The betting world largely leaned Chiefs; our predictive model boldly predicted: Eagles in a rout.
Not only did we have the Eagles dominating the 2025 Super Bowl – and they did 40-22 — as early as Week 9, we also called the final four playoff teams (yes, even the Commanders) months before most thought it was possible.
That’s not luck. That’s data.
At Predictive Analytics Group (PAG), we build models that help businesses reduce risk, anticipate change, and move ahead of the competition. But let’s be honest: We’re data nerds at heart and sometimes that means applying our skills to things we love like football.
Every season, a few of our developers build a predictive model to test ideas, have fun, and sharpen our approach. It’s become a sort of lab for us, a way to experiment without stakes and see just how accurate our methods can be.
📨 Stay Ahead of the Curve
Our NFL model will be sharing weekly predictions again this season—just like last year, when we nailed the full playoff picture and Super Bowl outcome weeks before anyone else.
Want to follow along?
[🔗 Subscribe on our LinkedIn Company Page] to get our latest predictive insights each Thursday, starting in Week 5.
But we don’t rush to judgment. Unlike fantasy football projections that drop in August, we typically wait until after Week 4 to publish our first real predictions. Why? Because last season’s data doesn’t account for this year’s rookie classes, coaching changes, trades, or updated rosters. Just like you can’t use sub-prime data to project prime performance, or consumer loan patterns to predict credit card behavior, we need current, relevant data that matches the environment we’re analyzing.
Last season?
- By Week 6, the model had predicted 10 of the 14 eventual playoff teams.
- By Week 14, it had nailed all 14—including their seeding.
- It called the final four teams ahead of the curve.
- And our season-long simulated “parlay” model (starting with a hypothetical 100) would’ve returned 1,340% by Week 18.
But this isn’t about betting. We don’t offer gambling advice. We never will.
This year presents a perfect example of why timing matters. While we’re seeing early signals pointing toward teams like Philadelphia, Detroit, Buffalo, and Kansas City, we’re holding our detailed projections until we have enough current-season data to make reliable calls.
This is about what’s possible when predictive analytics are done right, when they’re tuned, tested, and given room to evolve. Our NFL model doesn’t even use our proprietary GOBLIN analytics platform (we reserve that edge for our clients).
So why share this?
Here’s the thing: analytics are equal parts art and science. The numbers tell you what happened, but experience tells you what it means. Many companies get seduced by the quantitative side—the models, the algorithms, the clean mathematical outputs. But the real magic happens when you understand how those insights apply in the messy, unpredictable real world.
Because in football and finance, the rules are the same: Uncertainty is the opponent. Data is how you win.
The same techniques that let us forecast playoff brackets in October are the ones helping our financial clients make sharper loan decisions, spot risky portfolios early, and see around corners in an unpredictable market.
Imagine applying that level of foresight to your business:
- Adjusting your strategy in real time, just like a coach at halftime.
- Predicting where the market’s headed while your competitors are still reacting.
- Turning noise into signals, and patterns into confident plays.
Of course, when businesses need projections before sufficient operational data exists (like launching a new product line), we partner with credit bureaus and alternative data providers to access two years of performance data across all account types and risk levels. It’s like having film from every team in the league, even when your specific matchup data is limited.
Football gives us a sandbox. Business gives us the stakes. And GOBLIN gives our clients the edge.
Every great team has a playbook. Ours just happens to predict outcomes with precision.
Let’s talk about how predictive analytics can help you call smarter plays—before the next market shift.
