based on resume.
  • Illinois (KenPom #7, Seed #3): The #1 offense makes them dangerous. 3-seed could be a gift.

Teams Where Stats Disagree With Seed (Overseeded)

  • BYU (KenPom #23, Seed #6): Season-long stats mask the Saunders injury. Current form is more like an 8-9 seed.
  • North Carolina (KenPom #29, Seed #6): Without Wilson, their true level is closer to a 9-10 seed.
  • Texas Tech (KenPom #20, Seed #5): Without Toppin, their current form is closer to an 8-9 seed.
  • Georgia (KenPom #32, Seed #8): #80 defense is the worst of any team seeded 8th or higher.

Fascinating Statistical Profiles

  • Alabama: #3 offense, #67 defense. They score 91.7 PPG but give up 83.5. Every Alabama game is a shootout. They allow 30.4% offensive rebounding — opponents get constant second chances.
  • Nebraska: #7 defense, #55 offense. They hold opponents to 30% from 3 but can they score enough to beat elite teams? Their offense is the lowest-ranked of any 4-seed.
  • Northern Iowa: #24 defense, #153 offense. Their defense is legitimately elite for a 12-seed, but their offense is among the worst in the tournament field. Classic low-scoring upset profile.
  • Cornell (not in field): #47 offense, #343 defense. Hilarious split — would score with anyone but couldn’t stop a nosebleed.

Conference Strength Analysis

ConferenceTeams InAvg KenPomBest TeamConcern
SEC10~18.3Florida (#4)Depth is real but Alabama’s defense and Tennessee’s offense are liabilities
Big Ten9~15.8Michigan (#2)Strongest top-to-bottom. Michigan, Purdue, Michigan State, Iowa, Ohio State, Wisconsin all legitimate
ACC8~21.4Duke (#1)Top-heavy. Duke and UConn elite; Louisville, UNC, Clemson all dealing with injuries
Big 128~19.1Arizona (#3)Arizona and Iowa State are elite. Kansas has offensive issues. Texas Tech gutted by Toppin injury
Big East3~16.7UConn (#12)Only 3 bids but quality. St. John’s and Villanova are both tournament-tested programs
WCC3~14.3Gonzaga (#10)Gonzaga is the engine. Saint Mary’s and Santa Clara are undervalued mid-majors

Recommendation to Bracket Architect

Lock as champion contenders: Michigan, Duke, Arizona (all pass every statistical filter)

Strong value picks: Vanderbilt (massively underseeded), Houston (elite defense + home court), Iowa State (#4 defense)

Avoid as champion: Alabama (defense disqualifies them), Purdue (defense too weak), Illinois (borderline defense), Nebraska (offense too limited)

Statistical upset triggers: Any team with a key player out whose seed was earned with that player playing. The stats say Texas Tech, UNC, BYU, and Kentucky are all overseeded based on their CURRENT rosters.


Report prepared by Agent 1: The Statistician Methodology: KenPom adjusted efficiency, historical championship benchmarks, efficiency margin analysis