Human vs. ChatGPT NFL Week 2 betting picks: AI looks to rebound after disastrous 0-5 debut

Read the original at Fox News ↗
Fox News · collected 2026-09-17 · by Dan Zaksheske

Quick Summary

In a competition between human and AI NFL betting skills, the author predicted that ChatGPT would outperform him in week one but instead lost all five bets, finishing with a 5.23-unit loss. The author won three of his five picks, gaining 0.99 units, suggesting human intuition still holds an edge over current artificial intelligence for sports betting. The article outlines the rules and provides detailed results from both competitors’ first week of NFL predictions.
Written locally by qwen2.5:14b on 2026-09-18, using this article's own text rather than the other coverage of the same event (that is the story summary below).

AI analysis runs on qwen2.5:14b, locally

Story summary

In a unique NFL betting challenge for Week 2, Dan Zaksheske competed against ChatGPT, an AI model. In Week 1, Dan performed better than expected with a 3-2 record and gained 0.99 units, while ChatGPT went 0-5, losing 5.23 units. This initial week suggested that humans might still outperform AI in NFL betting predictions, at least for now. The competition adheres to strict rules where each bet must be priced no worse than -110 and is sized to win one unit. Both participants will continue their head-to-head bets throughout the season.

Written for “AI Football Betting Picks” on 2026-09-18, grounded in this article and the 0 other(s) covering the same event.

Signals How these are calculated →

Claims extracted
102
claim-shaped sentences
Uncertain
6%
6 of 102 hedged
Leaning
not political
takes no side on a contested political question
Correction & hedging signals
95.2
corrections and hedging in what we collected; not a measure of accuracy
Outlets on this story
1
Technology
Narrative spread
1
articles carrying this framing
Analyzed 2026-09-18 · how these are computed

Story

📰 AI Football Betting Picks
Technology · 1 article(s) covering the same event.

How this is being covered How these are calculated →

Article leaning vs. publisher reliability
Source leaning vs. consistency

Compared with similar articles

This article reads unscored and hedges 6% of its claims. Each row says how that neighbour differs.
New York Post
⚖️ leaning not scored 🔴 8% hedged 2 of 26 📰 publisher trust 59
“The articles discuss different aspects of NFL Week 2 betting and analysis, with Article A focusing on general picks and confidence pool rankings while Article B covers a human vs. AI betting competition.”

Publisher

Fox News · 801 article(s) · 1 correction(s) detected
Running correction rate · 1 correction(s)
2026-08-04
Pennsylvania school bus aide accused of sexually assaulting special needs teen

Who wrote this

Dan Zaksheske
6 article(s) here · 1 carrying a prediction
🔮 I also openly predicted that artificial intelligence would wipe the floor with me.
🔮 The congressman said he has not received a response from the NHL but is willing to give the league time to answer.
🔮 Yet, people are going to act like that one out of 17 games is plenty of information to inform us about how the rest of the season will go.
🔮 "Why would they build an entire program around blaming coaches when they know that that is not true?" she said.
🔮 At some point, you have to wonder whether Mother Jones can even imagine that people might sincerely believe women’s sports should be for women.
🔮 As sad as I am to see summer coming to an end, I couldn't be more excited that the best league on the planet is back and in full swing.
Also by Dan Zaksheske
Nothing else under this byline is closely related to this article, so these are simply their most recent.
All 6 articles by Dan Zaksheske →

Topics

ChatGPT FOX News Green Bay Packers NFL Packers

Subjects

Browns ORG · 6× ChatGPT ORG · 5× Cleveland Browns ORG · 3× Green Bay Packers ORG · 3× Buccaneers ORG · 2× GPT ORG · 2× Los Angeles Chargers ORG · 2× Packers ORG · 2× Tampa Bay Buccaneers ORG · 2× Washington Commanders ORG · 2×

Narrative

Week 1 NFL Betting Results Dan’s Week 1 NFL Betting Picks (3-2, +0.99 units) - Chicago Bears -3 (-106) ✅ (+1 unit) - Jacksonville Jaguars-Cleveland Browns Under 40½ (-110) ❌ (-1.1 units) - Washington Commanders +5½ (-106) ✅ (+1 unit) - Las Vegas Raiders-Miami Dolphins Under 41 (-110) ✅ (+1 unit) - Green Bay Packers Moneyline (+110) ❌ (-0.91 units) The Packers losing at plus money softened the damage, and hitting both the Bears spread and Raiders-Dolphins under secured a profitable opening week.
framing: assertive · carried by 1 article(s) · first seen 2026-09-18
🔮 I also openly predicted that artificial intelligence would wipe the floor with me.

Claims (102 extracted, 6 hedged)

Last week, I kicked off a season-long experiment in which I go head-to-head with ChatGPT on NFL bets. asserted
I → kick → bets
I also openly predicted that artificial intelligence would wipe the floor with me. asserted
intelligence → predict → me
It's important to be nice to our future overlords and pray that they show mercy. asserted
they → pray → mercy
Apparently, ChatGPT felt sorry for me because it proceeded to go 0-5. asserted
it → feel → 0
This page may contain affiliate links to legal sports betting partners. uncertain
page → contain → partners
If you sign up or place a wager, FOX News may be compensated. uncertain
News → sign → wager
This content was created by a team that works independently from the FOX newsroom. asserted
that → create → newsroom
Meanwhile, I finished 3-2 and gained 0.99 units. asserted
I → finish → units
That’s hardly the kind of week that makes the guys in Vegas suspicious, but it looks spectacular next to ChatGPT’s 5.23-unit loss. asserted
it → ’ → loss
One week is obviously far too small a sample to draw any sweeping conclusions. asserted
is → draw → conclusions
So, naturally, I’m going to draw one anyway: humanity is safe. asserted
humanity → go → one
The rules remain the same. asserted
rules → remain → ?
ChatGPT and I each make our five best bets using the same set of odds, and no selection can be priced worse than -120. asserted
selection → make → -120
Every wager is sized to win one unit, which means a -110 favorite risks 1.10 units to win one, while a +110 underdog risks approximately 0.91 units to win one. asserted
underdog → size → one
Before we get to the Week 2 picks, let’s review the opening week. asserted
’s → get → week
Week 1 NFL Betting Results Dan’s Week 1 NFL Betting Picks (3-2, +0.99 units) - Chicago Bears -3 (-106) ✅ (+1 unit) - Jacksonville Jaguars-Cleveland Browns Under 40½ (-110) ❌ (-1.1 units) - Washington Commanders +5½ (-106) ✅ (+1 unit) - Las Vegas Raiders-Miami Dolphins Under 41 (-110) ✅ (+1 unit) - Green Bay Packers Moneyline (+110) ❌ (-0.91 units) The Packers losing at plus money softened the damage, and hitting both the Bears spread and Raiders-Dolphins under secured a profitable opening week. asserted
Packers → lose → week
We were one garbage-time Deshaun Watson touchdown away from cashing a 4-1 week, but that's why they call it gambling. asserted
they → cash → it
Honestly, I felt pretty good about the process. asserted
I → feel → process
The Packers were the far superior team for a large portion of the game before laying an egg and, as mentioned, the over only hit in Browns-Jags because of a 46-yard touchdown in the final two minutes of a blowout. asserted
over → lay → blowout
Detroit won but failed to cover, while Houston, Los Angeles and Green Bay all lost outright. asserted
Houston → win → ?
The Buccaneers and Bengals also sailed past the total. asserted
Buccaneers → sail → total
ChatGPT spent nearly 20 minutes analyzing every eligible spread, total and moneyline before producing its card. asserted
ChatGPT → spend → card
Apparently, it needed another 20. asserted
it → need → 20
Cleveland Browns +8 (-108) Cleveland Browns at Tampa Bay Buccaneers, 1 p.m. ET Dan’s analysis: It's terrifying to back the Browns after that putrid Week 1 performance, I get it. asserted
I → back → it
But the lookahead line, which was posted before Week 1 results could influence the market, had the Browns +6, so oddsmakers have baked in how bad the Browns looked and moved them to more than a touchdown underdog. uncertain
Browns → post → underdog
And, let's be honest, the Buccaneers didn't play all that well, either. asserted
Buccaneers → let → ?
I think this is too much of an overcorrection and the public is chomping at the bit to bet against the Browns. asserted
public → think → Browns
We're going to fade the public here and take the points. asserted
We → go → points
Still, the Browns’ offensive-line concerns and Deshaun Watson’s turnover risk give Tampa Bay several paths to create separation, so I would lean toward taking the points but stop short of betting it. asserted
I → give → it
Dan’s analysis: Week 2 is a great time to fade Week 1 perceptions, and this game is a perfect example. asserted
game → fade → perceptions
Everyone is piling on the Packers after they blew a double-digit second-half lead to the Carson Wentz-led Minnesota Vikings (rightfully so). asserted
they → pile → Vikings
But the truth is that Green Bay is still a far superior team to New York, which managed to beat lowly Tennessee in Week 1. asserted
which → manage → Week
The lookahead line on this one was Packers -5½, so again, there's been an overcorrection. asserted
line → be → one
I just don't see it. asserted
I → see → it
I expect Green Bay to compete for an NFC North title and I expect the Jets to be better than last season, but not a team that should hang with an already desperate and hungry Packers team. asserted
that → expect → team
GPT’s analysis: I agree with fading the Week 1 overreaction, but crossing the key number of three makes this less appealing. asserted
this → agree → three
Green Bay has the more talented roster, but Geno Smith has repeatedly succeeded against Jonathan Gannon-coached defenses, so I would have needed Packers -3 to make the bet. asserted
I → have → bet
With Tua Tagovailoa still not practicing and uncertainty surrounding Atlanta's Week 2 starter, that's not good news for points. asserted
that → practice → points
The Falcons only scored 13 in Week 1, but this total is 44 because the Panthers surrendered 59 points to the Bears. asserted
Panthers → score → Bears
Obviously, I don't see a repeat coming. asserted
repeat → see → ?
…and 62 more, not listed.
💬 Give feedback
🕘 History 🎫 Support