Ball State at Kent State. No call on the spread and no call on the total.

Spread.The strongest explanation for Kent State -3.5 is home field plus quarterback continuity against a Ball State offense struggling to sustain drives. That supports a modest favorite, not necessarily a comfortable victory. Ball State's September 21 press conference identified 11 penalties in consecutive games and specifically highlighted Dru DeShields' ability to extend plays. Our -9.8 is 6.3 points beyond consensus, but I substantially discount that separation: Kent trailed Wofford entering the fourth quarter before its 22-point finish, something a final-margin model cannot distinguish from steady dominance. There is no uniform directional signal. September 23 FanDuel reporting corroborates -3.5. I lean Kent because of the quarterback and execution matchup, but those advantages appear substantially priced. My limit is -3 at -110; pass today. Confidence is 2/5 ordinal, not a probability.

Total.The market around 51 sensibly balances weak offenses against vulnerable defenses. The under case is specific: Ball State managed only 93 passing yards at Liberty, while Kent scored just 14 through three quarters against Wofford. Those performances question how reliably either offense can finish drives. The over case is substantial: Ball State surrendered 460 yards at Liberty, both teams now face easier defenses, and the kickoff forecast offers no meaningful weather suppression. Our 55.13 projection favors over, but its historical scoring inputs cannot resolve this matchup's offensive uncertainty. Captured totals moved in opposite directions across books, so I see no coherent market signal. I lean under only at 54.5 or higher, -110 or better, to allow for defensive breakdowns. Pass at the available 50.5 to 51.5. Confidence is 2/5 ordinal.

Where our number looks wrong.The -9.8 projection overstates the demonstrated separation. Its prior-driven ratings and final-margin update give insufficient weight to Kent's difficulty separating from Wofford before the fourth quarter. The flat home-field allowance and calibration do not establish an additional 6.3-point betting edge. Ball State's transfer quarterback also makes historical roster-based assumptions fragile.

• The displayed injury entries referred to September 19 games rather than confirmed September 26 designations.
• DraftKings displayed Ball State with 80% of moneyline handle and 50% of bets; sample size was unavailable, so this is not evidence of informed spread action.
• Ball State finished with 261 total yards, including 93 passing, while Liberty accumulated 460 yards.
• Kent State's quarter-by-quarter scoring against Wofford was 7, 7, 0 and 22.
• The matchup's noon ET forecast showed 65°F, partly cloudy conditions, 0% precipitation and NNE wind at 7 mph.
• The National Weather Service forecast sunny Saturday conditions with a high near 72°F in Kent.

Spread.The market's case for a short number: both teams are bad, Ball State won last year's meeting 17-13, Ball State's Keldric Luster (Texas State transfer) was efficient against Stony Brook, and Kent State has been shut out by South Carolina and held to 3 by Ohio State. The case for Kent State: it is the more stable program, returning about 64% of production under Mark Carney after a 5-7 season (4-4 MAC) that beat preseason expectations, with Dru DeShields accounting for 322 yards and three touchdowns against Wofford, while Ball State returns about 6% of its production and lost 51-15 at Liberty. Both schedules were brutal, so the FBS results tell us little. I lean Kent State at home at -3.5, but the model gap is the main input and that model has no ATS edge; this is a pass.

Total.The total is 50.5 to 51.5 and has barely moved; our model is 55.1, leaning over by 4.1. The over case: both defenses allowed 51 or more to better teams, and both offenses scored 36-plus against FCS opponents. The under case: last year's meeting ended 17-13, Kent State has scored 0 and 3 against FBS opponents and Ball State 3 and 15; the MAC opener at noon with mild wind should not add much. I lean slightly under because neither offense has shown it can score against FBS competition, but the model's disagreement and both defenses' weakness keep this a pass.

Where our number looks wrong.The model's -9.8 leans on Kent State's and Ball State's prior ratings and blowout losses to elite teams; it cannot tell how much of Ball State's 6% returning production has been replaced by transfers like Luster, and against Ohio State both teams produced near-identical results (3-56, 3-59).

• Ball State 17, Kent State 13 in 2025
• Kent State lost 0-57 and 3-59 to FBS opponents

Spread.Both teams are 1-2 and were blown out by the same Ohio State team (Kent State 3-59, Ball State 3-56), plus Ball State lost 15-51 to Liberty. This is a bottom-of-the-MAC matchup with thin coverage and no clear informational edge beyond the small home-field bump priced into Kent State -3.5. A published win-probability model projects a final margin of roughly 3 points, which lines up almost exactly with the market number, reinforcing that there isn't obvious mispricing here. I don't have a real lean on either side at this number.

Total.Both defenses have real, repeatable weaknesses that aren't purely a byproduct of playing Ohio State: Kent State allowed 15 to an FCS Wofford team and Ball State allowed 17 to a non-Power-conference Stony Brook squad. When two offensively limited teams with genuinely shaky defenses play each other, scoring tends to normalize higher than either team's season average against elite competition would suggest, since neither defense is likely to consistently stop the other.

• Kent State beat Wofford 36-15 and Ball State beat Stony Brook 41-17, both allowing double-digit points to lesser competition