In Week 7 of an ongoing comparison between human and artificial intelligence NFL betting predictions, the human bettor (Dan Zaksheske) achieved a 4-1 record and gained 3.11 units, while ChatGPT finished 2-3 and lost 1.33 units [1]. The human's successful picks included the Cincinnati Bengals -2½, San Francisco 49ers-Atlanta Falcons Under 49½, New York Giants-Houston Texans Over 42½, and Green Bay Packers +3½, though the Titans Moneyline pick failed [1]. ChatGPT's winning selections were limited to the Pittsburgh Steelers -3 and Cincinnati Bengals -2½, with losses on the Minnesota Vikings -6½, Denver Broncos-Arizona Cardinals Under 43½, and Kansas City Chiefs +3 [1].
Over the full season to date, the human maintains a significant advantage with a record of 19-15-1 and +3.20 units gained, compared to ChatGPT's 15-19-1 record and -5.36 units lost [1]. Breaking down by bet type, the human is 12-11-1 (+0.54 units) on spreads and moneylines, and 7-4 (+2.66 units) on totals, while ChatGPT shows 9-13 (-4.90 units) on spreads and moneylines, and 6-6-1 (-0.46 units) on totals [1]. The human's special "Plug Your Nose and Hit 'Place Bet' Bet of the Week" has produced a 4-2 record with +2.01 units [1].
For Week 8, both predictors submitted five selections from the same odds board with a maximum price of -120 [1]. The human's picks include Dallas Cowboys -6½ against Arizona, Baltimore Ravens +3½ against Buffalo, Indianapolis Colts-Jacksonville Jaguars Under 46½, Los Angeles Chargers +7, and Kansas City Chiefs-Denver Broncos Under 41 [1]. The human's rationale for the Cowboys pick emphasizes Dallas's offensive improvement and Arizona's red zone struggles, while the Ravens selection involves paying juice for value [1].
CONCLUSION
The human bettor has demonstrated consistent superiority over ChatGPT in NFL betting predictions, maintaining a +3.20 unit advantage over the season with a 19-15-1 record compared to ChatGPT's -5.36 unit deficit and 15-19-1 record [1]. Week 7's 4-1 performance extends the human's winning momentum heading into Week 8 [1]. This comparison suggests that human judgment and analysis currently outperform AI-generated betting selections in professional sports wagering.