NFL Week 1 Road Underdogs: The 68.9% ATS Edge and How to Find the Setup

To beat a -110 line consistently, you need to win 52.38% of your bets. That is the math of the vig, and the number does not move.

Road underdogs in the first two weeks of the NFL season have covered the spread at 68.9% since 2004. That figure, sourced from BetMGM's historical data, reflects 84 wins, 38 losses, and 5 pushes across 127 relevant games over 22 seasons. The rate is 16.5 percentage points above break-even. The season opens September 9, and the setups on the 2026 board are already visible.

The reason this edge still exists after two documented decades is worth understanding before you touch a Week 1 line.

What the Full Data Set Shows

All Week 1 underdogs, regardless of location, cover at 53% in Week 1 games since 2000, per FanDuel Research. The ATS record: 183 wins, 163 losses, 14 pushes. Pull the sample back to 1966 and the long-run rate holds at 52.1% across more than 750 games.

Home teams, who are typically the favorites in Week 1, cover at only 45.3% compared to 48.7% across all other weeks. The home field advantage that sportsbooks build into a spread is worth less in Week 1 than at any other point in the season.

The gap between the overall 53% and the road-dog-specific 68.9% shows the filter matters. Not all Week 1 underdogs generate the same edge. The road factor adds roughly 15 percentage points above the aggregate.

Why Sportsbooks Price Week 1 Poorly

No 2026 NFL game has been played. Books set their opening Week 1 lines from offseason rosters, depth chart projections, and preseason reps. The 2026 regular-season schedule dropped May 15. Most sportsbooks had lines posted that night, before training camp opened, before preseason Week 1, and before any of the personnel decisions that define a roster's actual Week 1 shape.

Home field advantage is a specific pricing problem in September. In the regular season, a sportsbook prices home advantage using how the team performed at home over the past two to three seasons: crowd noise effects, opponent road fatigue, weather history, the team's margin in its own stadium. In Week 1, none of those 2026 data points exist. The book uses historical norms. Historical norms are a coarser tool than in-season performance data.

That coarseness lands hardest on the road team. The road dog is priced as a double underdog: disadvantaged by the spread and disadvantaged by the public's mental model of road teams in openers. Both of those disadvantages are real. The question is whether they are fully priced.

Twenty-two seasons of data suggest they are not.

The Academic Evidence

Richard Borghesi published a study in Applied Economics (Vol. 39, No. 15, 2007) examining the NFL home underdog effect. The core finding extended beyond the statistical result. Borghesi identified that the bias had persisted for decades despite being visible to researchers and market participants. He attributed this to "limits of arbitrage": the transaction costs of exploiting the edge, combined with the steady stream of public money recreating the distortion, make full market correction economically difficult for books to achieve.

A separate paper posted to arXiv (1211.4000, examining 2,560 NFL games from 2002 to 2011) found home underdogs covered at 53.5%. A strategy of backing home underdogs in that period produced a profitable cumulative return above the 52.38% break-even threshold. The researchers ran the strategy across the full decade and the edge compressed but held.

Justin Davis, Andy Fodor, Luke McElfresh, and Kevin Krieger extended the analysis into Week 2 in a 2015 paper in the Journal of Prediction Markets (Vol. 9, Issue 1). Their finding: bettors place too much weight on Week 1 results when pricing Week 2 games. The team that covered big in Week 1 attracts disproportionate public money in Week 2. The road dog bias extends into the second week of the season for the same structural reason: the public bets recency and name recognition, not current-year fundamentals.

Why Road Dogs Specifically Outperform

The overall 53% Week 1 underdog ATS rate includes home underdogs, who drag the aggregate down. Home teams in Week 1 sit at only 45.3% ATS, well below their rest-of-season rate. Home underdogs face a compounding pricing problem: books assign them home field value, the public respects their home setting, and the actual pricing leaves them with less spread cushion than the situation warrants.

Road underdogs get treated differently by the market. The public undervalues them twice. First, because the narrative of "road disadvantage" is baked into how casual bettors think about spread pricing. Second, because the public in Week 1 prices team quality based on last year's record, not on current-year roster adjustments. A team with an 8-9 record last season but a meaningfully upgraded roster and a favorable early schedule gets priced partly on that 8-9 signal. The road setting amplifies the discounting.

Books shade their lines toward where the money flows. If 75% of early bets are on the home favorite, the book adjusts the number toward the favorite to manage liability. The road dog gets a slightly better price than a fully efficient market would set. Multiply that mechanical discount across 22 seasons, and the result is a 68.9% ATS rate for a category that the market consistently underprices.

Divisional Dogs: The Subset Worth Targeting

The divisional dog data is the most pronounced subset. Since 2004, division-rival underdogs in the first two weeks have gone 62-23-3 ATS, a 72.9% rate, per BetMGM historical data. That is 72.9% on 88 decided games across 22 seasons.

Divisional opponents play each other four times a year. Coaching staffs have extensive tape. Familiarity compresses talent gaps on the spread even when the actual quality difference between the teams is real. The public prices Week 1 divisional games on the headline talent gap. The rival, who spent the offseason preparing specifically for this opponent, covers more than the spread suggests.

When the road underdog in a Week 1 or Week 2 game is also a divisional opponent, both factors stack. Road discount plus familiarity factor. The 72.9% ATS rate reflects that combination.

Reading the 2026 Board

The 2026 NFL regular season opens September 9. Nine of the sixteen Week 1 games have spreads of three points or fewer, per reporting from ESPN and FOX Sports. The largest spread: Chargers as 11.5-point home favorites over the Cardinals.

Two situations stand out based on early public betting data.

The Chiefs and Broncos close Week 1 on Monday night, September 14. Early public betting data from SportsBettingDime shows Denver drawing 80% of early spread bets. Kansas City draws 55% of the early spread money. When tickets run heavily one way but dollars split more evenly or favor the other side, the dollar-weighted action reflects larger, more sophisticated wagers. In this case, 80% of bets are on Denver while the majority of money sits on Kansas City. The lines have not moved in the direction the 80% would predict. That is the signature of reverse line movement: public bettors loading one side while bigger money sits on the other.

The Cowboys opened as 2.5-point home favorites against the Giants. Early data shows Dallas taking 72% of bets and 63% of money. The gap between ticket percentage and dollar percentage is smaller here. When bets and money align closely, the signal is more uniform rather than a sharp-versus-public split. The road dog setup is less clean.

For the rest of the board, the filter works this way. Look for games where a road team draws fewer than 35% of public spread bets and the line has not moved toward the home favorite despite the imbalance. The road underdog structure is intact when the public is fading the road team but the price is not compressing toward the home side. That combination means books are not seeing enough sharp action on the favorite to justify tightening the number.

When to Pass

The road dog edge is a structural tendency across a large sample, not a guaranteed result in any individual game. BetMGM data identifies one notable exception: Week 1 games pitting two prior-year playoff teams have shown favorites going 33-21-1 ATS since 2005. When both teams carry recent playoff credibility, the public pricing is more accurate. Higher-profile matchups draw more betting volume and more analytical attention from sharps. The low-scrutiny condition that creates the road dog edge is less present when both fanbases are engaged and both teams draw serious early action.

Also pass on road dogs when the line has already moved in their favor. If a road dog opened at +7 and sits at +5.5 by the time you are shopping, sharp money has already discovered the value. You are chasing a number that partially corrects the original edge. The 68.9% historical rate applied to lines before the sharp correction, not after.

The setups worth targeting share four traits:

• Road team drawing fewer than 35% of public bets
• Line not moving toward the heavily bet home favorite
• Neither team qualified for the prior-year playoffs
• Spread between 3 and 10 points, where covering is realistic and the price still reflects the discount

The Two-Week Window

The Davis et al. paper on Week 2 bias carries one more practical implication. The early-season road dog edge does not hold with the same magnitude by Week 3. Two weeks of real game data gives books better calibration for Week 3 pricing. The public also has results to work with rather than prior-year narratives alone. The window of maximum pricing uncertainty is concentrated in Weeks 1 and 2, which is precisely what the 2004-2026 dataset captures.

For the 2026 board, the best prices are available now, before sharp money concentrates in the 48 to 72 hours before each kickoff. When professional bettors land on a number in the final 72 hours, lines move and the early price closes. The public-driven discount on road dogs exists because the public set the market. Once sharp volume corrects the number, the 68.9% historical rate no longer applies to the adjusted price.

Week 1 kicks off September 9. The road dogs that fit the four-factor filter, drawing low public action on a line that has not moved toward the favorite, are the ones carrying the historical edge into 2026. Check the public betting splits at BetMGM, DraftKings, or FanDuel in the week before kickoff. The games where the road team draws 20 to 30% of tickets and the number is not moving are the setups this data describes.