The SportsWire

Projections

Four projected margins for every game with a team page, side by side: SP+, Elo, Line, Our model.

Updated Tue Oct 13, 7:57 PM CT

This week week 7 · 22 games · 3 split

Friday, Oct 16

Saturday, Oct 17

Last week week 6 · 23 games · closest SP+ 0 · Elo 0 · Our model 0

Friday, Oct 9

  • Florida StateatLouisvillefinal
    Fri Oct 9, 6:00 PM CT / 7:00 PM ET
    Series: Florida State leads 17–5 · last: Florida State 16–6, 2023 · Florida State has won 2 straight
    SP+
    n/a
    Elo ✓
    Louisville by 3.1
    Line
    n/a
    Our model ✓
    Louisville by 1.8
    Actual: Louisville by 24 (Louisville 44, Florida State 20) · Closest: Elo
  • #17IowaatWashingtonfinal
    Fri Oct 9, 8:00 PM CT / 9:00 PM ET
    Series: Iowa leads 4–3 · last: Iowa 40–16, 2024 · Iowa has won 2 straight
    SP+
    n/a
    Elo ✓
    Iowa by 1.0
    Line
    Washington by 1.5
    Our model ✓
    Iowa by 0.7
    Actual: Iowa by 17 (Washington 24, Iowa 41) · Closest: Elo
  • Iowa Stateat#8BYUundersoldfinal
    Fri Oct 9, 9:15 PM CT / 10:15 PM ET
    Series: Iowa State leads 5–1 · last: BYU 41–27, 2025
    SP+
    n/a
    Elo ✓
    BYU by 3.7
    Line
    BYU by 14.5
    Our model ✓
    BYU by 19.1
    Actual: BYU by 14 (BYU 24, Iowa State 10) · Closest: Our model

Saturday, Oct 10

  • Texas A&Mundersoldat#13Missouriflatteredfinal
    Sat Oct 10, 11:00 AM CT / 12:00 PM ET
    Series: Texas A&M leads 11–7 · last: Texas A&M 38–17, 2025 · Texas A&M has won 3 straight
    SP+
    n/a
    Elo ✓
    Missouri by 1.5
    Line
    Missouri by 1.5
    Our model ✓
    Missouri by 4.8
    Actual: Missouri by 21 (Missouri 27, Texas A&M 6) · Closest: Our model
  • #21Wake ForestatNC Statemodels disagreefinalnot in the record
    Sat Oct 10, 11:00 AM CT / 12:00 PM ET
    SP+
    n/a
    Elo ✓
    Wake Forest by 3.6
    Line
    n/a
    Our model ✗
    NC State by 1.1
    Actual: Wake Forest by 24 (NC State 14, Wake Forest 38) · Closest: Elo
  • #5IndianaatNebraskafinal
    Sat Oct 10, 11:00 AM CT / 12:00 PM ET
    Series: Indiana leads 11–9–3 · last: Indiana 56–7, 2024
    SP+
    n/a
    Elo ✓
    Indiana by 12.0
    Line
    Indiana by 8.5
    Our model ✓
    Indiana by 6.5
    Actual: Indiana by 3 (Nebraska 17, Indiana 20) · Closest: Our model
  • UCFat#18Oklahoma Stateflatteredfinal
    Sat Oct 10, 11:00 AM CT / 12:00 PM ET
    Series: UCF leads 2–0 · last: UCF 17–14, 2025 · UCF has won 2 straight
    SP+
    n/a
    Elo ✗
    UCF by 3.8
    Line
    n/a
    Our model ✗
    UCF by 2.6
    Actual: Oklahoma State by 14 (Oklahoma State 42, UCF 28) · Closest: Our model
  • North CarolinaatPittsburghfinal
    Sat Oct 10, 11:00 AM CT / 12:00 PM ET
    SP+
    n/a
    Elo ✗
    Pittsburgh by 12.2
    Line
    n/a
    Our model ✗
    Pittsburgh by 10.1
    Actual: North Carolina by 4 (Pittsburgh 17, North Carolina 21) · Closest: Our model
  • Ball Stateat#25Northwesternfinalnot in the record
    Sat Oct 10, 11:30 AM CT / 12:30 PM ET
    SP+
    n/a
    Elo ✓
    Northwestern by 28.7
    Line
    n/a
    Our model ✓
    Northwestern by 25.7
    Actual: Northwestern by 31 (Northwestern 45, Ball State 14) · Closest: Elo
  • South CarolinaatFloridafinal
    Sat Oct 10, 11:45 AM CT / 12:45 PM ET
    Series: Florida leads 31–10–3 · last: Florida 41–39, 2023 · Florida has won 2 straight
    SP+
    n/a
    Elo ✗
    Florida by 3.3
    Line
    Florida by 14.5
    Our model ✗
    Florida by 9.2
    Actual: South Carolina by 19 (Florida 19, South Carolina 38) · Closest: Elo
  • #20HoustonundersoldatKansas Statemodels disagreefinal
    Sat Oct 10, 2:30 PM CT / 3:30 PM ET
    Series: tied 1–1 · last: Houston 24–19, 2024
    SP+
    n/a
    Elo ✗
    Kansas State by 7.7
    Line
    n/a
    Our model ✓
    Houston by 3.5
    Actual: Houston by 19 (Kansas State 16, Houston 35) · Closest: Our model
  • Stanfordat#2Notre Damefinal
    Sat Oct 10, 2:30 PM CT / 3:30 PM ET
    Series: Notre Dame leads 25–14 · last: Notre Dame 49–20, 2025 · Notre Dame has won 3 straight
    SP+
    n/a
    Elo ✓
    Notre Dame by 36.9
    Line
    n/a
    Our model ✓
    Notre Dame by 37.8
    Actual: Notre Dame by 51 (Notre Dame 58, Stanford 7) · Closest: Our model
  • #7TexasvsOklahomafinal
    Sat Oct 10, 2:30 PM CT / 3:30 PM ET · neutral site
    Series: Texas leads 62–51–5 · last: Texas 23–6, 2025 · Texas has won 2 straight
    SP+
    n/a
    Elo ✗
    Texas by 8.0
    Line
    Texas by 6
    Our model ✗
    Texas by 4.2
    Actual: Oklahoma by 3 (Oklahoma 27, Texas 24) · Closest: Our model
  • #12UCLAundersoldatOregonflatteredfinal
    Sat Oct 10, 2:30 PM CT / 3:30 PM ET
    Series: UCLA leads 40–33 · last: Oregon 34–13, 2024 · Oregon has won 5 straight
    SP+
    n/a
    Elo ✗
    Oregon by 9.9
    Line
    Oregon by 12.5
    Our model ✗
    Oregon by 17.5
    Actual: UCLA by 2 (Oregon 15, UCLA 17) · Closest: Elo
  • #9Ole MissatVanderbiltfinal
    Sat Oct 10, 2:30 PM CT / 3:30 PM ET
    Series: Ole Miss leads 56–37–2 · last: Ole Miss 33–7, 2023 · Ole Miss has won 5 straight
    SP+
    n/a
    Elo ✓
    Ole Miss by 0.5
    Line
    n/a
    Our model ✓
    Ole Miss by 2.9
    Actual: Ole Miss by 35 (Vanderbilt 17, Ole Miss 52) · Closest: Our model
  • #14TennesseeatArkansasfinal
    Sat Oct 10, 3:15 PM CT / 4:15 PM ET
    Series: Tennessee leads 14–7 · last: Tennessee 34–31, 2025
    SP+
    n/a
    Elo ✓
    Tennessee by 10.1
    Line
    n/a
    Our model ✓
    Tennessee by 14
    Actual: Tennessee by 7 (Arkansas 17, Tennessee 24) · Closest: Elo
  • Marylandat#4Ohio Statefinal
    Sat Oct 10, 3:15 PM CT / 4:15 PM ET
    Series: Ohio State leads 9–0 · last: Ohio State 37–17, 2023 · Ohio State has won 9 straight
    SP+
    n/a
    Elo ✓
    Ohio State by 33.2
    Line
    n/a
    Our model ✓
    Ohio State by 23.4
    Actual: Ohio State by 29 (Ohio State 45, Maryland 16) · Closest: Elo
  • #22LSUat#15Kentuckyfinal
    Sat Oct 10, 6:00 PM CT / 7:00 PM ET
    Series: LSU leads 40–17–1 · last: Kentucky 42–21, 2021
    SP+
    n/a
    Elo ✗
    LSU by 4.6
    Line
    LSU by 10
    Our model ✗
    LSU by 10.4
    Actual: Kentucky by 1 (Kentucky 35, LSU 34) · Closest: Elo
  • #1Georgiaat#6Alabamafinal
    Sat Oct 10, 6:30 PM CT / 7:30 PM ET
    Series: Alabama leads 45–26–3 · last: Georgia 28–7, 2025
    SP+
    n/a
    Elo ✗
    Alabama by 0.1
    Line
    Georgia by 2.5
    Our model ✗
    Alabama by 5.9
    Actual: Georgia by 3 (Alabama 17, Georgia 20) · Closest: Elo
  • #19USCundersoldatPenn Stateflatteredmodels disagreefinal
    Sat Oct 10, 6:30 PM CT / 7:30 PM ET
    Series: USC leads 6–5 · last: Penn State 33–30, 2024
    SP+
    n/a
    Elo ✗
    Penn State by 3.9
    Line
    Penn State by 1.5
    Our model ✓
    USC by 3.1
    Actual: USC by 8 (Penn State 23, USC 31) · Closest: Our model
  • SyracuseatVirginiafinal
    Sat Oct 10, 6:30 PM CT / 7:30 PM ET
    Series: tied 3–3 · last: Syracuse 22–20, 2022
    SP+
    n/a
    Elo ✓
    Virginia by 14.3
    Line
    n/a
    Our model ✓
    Virginia by 7.2
    Actual: Virginia by 3 (Virginia 28, Syracuse 25) · Closest: Our model
  • Kansasat#11Utahfinal
    Sat Oct 10, 7:00 PM CT / 8:00 PM ET
    Series: Utah leads 3–2 · last: Utah 31–21, 2025 · Utah has won 3 straight
    SP+
    n/a
    Elo ✓
    Utah by 14.2
    Line
    n/a
    Our model ✓
    Utah by 7.5
    Actual: Utah by 17 (Utah 41, Kansas 24) · Closest: Elo
  • #23Boise StateatFresno Statemodels disagreefinal
    Sat Oct 10, 9:30 PM CT / 10:30 PM ET
    Series: Boise State leads 17–10 · last: Fresno State 30–7, 2025 · Fresno State has won 3 straight
    SP+
    n/a
    Elo ✗
    Fresno State by 1.1
    Line
    n/a
    Our model ✓
    Boise State by 8.2
    Actual: Boise State by 7 (Fresno State 31, Boise State 38) · Closest: Our model

Season weeks 2, 3, 4, 5, 6

Over the 79 games where all three of SP+, Elo and the line existed before kickoff: the average miss in points, and how often SP+ or Elo was closest (a tie credits both). The line is the benchmark: its miss only, never closest.

modelgamesavg missclosest
SP+7910.647 of 79
Elo7912.333 of 79
Line7911.1—

Line only (no SP+ or Elo rating for an FCS opponent): 8 games, line avg miss 14.5; not in the table.

Our model experimental · weekly snapshots for weeks 4, 5, 6, 7 · 63 eligible games scored

vs SP+: 40 games, avg miss our model 11.8 · SP+ 11.4, closest 18 / 22

vs Elo: 61 games, avg miss our model 12.1 · Elo 13.3, closest 35 / 27

vs Line: 50 games, avg miss our model 11.9 · Line 11.7, closest 24 / 26

All four: 40 games, our model closest 14 of 40 (among the three models, ties credit each; the line never counts)

FCS opponents, our model vs the line: 0 games

Our call · 40 games where SP+, Elo and Our model all existed before kickoff, each model on those same games, and the line for comparison (its miss only: a benchmark is never counted as naming the winner):

modelgamesavg missnamed the winner
Our call4011.930 of 40
SP+4011.429 of 40
Elo4013.329 of 40
Line (benchmark)4011.6—
Our model4011.832 of 40

Against its benchmark, the line: Our call was at least as close in 15 of 40, the line in 25 (a tie counts for both).

How to read this
  • SP+ is a rating of how good a team is, in points better or worse than an average team, built from its play-by-play; the gap between two ratings is a projected margin. Elo rises with wins and falls with losses, weighted by the opponent's strength; the gap is divided by 25 for points. The line is the pregame line, the margin the market expected; it is not a model, and is shown for comparison only, as the benchmark. Our model is this site's own experimental rating from opponent-adjusted play-by-play blended with a preseason estimate.
  • In SP+, Elo and our model the home team gets 2.5 points; the line already includes the market's home edge. Each cell names the favored team and the margin; Even is a margin of zero.
  • Each cell is one method's estimate, shown side by side. "Models disagree" means the models shown do not all favor the same team; the line never counts in that. Our call, the site's prediction on the team pages and the big game pages, is the simple average of SP+, Elo and Our model (the line is left out and kept as its benchmark); the season section scores it on the games where all three existed.
  • For a finished game, ✓ marks each model that named the winner and ✗ each that did not (the line gets no mark); Closest: names the model whose margin was nearest the real one among the models shown (never the line), our model included when it had a projection before kickoff; an exact tie goes to the first listed. The closest counts in each section's heading and the season table below keep the accuracy report's scoring of SP+ and Elo, where a tie credits both and our model is credited when it was at least as close as all three. Avg miss in the season table is that model's average distance from the real margin, in points, over the games where every model existed before kickoff.
  • A game counts in the record only if one of its teams had a page before that week's kickoffs; a game shown as not in the record belongs to a team that gained its page afterwards, and is displayed but counted nowhere.
  • SP+ and Elo show n/a until the week's ratings have been captured, which happens midweek once the data source updates them, with one exception: until a game is final, an SP+ cell without the week's own rating shows the newest earlier one, marked week N snapshot, the same number and label as the team page. The season record counts only games where the week's own ratings existed. Our model counts a game only when its weekly snapshot was written before kickoff; "excluded" marks a game that had already kicked off.
  • A small flattered or undersold tag after a team's name is the regression watch from its page: last game's score ran more than 7 points ahead of, or behind, the play. More ›
  • Metrics explained: the four models and their track record ›