ChatGPT Shocks College Football Fans with Unbelievable Week 13 Upset Predictions Against the Spread

You ever wonder if a super-smart AI can outthink a seasoned bettor when it comes to college football spreads? Well, last weekend I threw ChatGPT into the ring to see if it could out-pick me. Spoiler alert: it flopped pretty hard, managing just a 40% success rate while I edged out a modest winning record. But hey, if at first you don’t succeed, why not let the computer have another go, right? This week, I nudged the bot to factor in injuries and even tossed in a bit of “analysis” to see if it could level up its game. Did the machine finally crack the code, or are both man and machine just fumbling down the stretch? Buckle up — the final weeks of the season are shaping up to be a showdown neither of us will forget. LEARN MORE.

In case you may have missed it last weekend, I turned to ChatGPT for a little assistance in the ol’ picking college football games against the spread department. While I was fully expecting the computer to help me out in picking some winners, it failed miserably, posting a record of 14-21 against the spread for a success rate of just 40%. I went 19-16 with my picks last week.

That, my friends, will not get the job done, but if at first you don’t succeed, you try (and have a computer make your life easier) again.

The thought process for going back to ChatGPT for pick ‘em assistance this week was simple: it can’t get any worse, and while that came to be true, the computer didn’t exactly knock it out of the park.

In total, I had it select 33 games to pick this week. I kept the prompt with ChatGPT straightforward, telling it to make picks against the spread, which I provided, and to take into consideration key injuries. The one difference from this week to last week is that I also told ChatGPT to “use analysis” while picking the games as well, which the chat took as considering home-field advantage and look-ahead spots on the schedule.

Like last week, ChatGPT also ranked each game with a confidence level between 1 and 5, unprompted to do so, with 5 being the most confident.

ChatGPT didn’t rank a single game with a confidence level of 5, but did list two games with a confidence level of four and then eight more picks with a confidence level of 3.

Confidence Level 4

Washington -10.5 @ UCLA – Winner

Miami -17.5 @ Virginia Tech – Loser

Confidence Level 3

Duke -6.5 @ North Carolina – Winner

Nebraska +9.5 @ Penn State – Loser

Tennessee -4.5 @ Florida – Winner

USF -21.5 @ UAB – Winner

Illinois -8.5 @ Wisconsin – Loser

BYU -2.5 @ Cincinnati – Winner

North Texas -18.5 @ Rice – Winner

Kansas State @ Utah -17.5 – Loser

While ChatGPT went 6-4 with its most confident picks, it posted an overall record of 17-16 against the spread this week. Surprisingly, I only disagreed with three of its picks this week, which I flipped, and unfortunately lost to the machine as I went 16-17 against the spread.

The takeaway here is that both I and AI suck at picking college football games.

With just one week left in the college football regular season, I think the next move is fully fading ChatGPT next weekend, which will surely come back to haunt me.