Edge Labs EDGE LABS Market intelligence
CFB NFL
← The slate
CFB · Week 2 · Sat, Sep 12 · 7:45 PM ET · Jordan-Hare Stadium · FINAL Southern Miss 8, Auburn 43

Southern Miss at Auburn. Our picks: Southern Miss +33.5 (medium). Under 56.5 (low).

The picks come first on this page. The market's number, our number and the gap between them sit under them as MODEL CONTEXT: that is the input our analyst panel argued from, not a call of its own. Tiers are how decisively the evidence agreed, never a win probability.

ATS pick
Southern Miss +33.5
medium LOSS
locked Wed 6:59pm ET
locked +33.5 · closed +32.5
Southern Miss 8, Auburn 43
Total pick
Under 56.5
low WIN
panel 2-1

Panel reasoning

On the spread

Southern Miss at Auburn, Auburn -33.5. The line moved from -29 to -33.5 in three days at every book on lopsided Auburn money after a 17-16 escape against Baylor in which Byrum Brown threw three interceptions and Auburn turned five Baylor giveaways into three points. Alex Golesh is in his first season (our prior has that flag wrong), the interior line was inconsistent by his own account, receivers Williams and Neptune are questionable, and Womack is day to day, though Skinner, Strand and Murray should be back. Southern Miss has quarterback Ethan Hampton healthy after his hand injury, a 49-3 opener and most of last year's production returning. Our number leans Southern Miss by 13 points; the rating is preseason-heavy, so discount its size, but 33.5 asks a team that scored 17 to win by five touchdowns. The counter-case is the SEC talent gap and an FCS opener on the other side. Take Southern Miss +33.5, available everywhere.

On the total

Under 56.5. The total climbed from 52 to 56.5 (57.5 at Pinnacle) with the spread, and 56.5 implies Auburn near 45; Auburn has not shown 45, and its goal-to-go completion rate was the nation's worst last year. The forecast has changed since our capture: showers and possibly a thunderstorm are likely through the 6:45pm Central kickoff. Our totals model reads 49, which is too low to take literally but points the same way. The over case, and one panel took it, is Golesh's tempo, Southern Miss scoring late against reserves and a thin, low-liquidity number. Under 56.5 at retail, 57.5 at Pinnacle.

Model context

The numbers the panel argued from. Our models price a game, the panel makes the call, and the picks above are the call. Nothing in this block is a pick.

Locked line
Southern Miss +33.5 · Auburn -33.5
the spread we locked at, it does not move
Our model number
Auburn by 20.1
the calibrated EL CFB rating number, home field included
Spread gap
12.4 points
the distance between our model number and the market, on the Southern Miss +32.5 side. Context for the panel, not a call of its own.
Market total
57
consensus of 5 books at the model run, Sep 12, 12:38 PM ET
Our total
48.9
EL CFB Totals v1
Total gap
-8.1
the distance between our model total and the market total. Context for the panel, not a call of its own.

Context, not a call. The spread number is our rating difference plus home field. The totals number is a registered candidate that loses to the closing total: across 2,264 held out games it missed the combined score by 12.862 points on average and the closing total missed by 12.635, ahead of us in all 3 of those seasons, so it never posts to the board and never enters the record on its own.

Over and under splits

Descriptive context on this game, every line with the games it counts. None of it is a reason a pick wins and none of it enters a tier.

  • Early-season total of 56.5 fits the validated under window, unders 55% historically (n=943)
    total · under · moderate · n=943
    H64 (validated): Aug through mid-Sep games with a closing total of 47+ landed under 55% of 943 games, 2015-2025. Recent seasons: 2022: 58% (n=92) · 2023: 53% (n=81) · 2024: 57% (n=103) · 2025: 52% (n=119). 2025 ran soft at 52.1%, below the -110 breakeven, and the 2026 forward test is pre-committed on the early-totals page. Early-season totals are shaded up by public over-action; the market corrects as real form data arrives.
  • Auburn games go UNDER 57% at home (n=80)
    total · under · moderate · n=80
    Multi-season split: Auburn at home landed under 57% of 80 gradable games (2013-2025). A genuine venue tendency, one of several reasons to lean under, not a standalone edge.
  • Auburn have gone UNDER in 7 of their last 10
    total · under · lean · n=10
    Auburn's last 10 gradable games landed under 70% vs the consensus total. Streak continuation tested 2013-2025: 49.5% train / 46.4% held-out 2025, it does NOT persist. A real recent pattern, weigh with the matchup, not alone.

The panel

Panel A

Southern Miss +33.5

Under 57.5

Southern Miss +33.5 is a modest price-based lean. Our Auburn -20.1 projection differs by 13.4 points, but its unproven ATS performance and incorrect coaching input make that gap substantially less persuasive than it appears. The captured move from roughly Auburn -29 to -33.5 already prices a sizable mismatch; I could not verify betting splits or attribute the move to informed money. Auburn's 17-point opener understates its offensive potential: Baylor reported 389 Auburn yards and three interceptions, so cleaner execution could produce a considerable rebound against weaker opposition. Conversely, laying nearly five touchdowns requires sustained separation from an offense still establishing itself. Southern Miss's 49-3 opener provides little evidence against SEC competition, but taking the expanded cushion allows for substantial Auburn improvement. Optimistic Auburn injury reporting subtracts from the underdog case. This is a small lean, not a high-tier opportunity justified by the model gap.

Total. Under 57.5 is a narrow lean using the dossier's Pinnacle quote, corroborated as an available total level by Covers. The 49.04 model forecast is too low to accept literally: Golesh's tempo, Auburn's turnover-suppressed opener and the weaker opposing defense support the over. Southern Miss scoring late against reserves is another meaningful risk. Against that, the market has already raised totals from 52 to 53.5 to 56.5 to 57.5, requiring a much stronger scoring environment. Auburn's offensive execution remains uncertain, and Southern Miss's FCS opener does not establish reliable scoring against Auburn. Possible evening showers offer only minor support; there is no verified strong-wind argument. I prefer the higher under number, with limited conviction because the coaching transition makes historical scoring data less informative.

Where our model looks wrong

Auburn's first_year_head_coach=false input is demonstrably incorrect: Alex Golesh debuted September 5. The model also cannot adequately represent the Golesh/Byrum Brown offense through last year's production and one final score. Southern Miss's 90.2% returning-production assumption was not independently verified and should be audited before trusting this gap. The Alcorn blowout is weak evidence for narrowing the talent differential. The kickoff field is also wrong: official scheduling says 7:45 p.m. ET, not midnight.

Panel B

Southern Miss +33.5

Under 56.5

Our number is Auburn -20.1 against a market of -33.5, and every rating path lands near the same place: preseason priors of Auburn +12.0 and Southern Miss -7.1 plus 2.5 home field is about 21.5, so the market is roughly 12 points above any power-rating projection. The counter-signal is real and I am pricing it rather than ignoring it: the line moved -29.5 to -33.5 between 9/6 and 9/9 on 99% of the money (cleatz splits, 9/9), which is a sharp move, not a public one. But the football case for the dog is concrete. Auburn scored 17 on Baylor while Byrum Brown threw three interceptions and the offense turned five extra possessions into three points (ESPN recap, 9/5). Golesh blamed the guards for not being physical enough (9/7 presser) and both guard spots carry OR designations on the released depth chart (9/7). Southern Miss returns 90.2% of production, led the nation with 23 interceptions in 2025, and is the veteran profile that backdoors when Auburn empties the bench. Laying 34 with a one-game, 17-point offense is the wrong price.

Total. 56.5 implies roughly Auburn 45, Southern Miss 11.5. Auburn has not shown 45. It moved the ball on Baylor (Brown 26-of-35 for 259) but turned five extra possessions into three points, and FOX's game page notes Auburn had the nation's worst goal-to-go completion rate in 2025. The weather has genuinely changed since our capture: NWS Birmingham for Auburn on 9/12 reads showers likely and possibly a thunderstorm after 1pm, 70% chance, high near 90, then 40% Saturday night mainly before 1am, for a 6:45pm CT kickoff. Our dossier's captured 25%/74F forecast is stale. Southern Miss's quarterback is unsettled against an Auburn defense returning most of a top-10 2025 unit, so the dog's half of this number is soft. The total was bought from 52 up to 56.5 in three days; I am selling that inflated number. Golesh's tempo and a garbage-time blowout are the honest risk.

Where our model looks wrong

Three places. (1) The dossier flags Auburn first_year_head_coach: false, which is simply wrong - Alex Golesh is in his first season at Auburn and said so himself on 9/7; a first-year staff installing a new tempo offense is exactly the volatility the prior should be widening for, and it is being credited to a continuity profile instead. (2) The in-season v1 refit is overreacting to a single week: Auburn dropped 12.02 to 9.91 for winning by one over a real P4 team, while Southern Miss climbed -7.12 to -4.76 for beating an FCS opponent by 46. Those two adjustments move in the wrong directions relative to opponent quality and account for roughly two points of our gap. (3) The flat 2.5 home field and the rating's blindness to depth mean it structurally cannot price a P4-vs-G5 blowout spot where Auburn's blue-chip ratio is 0.52 against 0.015 - the market's extra 12 points is largely a depth and attrition premium the model has no term for. I am taking the dog on the price and on the news, not because the model number is trustworthy; its own holdout says 46-50% ATS at every gap size.

Panel C

Southern Miss +33.5

Over 56.5

Market moved from Auburn -29/-30 (9/6) to -33.5 (9/9) across DK, BetMGM, FanDuel and Pinnacle alike, and total climbed with it from ~52 to 56.5-57.5 -- both up roughly 4-4.5 points in three days (Edge Labs line-history capture, 9/9). That kind of synchronized push on a thin, low-liquidity Sun Belt-at-SEC line looks more like public 'bounce-back favorite' money after Auburn's unconvincing 17-16 escape over Baylor than hard new information; Southern Miss's 49-3 romp came against FCS Alcorn State and tells us little. Auburn is trending healthier up front (OL Skinner/Strand, DE Murray all cleared; Golesh called the team healthy, Yahoo Sports 9/7-9/8), but Covers' 9/9 scan lists three Auburn WRs (Williams Jr., Neptune, Smith) and two RBs (Davenport, Meadows) questionable -- real ceiling risk on a night they're being asked to cover five touchdowns. Laying 33.5 in a non-conference get-right spot invites the garbage-time/backdoor risk that plagues extreme numbers. Our own model, while explicitly unproven ATS, flags this as its largest gap of the week. Taking the points.

Total. Total climbed with the spread -- roughly 52 to 56.5/57.5 across books between 9/6 and 9/9 (Edge Labs line history) -- but our totals model's own gap (7.7-8.5 pts) doesn't cross its 12.86-point action threshold, so it stays silent here; I'm not leaning on it either way. Southern Miss debuted an explosive offense (49 points) with a healthy Hampton at QB, and even a Sun Belt defense that held Alcorn State to 3 hasn't been tested by SEC speed yet. Auburn, even missing pieces at WR/RB, should still move the ball on a clearly overmatched defense at home, and a lopsided script typically adds mop-up scoring from both sides late. Weather (mild, low wind, roughly 25-30% chance of an isolated early shower) is not a total-suppressing factor here. Taking the over.

Where our model looks wrong

The model's -20.1 fair number vs a -33.5 market implies Southern Miss should be getting 13.4 more points than the market offers -- its largest gap class ('high tier'). But the model is explicitly non-predictive ATS at every historical gap bucket (46-50%), ignores this week's injuries entirely, and one game into the season is still mostly a preseason prior for both teams. I am not treating its gap as a market-inefficiency signal on its own; the ATS lean here rests on market-structure reasoning (thin liquidity, likely public favorite money, extreme-number backdoor-cover risk) plus the fresh WR/RB questionables, not on the model disagreeing with Vegas.

The write-up

The case

Southern Miss at Auburn, Auburn -33.5. The line moved from -29 to -33.5 in three days at every book on lopsided Auburn money after a 17-16 escape against Baylor in which Byrum Brown threw three interceptions and Auburn turned five Baylor giveaways into three points. Alex Golesh is in his first season (our prior has that flag wrong), the interior line was inconsistent by his own account, receivers Williams and Neptune are questionable, and Womack is day to day, though Skinner, Strand and Murray should be back. Southern Miss has quarterback Ethan Hampton healthy after his hand injury, a 49-3 opener and most of last year's production returning. Our number leans Southern Miss by 13 points; the rating is preseason-heavy, so discount its size, but 33.5 asks a team that scored 17 to win by five touchdowns. The counter-case is the SEC talent gap and an FCS opener on the other side. Take Southern Miss +33.5, available everywhere.

Confidence: medium

Three independent analyst panels (A, B and C) worked the same evidence pack and the week's news independently, each returned a side, and each then ranked its five strongest picks for the week. Every panel was on this side. Medium is a decisiveness label for how firmly the panel agreed, never a win probability: high means every panel was on this side and ranked it among the week's best; only high picks count on the record and go free.

Our own number leaned Southern Miss +33.5 (high) before the panel sat. In college football a model lean becomes a pick only when the panel majority lands on the same side, which it did here.

The total: Under 56.5 (low)

Under 56.5. The total climbed from 52 to 56.5 (57.5 at Pinnacle) with the spread, and 56.5 implies Auburn near 45; Auburn has not shown 45, and its goal-to-go completion rate was the nation's worst last year. The forecast has changed since our capture: showers and possibly a thunderstorm are likely through the 6:45pm Central kickoff. Our totals model reads 49, which is too low to take literally but points the same way. The over case, and one panel took it, is Golesh's tempo, Southern Miss scoring late against reserves and a thin, low-liquidity number. Under 56.5 at retail, 57.5 at Pinnacle.

Our number

edgelabs EL rating (preseason v1 until a team has played, then the in-season v1 Elo-style refit, Sun 23:00 and Tue 02:00 PT) run through the 2026-09-07 CFB calibration map; home field 2.5. Home perspective: negative = home favored. Ours: Auburn -20.1; market at synthesis: -33.5. Held-out 2024-2025 reconstructions (9,927 FBS games): 46-50% against the spread in every gap bucket, calibrated margin error 12.7 points. No against-the-spread edge at any gap size. Treat the number as context, not as a priced edge.

written before kickoff, frozen at kickoff

Trends

Descriptive context, every line with the games it counts. Filters like these did not hold up as predictors in our testing, so none of this is in the confidence read and none of it is a reason a pick wins. Situational splits look back 3 seasons and need at least 8 decided games to be shown at all. Records against the number and on the total join these once our closing-line history is restored; the straight-up splits are live now.

Southern Miss
  • won 1 of their last 5 (n=5, 2025 to 2026)
  • 0-10 straight up in divisional games after a straight up loss (n=10, 2024 to 2025)
Auburn
  • won 2 of their last 5 (n=5, 2025 to 2026)
  • 1-9 straight up in day divisional games (n=10, 2024 to 2025)
The rest of the splits we can compute on this game
Southern Miss
  • won 6 of their last 10 (n=10, 2025 to 2026)
  • 2-10 straight up in day games after a straight up loss (n=12, 2024 to 2026)
  • 4-14 straight up after a straight up loss (n=18, 2024 to 2026)
  • 2-6 straight up in home divisional games (n=8, 2024 to 2025)
  • 3-8 straight up in day divisional games (n=11, 2024 to 2025)
  • 3-8 straight up on the road (n=11, 2024 to 2025)
  • 5-11 straight up in day games (n=16, 2024 to 2026)
  • 5-11 straight up in divisional games (n=16, 2024 to 2025)
  • 3-5 straight up in home day games (n=8, 2024 to 2026)
  • 3-5 straight up in road divisional games (n=8, 2024 to 2025)
  • 4-6 straight up at home after a straight up loss (n=10, 2024 to 2026)
  • 4-6 straight up in night games (n=10, 2024 to 2025)
  • 6-8 straight up at home (n=14, 2024 to 2026)
Auburn
  • won 3 of their last 10 (n=10, 2025 to 2026)
  • 1-7 straight up in home divisional games (n=8, 2024 to 2025)
  • 3-13 straight up in divisional games (n=16, 2024 to 2025)
  • 2-7 straight up in divisional games after a straight up loss (n=9, 2024 to 2025)
  • 2-6 straight up in road divisional games (n=8, 2024 to 2025)
  • 5-10 straight up in day games (n=15, 2024 to 2026)
  • 3-6 straight up on the road (n=9, 2024 to 2025)
  • 6-4 straight up in night games (n=10, 2024 to 2025)
  • 4-5 straight up in day games after a straight up loss (n=9, 2024 to 2026)
  • 8-7 straight up after a straight up loss (n=15, 2024 to 2026)
  • 7-8 straight up at home (n=15, 2024 to 2025)
  • 4-4 straight up in home night games (n=8, 2024 to 2025)

The signals and the work

Members

7 independent signals on this game

The signal tally per side, the totals read, every angle with its sample size, and the line movement from open to now. 3 picks are released free on the slate every week. $24/mo, founder rate.

The two numbers underneath

TeamEL ratingConference
Southern Miss -4.76 Sun Belt
Auburn 9.91 SEC

Rating difference plus 2.5 points of home field is the rating read above. Full board: the power ratings.

Conditions at kickoff

Partially cloudy. 79F · wind 3 mph · 23% chance of precipitation

Provenance: schedule and finals from the Edge Labs database (2026 season); the line is the latest capture for this game with the book named; our number is the EL rating read (in-season once a team has played, preseason before that) plus 2.5 home field, none at neutral sites, and it is context on this page rather than the call. Signals come from the trend engine, each with its real sample; a tier is how decisively the evidence agreed and is never a win probability. Once the game kicks off this page stops computing and reads our pick lock ledger instead: the pick and the spread we were locked at, written once at kickoff, never updated, graded against that same number. Line movement comes from our permanent capture log, one book across both ends and never a capture taken at or after kickoff, so the second number is the last pregame line and after kickoff it is the close; a game the log has captured only once shows no movement rather than an invented one. Weather is the captured forecast for the venue, and the chip appears only when it is worth saying (wind at 12 mph or more, a 50 percent or better chance of rain, or 35F or colder), never on an indoor venue. Injury counts are our latest daily scan, real report rows only, and a team the scan does not cover is left out rather than shown as zero. The write-up is assembled from those same stored rows, never written around them: the case comes from the lock row, the signals from the trend engine with their own samples, the series from our game database from 2013 forward, and the scoring profiles from completed games only, each with the number of games it averages. A section with no data behind it is left out instead of filled in, it refreshes while the game is pregame, and it can never be edited once the game has kicked off. Trends are descriptive only, computed by the same shared module the write-up uses so the two cannot disagree: straight-up splits come from completed games in our database, situational splits look back three seasons and need at least eight decided games, records against the number and on the total arrive with the closing-line restore, and every line carries the games it counts. Filters like these did not hold up as predictors in our testing, so none of them enter the confidence read. The totals block is a read and not a pick: our number is EL CFB Totals v1 (edgelabs.el_cfb_totals, season 2026, model_version v1, docs/el-cfb-totals-v1.md), a registered candidate that loses to the closing total, so it never posts to the board or the record; the market total beside it is the same capture the spread comes from before kickoff and the frozen consensus the model priced against after it, and the gap is measured against whichever number is printed. The totals trends under it are the trend engine's own rows for this game, deduped on the headline, each carrying its sample. On the games the analyst panel works each week (the games our model prices off the market plus the ranked and Power Four games, at most twenty), the pick and the total call are the panel's synthesis: three independent analyst panels, the same evidence pack, their own research, one side each, tiered by how firmly they agreed, never a win probability. A panel majority on a side is what makes a pick; without one the game is a published no-pick, printed as No pick rather than left blank. Research and context, never a guarantee, 21+.