Predicting the Win: What the NFL Draft Teaches Us About Data, AI, and the Future of Work

Every organization—whether it’s an NFL franchise or a financial services firm—faces the same fundamental challenge: making high-stakes decisions about people, roles, and the future with incomplete information, under time pressure, and against competitors who have access to the same data you do. The question isn’t who has the data. It’s who has the better framework for using it.

At Innovation Summit 2026, NFL Network’s Predictive Analytics Expert Cynthia Frelund delivered one of the most practically applicable talks of the event—using the NFL Draft as a lens to explore how organizations should think about AI adoption, talent strategy, data-driven decision-making, and the irreplaceable role of human judgment. Her core message: AI doesn’t replace the decision-maker. It empowers them to make better decisions, faster, with less bias and more context.

For organizations building future-ready operations, the parallels are impossible to ignore. Whether you’re drafting a quarterback or onboarding a new data science team, choosing a technology vendor or restructuring a workflow around AI, the principles Frelund outlined apply directly. Here are the key takeaways.

About the Session

Cynthia Frelund is the NFL’s On-Air Predictive Analytics Expert and one of the most recognizable voices in sports data science. Her background spans biology, private equity, and finance before landing in football—a path that shaped her belief that the best analytical work sits at the intersection of quantitative rigor and qualitative judgment. She spoke virtually at Innovation Summit 2026 in Pittsburgh a month before the NFL Draft came to the city.

More Data Isn’t Better. Better Questions Are.

One of Frelund’s most counterintuitive points was a direct challenge to the assumption that winning the data game means having the most data. The NFL combine—where hundreds of prospects are measured, timed, weighed, and evaluated in an elaborate two-week spectacle—explains only about 5% of career success at the next level, she argued. It’s a job interview, not a performance preview.

The 40-yard dash gets enormous attention, but as Frelund pointed out, when’s the last time you watched an offensive lineman run 40 yards in a game? The data being collected is real. The problem is it’s the wrong data, stripped of the context that makes it meaningful. Measurements taken in a controlled environment, on a single day, tell you very little about what a player will do under the conditions that actually matter.

This is a trap organizations fall into constantly. When companies invest in data infrastructure and analytics capabilities, the temptation is to measure everything and surface as much as possible. The dashboards get bigger. The reports get longer. But the decisions don’t necessarily get better—because the questions driving the analysis haven’t been refined.

Frelund’s framework is disciplined simplicity: identify the traits that actually predict success in your specific context, remove the noise, and build your model around those signal points. For a team running an outside zone rushing scheme, a guard’s lateral foot speed matters more than his 40 time. For a law firm evaluating a lateral hire, a candidate’s track record on matters similar to your client base tells you more than their law school ranking.

For HIKE2’s clients, this principle drives how we approach every data and AI engagement. We don’t start by asking what data you have—we start by asking what decisions you’re trying to improve. The architecture follows the question, not the other way around.

AI Is a Tool for Empowering Human Judgment—Not Replacing It

Perhaps the single most important idea in Frelund’s talk was her insistence that AI’s proper role is to sharpen human decision-making, not to automate it away. She was emphatic on this point, and it’s worth sitting with—because the business conversation around AI often gravitates toward efficiency, headcount, and automation, while the real competitive advantage lies somewhere more nuanced.

“You’re never taking the human out of the decision. You’re trying to empower the human to make a better decision—or at least know when they’re making a low probability decision.” — Cynthia Frelund, NFL Network Predictive Analytics Expert

Frelund illustrated this with the way AI is changing how scouts evaluate film. What used to require grinding through hours of footage—manually rewinding, re-watching, and re-evaluating play after play—can now be dramatically compressed. An analyst can focus on the specific plays and matchups that are most relevant to the player they’re evaluating, flag patterns that align or conflict with their instincts, and surface the things the human eye might have missed or the tired mind might have discounted.

But the scout still makes the call. The coach still calls the play. The general manager still picks up the phone. What changes is that they do so with a clearer picture, in less time, and with less cognitive load—which means the quality of the decision goes up even as the conditions of the decision remain challenging.

This framing is foundational to how HIKE2 designs AI and automation solutions for clients. In regulated industries like insurance and financial services, where human accountability is embedded in the process, or in professional services like law where judgment is the product, AI that tries to eliminate the human creates risk and resistance. AI that augments the human—surfacing relevant precedent faster, flagging anomalies for review, reducing the manual work that consumes expert time—creates competitive advantage and workforce buy-in. The goal is always a better-equipped decision-maker, not a removed one.

Context Is the Competitive Edge

Frelund returned repeatedly to a theme that doesn’t show up in most AI conversations: context.

A model trained on incomplete or unrepresentative data will give you confident-sounding answers that are wrong. A model that produces the right output but is applied in the wrong situation is just as dangerous. Knowing when to trust the model—and when the situation calls for something the model can’t see—is where real expertise lives.

She offered the AFC Championship game in Denver as an example: commentators attributed the outcome to white jerseys in the snow. Frelund’s read was more layered—a backup quarterback making his first start of the season, a coach playing high-risk-high-reward because the situation demanded it, a whole set of contextual factors that no model would have flagged but any experienced observer would have weighed.

This matters deeply for organizations deploying AI in client-facing or high-stakes operational settings. A credit risk model trained during a low-volatility economic period will misfire during a sudden contraction. A customer churn predictor built on historical behavior won’t catch a competitor’s aggressive new product launch. The model is always operating on past data. The human is operating in the present.

“The context, the contextualization—how the context of everything really, really matters. The model matters, how you’re modeling, what you’re selecting for, the questions you’re asking and how you’re asking the question—all of those things matter.” — Cynthia Frelund, NFL Network Predictive Analytics Expert

Frelund’s cautionary tale about the Cleveland Browns makes the point sharply. Paul DePodesta—of Moneyball fame—was brought in with the tools and the frameworks. The data was right. The models were sophisticated. But action is the step that bridges analysis and outcome, and without a decision-making culture that trusts and acts on the model’s output, the infrastructure doesn’t deliver. Having the capability and building the organizational will to use it are two very different challenges.

At HIKE2, this is why our advisory approach is human-centered by design. Deploying a new AI capability is only part of the work. We also help clients build the decision architecture—the governance, the training, the change management, and the cultural alignment—that determines whether the technology actually changes behavior and outcomes.

The World of Work Is Being Redrafted. Are You Ready?

Frelund closed with a direct pivot from the NFL to the broader workforce—and it landed like a well-placed second-round pick.

The same forces reshaping the value of positions in professional football are reshaping the composition of teams inside every organization. Roles that were specialized and stable are becoming hybrid and fluid. The traits that define high performance are changing faster than job descriptions can keep up.

She pointed to projections from McKinsey and PwC suggesting that by 2030—a date she noted feels closer than it sounds—productivity expectations and the nature of work itself will have shifted dramatically. The NFL’s version of this change is visible and fast: a pocket passer who can’t run was a premium asset a decade ago; today it’s a limitation. The same inflection is coming for knowledge work, and the organizations that are mapping it now will be in a far better position than those waiting for it to arrive.

The parallel to talent strategy is direct. Just as NFL teams have to evaluate prospects not just for what they do today but for how their skills will translate to a changing game, organizations need to evaluate their workforce—and their hiring criteria—against the game they’ll be playing in three to five years, not the one they’re playing now. What roles are becoming more valuable? What skills are being amplified by AI? Where are you drafting for the wrong position?

For HIKE2 clients, this is the workforce transformation question we help answer. Building a future-ready digital workforce isn’t just about training people to use new tools—it’s about rethinking the structure of teams, the definition of roles, and the criteria for success in a world where AI handles a growing share of the repeatable work. The organizations getting this right are the ones thinking about it now, not reacting to it later.

Watch the Full Session

Cynthia Frelund’s full presentation goes deeper on the NFL’s proprietary tracking technology, how teams build and pressure-test predictive models, the specific data challenges introduced by the transfer portal and NIL era, and what the 2026 draft class reveals about the state of the art in talent analytics. It’s a masterclass in applied data thinking—and one of the most directly relevant sessions for any leader navigating AI adoption.

Ready to Build a Better Decision Architecture?

The gap between organizations that use AI effectively and those that don’t isn’t usually a technology gap—it’s a framework gap. The teams that win are the ones who know what they’re selecting for, build models around meaningful signal rather than convenient data, and design their workflows so that human judgment and AI capability reinforce each other.

That’s the work HIKE2 does every day—with clients in financial services, insurance, the public sector, law, and high-tech. If you’re thinking about how to build the right AI foundation, develop a future-ready workforce, or make better decisions faster with the data you already have, we’d love to talk.

Contact HIKE2 to start the conversation →