2026 March Madness Second Round Picks & Predictions

Sunday, March 22

Before diving into Sunday’s slate, a quick recap of where we stand with the model and our first-weekend totals plays. This has been a strong tournament for the efficiency-based projections: High Point was called to beat Wisconsin straight up in Round 1, and it happened — 83-82. VCU was called to beat North Carolina, and it happened in overtime. On the totals side, our historically reliable first-weekend under plays finished either 2-2 or 3-1 depending on how the Michigan/Saint Louis under at 161.5 resolved Saturday.

There are no designated under plays for Sunday’s games — that closes out the first-weekend totals series at four plays. The focus Sunday is entirely on sides and where the model diverges most sharply from the market.

The Games

Purdue -7.5 vs. Miami (FL) | 12:10 p.m. | CBS | Total: 148.5 Model: Purdue 79.0, Miami 77.3

The market has Purdue laying 7.5 points. The model has Purdue winning by 1.7. That’s a six-point gap entirely in Miami’s favor from a cover standpoint, and it’s the second-strongest spread signal on Sunday’s board. Miami knocked off Missouri in the first round, Malik Reneau put up 24 points, and the Hurricanes bring a physical, experienced ACC roster into this game. Purdue’s defensive efficiency (1.048) has been a persistent vulnerability, and Miami’s front line can exploit it. Purdue wins in the model, but Miami and the 7.5 points is where the value is. Model-projected total of 156.3 vs. 148.5 is a 7.8-point over lean — notable but not a formal play.

Iowa State -4.5 vs. Kentucky | 2:45 p.m. | CBS | Total: 145.5 Model: Iowa State 79.4, Kentucky 69.6

The model’s most decisive result of the day. Iowa State wins by nearly 10 points, comfortably covering the 4.5-point spread. The Cyclones own the best defensive efficiency on Sunday’s slate (0.934), and Kentucky is coming off a draining overtime win against Santa Clara on Friday — playing back-to-back high-effort games with little margin for recovery. Iowa State’s offensive efficiency (1.173) is also legitimate, and the model sees this as a comfortable separation game in the second half. Iowa State covers.

St. John’s -3 vs. Kansas | 5:15 p.m. | CBS | Total: 144 Model: St. John’s 73.3, Kansas 68.7

The model has St. John’s winning by 4.6, which covers the 3-point spread. Rick Pitino’s team has back-to-back Big East regular season and tournament titles and enters with genuine Final Four ambitions. Kansas has the talent — Darryn Peterson is a potential top NBA Draft pick — but the Jayhawks have been inconsistent all season and escaped a scare against Cal Baptist in Round 1. The model trusts the Johnnies here. Model total of 142 is 2 points below the 144 line — a mild under lean but not enough to act on.

Virginia vs. Tennessee -1.5 | 6:10 p.m. | TNT | Total: 137.5 Model: Virginia 75.0, Tennessee 72.4

Virginia wins outright in the model despite being listed as a 1.5-point underdog. The Cavaliers are 29-5 and bring elite defensive efficiency (0.979) into a matchup where Tennessee (1.137 OE) has the lowest offensive efficiency of any team on Sunday’s slate. Tony Bennett’s system is built for exactly this environment — neutral court, slow pace, physical defense, low-margin basketball. Tennessee dominated Miami (Ohio) in Round 1 behind 53% shooting, but the model doesn’t expect that to carry against a completely different defensive identity. Rick Barnes has a well-documented rough history against the spread in the NCAA Tournament. The model takes Virginia. Model total of 147.4 vs. 137.5 shows a large over lean, but that’s a reflection of how low the market has set this number rather than a formal play signal.

Florida -10 vs. Iowa | 7:10 p.m. | TBS | Total: 145.5 Model: Florida 77.2, Iowa 72.4

Florida wins by 4.8 in the model, laying 10. Iowa covers by roughly five points per the model’s projection — the single strongest cover signal of Sunday’s eight games. Iowa held opponents to 66 points per game this season, shoots 49.1% from the field, and is built to control pace and keep games in the low-70s range regardless of opponent. The defending champion Gators have the interior trio of Haugh, Condon, and Chinyelu to impose their will, but Florida at -10 requires a type of blowout the model simply doesn’t project here. Iowa and the 10 points is the play.

Arizona -11 vs. Utah State | 7:50 p.m. | truTV | Total: 156.5 Model: Arizona 79.5, Utah State 73.3

Arizona wins by 6.2 in the model, but the market has them laying 11. Utah State covers according to the projection. The Aggies are 28-6, knocked off Villanova as a 9-seed, and their offensive efficiency (1.162) is legitimate against anyone. Arizona is the most complete team in the tournament with Burries and Bradley as the tournament’s best backcourt, and they advance — but the model doesn’t see the double-digit blowout the spread requires. Utah State and the points.

UConn -4 vs. UCLA | 8:45 p.m. | TNT | Total: 137.5 Model: UConn 75.4, UCLA 69.3

UConn wins by 6.1 and covers the 4-point spread. The Huskies’ edge over UCLA is almost entirely defensive (0.948 DE), and both teams carry nearly identical offensive efficiency marks. Tarris Reed Jr. was a force against Furman with 31 points and 27 rebounds, and Dan Hurley’s teams typically elevate from round to round. UCLA is a quality opponent — 38.2% from three makes them dangerous — but the model sees UConn controlling pace and pulling away. The model total of 144.7 vs. 137.5 is a 7.2-point over lean, which is notable even if not a formal play. UConn covers.

Alabama -1 vs. Texas Tech | 9:45 p.m. | TBS | Total: 165 Model: Texas Tech 84.7, Alabama 84.4

The night’s final game is the one the model refuses to decide. Texas Tech wins by 0.3 points — genuinely indistinguishable from a coin flip. What makes this more interesting is that the market now has Alabama as a 1-point favorite, while the model barely leans the other way. Alabama’s defensive efficiency (1.081) is a concern in a tournament setting, and Texas Tech has been playing inspired basketball since Christian Anderson stepped into J.T. Toppin’s role. But with a 0.3-point projection, there’s no conviction in either direction. Both offenses are in the 1.15+ range, both teams push pace, and the model total of 169.1 is 4.1 points above the 165 line — a mild over lean. Watch the line movement on this one and trust your read. The model has nothing to offer beyond “expect a game.”

Sunday Summary

Strongest model-vs.-market gaps: Iowa covers Florida (-10), Miami covers Purdue (-7.5), Utah State covers Arizona (-11), Virginia beats Tennessee outright.

Model advances to the Sweet 16: Iowa State, St. John’s, Virginia, Florida (barely), Arizona, UConn, Texas Tech (barely), and the toss-up between Purdue and Miami.

Enjoy the final day of the first weekend. The Sweet 16 preview is coming.