South Alabama at Tulane. Lean: South Alabama +9.5. Lean: Under 49.5.

Spread.South Alabama +9.5 is my lean. EL makes Tulane -4.5, a five-point disagreement, but its unproven ATS record prevents treating that difference as an edge estimate. The stronger argument is Tulane's offensive transition: a first-year head coach, a newly named starting quarterback and unresolved availability for Jamauri McClure. South Alabama's possession-heavy rushing approach offers a plausible route to keeping this within one score. The market has already recognized some of that risk: the dossier records movement from Tulane -10 to -11.5 toward -9.5, so this is a diminished opportunity. The opposing case matters: South Alabama's successful opener came against FCS competition, while Tulane faced Duke, and Semonza could improve the offense with starter preparation. I would therefore place less weight on the opening scores than EL's margin-based update might. I still prefer taking more than a touchdown against a favorite whose offensive execution remains unproven, without assuming McClure will miss Saturday.

Total.Under 49.5 is a modest lean, driven by South Alabama's possession profile and Tulane's unsettled offensive execution. EL's 48.73 projection is essentially market agreement, not meaningful independent support. The under case is sustained rushing possessions and another inefficient Tulane outing. The over case is credible: Semonza receives starter preparation, South Alabama offers easier opposition than Duke, and efficient rushing can produce touchdowns as well as consume clock. Weather does not justify a large deduction: forecasts disagree on kickoff rain, with no verified strong-wind threat. Most importantly, the dossier shows the total already falling from 50.5 to 55.5 to 49.5. That move prices much of the offensive concern and makes chasing the under unattractive at high conviction. I retain a small under preference, without counting McClure as out.

Where our number looks wrong.The dossier's first_year_head_coach=false for Tulane is factually wrong: Will Hall is in his first season. The rating also misses the confirmed switch to Semonza and unresolved McClure availability. Conversely, Tulane -4.5 may overreward South Alabama's FCS opener relative to Tulane's road loss at Duke; I do not endorse that fair spread. The totals documentation also contradicts itself: it says no weather input while explicitly applying a -1.22-point wind adjustment at just 6.2 mph.

Spread.Our number is Tulane -4.5 into a -9.5/-10 market, five points toward the dog, and the model is if anything understating it: its Tulane prior carries first_year_head_coach=false, but Will Hall is in year one after Jon Sumrall left, with 21% returning production and one returning offensive-line starter. Week 1 confirmed the rebuild: 3 points, 54 rushing yards, an FBS-high 15 penalties at Duke. The switch to Kadin Semonza helps the feel, but it is his first start since 2024 behind that line, and lead back Jamauri McClure has not played since an August foot injury. South Alabama is the opposite roster: 78.6% returning production, a veteran starter in Bishop Davenport, and a three-back ground game that produced 241 rushing yards and 40:09 of possession in week 1, pointed straight at a Tulane front Duke's Nate Sheppard ran through for 223 on 32 carries. The line has already come from -11/-11.5 to -9.5 while 70% of spread tickets sit on Tulane, reverse movement in our direction, and thunderstorms shorten the game. Take the points.

Total.The market has already done most of this work, dropping from 52.5-55.5 at open to 49.5 at all five books (Edge Labs captures 9/4-9/10), so I am taking a number that has come to me, which is why this is a lean and not a bet. The script still points down. South Alabama held the ball 40:09 in week 1 with a three-back run game; Tulane ranked 127th in time of possession and scored 3 points. Fewer possessions, and one of the two offenses has shown nothing behind a rebuilt line. NWS New Orleans (9/10) has showers and thunderstorms likely through the 6pm CT kick, 80% precipitation and a quarter to half inch of rain, with a real lightning-delay risk. The honest counterweight: 85% of total tickets are already on the under and FanDuel is charging -114 for it, so the cheap side is gone.

Where our number looks wrong.Two concrete errors, both pointing the same way. First, the Tulane preseason prior sets first_year_head_coach=false when Will Hall is in fact a first-year head coach inheriting a 21%-returning-production roster; correcting that would lower Tulane's rating and widen our gap toward South Alabama, so the model's own 'pass' verdict is soft. Second, the in-season refit rewarded South Alabama for a 39-14 win over FCS SE Louisiana and only mildly punished Tulane for a road loss to a decent ACC team, so both moves are noise dressed as signal. And as always the rating sees no injuries: it does not know Tulane's lead back has not played, that its offensive line returned one starter, or that the quarterback changed this week. The 2.5-point flat home field is also probably light for a night home opener in New Orleans, which cuts the other way. Net: the direction is right, the number is not evidence.

Spread.The fresh news reinforces the gap rather than closing it. Tulane was shut out through the air and on the ground for a half, drew 15 penalties, and lost 17-3 at Duke; HC Will Hall then benched Week 1 starter Zeon Chriss-Gremillion for Kadin Semonza, who will make his first start since 2024. RB Jamauri McClure's return from a foot injury is reported but not officially confirmed. South Alabama, by contrast, ran for 241 yards and controlled the clock for 40+ minutes in its opener, and came within a missed two-point conversion of forcing overtime against this same Tulane program a year ago. The market has barely budged off double digits despite all of that negative Tulane news, which reads more like brand/rating inertia (Tulane's 11-3 2025 season, ESPN's [figure removed]) than freshly priced information. Take the extra score of cushion against a shaky, QB-changing favorite.

Total.New Orleans is forecast at roughly an 80% chance of showers and thunderstorms through the evening at kickoff (NWS, retrieved 9/10/2026), matching Edge Labs' own captured forecast of rain/partly cloudy conditions. Tulane was shut out through the air in its opener and is now breaking in a new starting quarterback making his first start since 2024; South Alabama is a run-first, clock-control offense that averaged 40+ minutes of possession in week 1. Both point toward fewer possessions and messier execution rather than a shootout. Note the total has already fallen hard, from a 52.5-55.5 open range to a uniform 49.5 across every book tracked, so much of this under-case is already priced in, that caps conviction, but the direction still favors the under.

Where our number looks wrong.The rating never saw Tulane's in-game QB benching or the penalty/O-line collapse at Duke, it only updates on final margins and the preseason prior, which is exactly the blind spot listed in its own known weaknesses, and it can't see McClure's murky return timeline either. Those specific facts cut both ways (a fresh starter or a returning workhorse RB could spark Tulane, not just hurt it), so we would not trust the model's exact -4.5 number on its own. But the market's refusal to move off -9.5/-10 despite an ugly, penalty-riddled shutout loss looks more like reputation/brand pricing (Tulane's 2025 record, ESPN's [figure removed]) than a number that has actually digested this week's news, so the direction of our model's disagreement with the market is more credible than its magnitude.