NFL, NBA and MLB: three different ways to make a signing decision with data

When football looks to American sports for analytical inspiration, it often misreads them. It treats the three majors, American football, basketball and baseball, as if they were one thing, a generic block called data sports. They are not. Each one developed its own way of making a signing decision, shaped by the rules of the game, by the economic structure of the league and by the kind of data that sport produces. Understanding the differences is more useful than celebrating the similarity, because the useful lessons live in the differences.

MLB: high repetition and statistical predictability

Baseball is the founding case, and not by chance. It is a sport of statistical nature, made of discrete and repeated events, one pitch at a time, one at-bat at a time, in enormous volume across a season of more than one hundred and sixty games. This high-repetition structure produces large samples and isolates individual contribution with a clarity few sports achieve. The greater the number of comparable events, the greater the statistical predictability, because each new data point confirms or corrects the previous reading. That is why the analytical revolution began there.

Decades of advanced statistical analysis of baseball, the field known as sabermetrics, gave franchises tools to estimate how many wins a player adds to the team and what that is worth in money. The signing decision in modern baseball rests on two complementary fronts. The first is performance projection: with so many repeated events, it is possible to anticipate, with a good margin, what a player tends to deliver. The second is price reading: assessing whether a player is being priced in line with what he produces. The baseball front office combines these two readings, identifies undervalued assets and accepts living with the unpopularity of not signing the name the fans demand. The lesson baseball offers is that repetition generates predictability, and that much of the market error comes from ignoring that predictability when assessing talent and price.

The point to watch is precisely the contrast with football. Football is less repetitive and less predictable: it has fewer isolated comparable events, more interdependence among the athletes and more relevant external variables in each match. For that reason, baseball models do not import automatically. The principles are transferable; the formulas are not. Football requires its own models, built for its structure.

NBA: signing as trajectory management

Basketball is another scenario of analysis. It is a game of continuous space, not discrete events, and for a long time it resisted quantification precisely for that reason. The turning point came with tracking technology, cameras that record the position of every player and the ball many times per second, generating spatial data that describes not only what happened, but how it happened. Where the player positions himself, what kind of shot he creates, how much he protects the paint without having to block, how much his positioning improves the team's collective performance on the court.

The particularity of the NBA decision lies in the management of career trajectory. With short rosters and long, expensive contracts, a basketball franchise decides not only about the player of today, but about who he will be three or four years from now. Minutes management, load planning and care with the development of the young athlete all enter the signing calculation. The NBA thinks of the player as an asset whose value curve rises and falls over time, with phases of appreciation and depreciation, and actively manages that curve, attentive even to the moments of peak, decline and recovery. The lesson for football is direct: signing is not photographing a player in one instant, it is assessing a trajectory, and following that trajectory after the signing is part of the decision.

NFL: the allocation problem under a cap

American football is the most collective of the three and the one whose individual sample is hardest to isolate. A single play involves eleven athletes in highly specialized roles, and the outcome of each play depends heavily on collective execution. This makes individual evaluation a complex problem and shifts the focus to building the whole.

The NFL decision is, essentially, a problem of capital allocation under a rigid constraint. The salary cap forces each franchise to distribute a fixed budget across dozens of positions, and the annual draft adds the dimension of acquiring young, cheaper talent to balance the high contracts of veterans. This is where data takes on a specific role: it helps find cheap, complementary players, functional pieces that deliver above what they cost and that sustain the roster around the stars. The American football front office looks less like a talent hunter and more like a portfolio manager, deciding where to concentrate resources, where to save and how to balance stars and squad. The lesson for club football is the most institutional of the three: signing well is not getting one isolated decision right, it is managing an entire budget coherently, balancing risk, cost and need across several windows.

What football can learn from these sports

Each of the three sports offers a practical lesson that football can adapt to its own reality.

From baseball, the notion that repetition generates predictability and that talent and price should be read together. Wherever there is comparable data in volume, performance projection gains reliability, and ignoring it means giving up valuable information.

From basketball, the idea that a signing is the assessment of a trajectory, not of an instant, with attention to the curves of appreciation and depreciation and to the athlete's career phase. Following up after the signing is part of the decision.

From American football, the understanding that the individual decision only makes sense within a strategy of allocation for the whole, with balance between stars and squad and conscious management of risk and budget.

The three, taken together, describe a mature way of deciding: treating the signing with method, price reading and horizon. Football does not need to become baseball, basketball or American football. Each sport has its own structure, and the path is not to transplant formulas, but to build models suited to the complexity of football. Well-applied data analysis makes signing decisions more efficient, without promising certainty and without losing sight of the natural uncertainty of the game.