A Sports Writer Is Betting Against ChatGPT All Season to See Who Actually Knows Football

A Sports Writer Is Betting Against ChatGPT All Season to See Who Actually Knows Football

Follow America's fastest-growing news aggregator, Spreely News, and stay informed. You can find all of our articles plus information from your favorite Conservative voices. 

A Fox Sports/Outkick writer has launched a season-long experiment pitting his own NFL betting picks against ChatGPT’s, using the same odds board each week to see which one actually produces winners.

The rules are simple. Each week the writer picks five bets. Then ChatGPT, fed the same live odds and given the same instructions, picks its own five. Both records get tracked all season. No excuses, no moving the goalposts — if the AI gets a pick wrong, it’s just wrong.

For Week 1, the writer used GPT-5.6 in “Sol Max” mode, handed it the OddsChecker board as it stood Saturday at 6 p.m. ET, and told it to pick its five best bets with no line worse than -120, explaining its reasoning for every pick it made and every one it passed on. The bot took nearly 20 minutes to respond.

The Human’s Five Picks

The writer laid the Chicago Bears -3 on the road against the Carolina Panthers, backed an under between the Cleveland Browns and Jacksonville Jaguars, took the Washington Commanders +5½ against the Philadelphia Eagles, played an under on the Miami Dolphins-Las Vegas Raiders total, and bet the Green Bay Packers moneyline over the Minnesota Vikings — calling that last one his favorite play on the board.

ChatGPT’s Five Picks

The AI took Houston Texans +105 on the moneyline over the Buffalo Bills, an under in Tampa Bay Buccaneers-Cincinnati Bengals, a fade of the Detroit Lions laying 7½ against the New Orleans Saints, a fade of the Los Angeles Chargers laying 9½ against the Arizona Cardinals, and — matching the writer’s pick — the Green Bay Packers, though at moneyline +110 rather than the -½ spread.

ChatGPT backed its Bills pick by pointing to Houston’s defensive success against Josh Allen and ESPN’s FPI projecting Buffalo by only one point. On Tampa Bay-Cincinnati, it noted both quarterbacks barely played in the preseason and both defenses added personnel, making the 51-point total tough to clear. It leaned on Detroit’s home offense and New Orleans’ banged-up pass protection for the Lions game, and cited Arizona’s leaky defense — 35.2 points allowed per game over its final nine contests last season — against a Chargers team projected to win by 9.6 points per FPI.

The writer pushed back on several of the AI’s calls. He said he

Share:

GET MORE STORIES LIKE THIS

IN YOUR INBOX!

Sign up for our daily email and get the stories everyone is talking about.

Discover more from Liberty One News

Subscribe now to keep reading and get access to the full archive.

Continue reading