Preseason Week 3 by the Numbers: The Overs Finally Won a Week, and Completed a Perfect Lesson in How Markets Chase

The 2026 preseason closed its books with the overs' first winning week: 9–7 against closing totals, no pushes. The sixteen finales averaged 39.7 points against an average line of just 36.2, the lowest totals of the summer, and the market's biggest miss since opening weekend. Put the three weeks side by side and you get something better than a betting recap: a complete, self-contained lesson in how lines chase results.

Every game, every number

The overs: Cardinals–Packers was the game of the summer, 80 combined points (42–38 Green Bay) against a 36.5 total, over by 43.5 and the highest-scoring game of the entire preseason. Steelers–Bills cleared 35.5 at a walk (28–27 Buffalo, 55 points). Browns–Patriots went 50 against 35.5 on Cleveland's 37–13 rout. Ravens–Commanders (41–3, total 44 vs. 33.5), Bengals–Eagles (30–13, 43 vs. 36.5), Saints–Cowboys (27–24, 51 vs. 37.5), Vikings–Broncos (34–6, 40 vs. 37.5), Lions–Colts (25–16, 41 vs. 36.5), and Bears–Titans (24–15, 39 vs. 37.5) rounded out the nine.

The unders: Bucs–Jaguars produced 19 points against 34.5, Tampa Bay was shut out 19–0, and Seahawks–Chiefs matched it, an 18-point miss on a 33.5 total that ended in the week's lone deadlock, a 9–9 tie. 49ers–Raiders (30 vs. 37.5), Rams–Chargers (38 vs. 38.5, a half-point sweat), Texans–Panthers (29 vs. 35.5), Falcons–Dolphins (29 vs. 36.5), and Giants–Jets (29 vs. 36.5) filled out the seven.

The three-week arc, in one table

Week 1: 12–4 under, average total 38.25, average scored 33.9. Week 2: 8–8, total 37.4, scored 37.2. Week 3: 9–7 over, total 36.2, scored 39.7. Read it as a story: the books opened high, August's unders embarrassed them, and they cut totals by two full points over three weeks, exactly as actual scoring rose nearly six points a game, because starters play more and offenses sharpen as September approaches. The market chased last week's result down while reality climbed the other way, and the two lines crossed somewhere around cutdown weekend.

Home teams were no refuge either

Home teams went 7–8–1 straight up and 7–9 against the spread in Week 3. If you carried a 'take the home team in the preseason finale, they want to send the fans home happy' theory into last weekend, it paid the vig and nothing else. August home-field advantage is a myth for the same reason August totals wobble: nobody is game-planning, and the fourth quarter belongs to players fighting for roster spots, not the scoreboard.

What this means now that the games count

The preseason's parting gift is a rule worth pinning in the league chat: markets overcorrect toward whatever just happened, and the correction is usually late. It'll happen all season, a team covers three straight, its line inflates, and the value quietly switches sides. When your league debates its Week 1 pick this week, the question isn't 'what happened last week?' (the market already priced that); it's 'what does the market still have wrong?' That's a harder question. It's also the only one that pays.

Play responsibly. If gambling stops being fun, call or text 1-800-GAMBLER.