Offseason — portal net flow

The 2026 portal cycle, in ledger form.

Talent-weighted transfer-portal net flow for the 2026 recruiting cycle, standardized (mean-centered, scaled by standard deviation) across the 138 FBS teams that appeared in the portal. A positive value means a program added more talent-weighted value than it lost; a negative value means the reverse. The signal is applied as a small additive modifier (≈ 0.20 points of projection per standardized unit) on top of Bill Connelly's preseason SP+ — the only residual signal in the stack that is not already priced into his August number. The distribution is right-skewed by a handful of record classes and is not Gaussian; the sign and rank order are reliable, the exact standardized value is a scale marker rather than a probability statement.

↑ Top gains

by standardized net flow
+5.9
LSU Elite SEC
5★ QB Sam Leavitt (from Arizona State)
applied
+1.18
83.9
+2.9
Texas Tech Elite Big 12
5★ QB Brendan Sorsby (from Cincinnati)
applied
+0.58
87.5
+2.7
UCLA Solid Big Ten
4★ WR Aidan Mizell (from Florida)
applied
+0.54
56.7
+2.6
Purdue Building Big Ten
4★ EDGE Elo Modozie (from Georgia)
applied
+0.53
48.3
+2.6
Kennesaw State Rebuilding Conference USA
4★ QB Rickie Collins (from Syracuse)
applied
+0.52
34.0
+2.3
James Madison Building Sun Belt
3★ QB Arrington Maiden (from Memphis)
applied
+0.47
46.4
+2.2
Kentucky Solid SEC
4★ QB Kenny Minchey (from Notre Dame)
applied
+0.45
59.1
+2.2
Missouri State Rebuilding Conference USA
3★ QB Skyler Locklear (from UTEP)
applied
+0.45
19.7
+2.2
Ole Miss Contender SEC
4★ QB Deuce Knight (from Auburn)
applied
+0.43
74.2
+2.1
Memphis Building American Athletic
4★ WR Alante Brown (from Michigan State)
applied
+0.43
48.9
+2.1
UConn Rebuilding FBS Independents
3★ QB Jake Merklinger (from Tennessee)
applied
+0.42
33.1
+2.1
Virginia Solid ACC
4★ QB Beau Pribula (from Missouri)
applied
+0.42
61.7
+2.1
Indiana Elite Big Ten
4★ QB Josh Hoover (from TCU)
applied
+0.41
87.3
+2.0
Wisconsin Solid Big Ten
3★ QB Deuce Adams (from Louisville)
applied
+0.40
56.2
+1.9
Ball State Rebuilding Mid-American
3★ QB Tyler Mizzell (from Concord University)
applied
+0.39
13.0

↓ Top losses

by standardized net flow
−4.2
Oregon Elite Big Ten
4★ QB Bryson Beaver (to Georgia)
applied
−0.84
95.6
−3.7
Florida State Solid ACC
4★ WR Lawayne McCoy (to Louisville)
applied
−0.74
66.7
−3.3
Michigan State Building Big Ten
4★ QB Aidan Chiles (to Northwestern)
applied
−0.66
51.3
−3.2
Iowa State Building Big 12
4★ QB Rocco Becht (to Penn State)
applied
−0.63
53.1
−3.0
Oregon State Rebuilding Pac-12
4★ QB Gabarri Johnson (uncommitted)
applied
−0.59
40.2
−2.8
Florida Contender SEC
4★ QB DJ Lagway (to Baylor)
applied
−0.56
72.1
−2.6
North Carolina Solid ACC
4★ LB Khmori House (to Arkansas)
applied
−0.53
57.8
−2.6
Penn State Contender Big Ten
5★ EDGE Chaz Coleman (to Tennessee)
applied
−0.51
74.0
−2.5
Northern Illinois Rebuilding Mountain West
4★ S James Finley (to Illinois)
applied
−0.50
22.8
−2.3
Ohio State Elite Big Ten
4★ WR Mylan Graham (to Notre Dame)
applied
−0.45
99.0
−2.2
Auburn Solid SEC
5★ WR Cam Coleman (to Texas)
applied
−0.45
67.0
−2.1
South Alabama Rebuilding Sun Belt
3★ QB Bubba Thompson (to West Florida)
applied
−0.43
29.5
−2.1
USC Contender Big Ten
4★ QB Husan Longstreet (to LSU)
applied
−0.42
78.5
−2.0
Alabama Contender SEC
4★ EDGE Noah Carter (to Georgia Tech)
applied
−0.41
79.2
−2.0
Syracuse Building ACC
4★ QB Rickie Collins (to Kennesaw State)
applied
−0.41
51.2

Star chips reflect portal rating, not the player's high-school recruiting composite. A player who arrived as a 3★ high-school prospect can appear as a 5★ portal transfer after producing at the college level (Brendan Sorsby is the canonical case). The two ratings measure different things and are not expected to agree.

Portal weights: 5★ = 5.0, 4★ = 3.0, 3★ = 1.0, 2★ / unrated = 0.3. QB and edge positions weighted higher (1.5x / 1.3x). Immediate-eligibility transfers weighted 1.0; redshirt 0.5. Withdrawn entrants (players who returned to their original team) are excluded. Uncommitted portal entrants (destination not yet reported by CFBD) are shown as “uncommitted” rather than assigned to a school. See methodology for the full residual-signal story.