Santos Laguna's AI Scouting System: An Audit Still Missing Three Numbers

There are two verbs in Gonzalo Pineda's statement that I read four times: "explorar" and "desarrollar." To explore. To develop. When a head coach talks about technology in the continuous tense, he is usually describing a construction site, not a building that has been handed over. That was the first detail I wrote in my notebook.

RÉCORD published an exclusive report: Santos Laguna is building a scouting system supported by Analytics and artificial intelligence, covering both the first team and Fuerzas Básicas, the club's youth academy network. Pineda, the head coach of Los Guerreros, is the man explaining the project. He says Analytics serves as a filter, narrowing the list before scouts travel to watch players in person. He also names two people building the network: Omar Tapia and Andrés Bejarano. The expanded scope covers Mexico and the United States.

Three names, two countries, one system. And not a single number.

Santos Laguna is one of the clubs with a genuine development tradition in northern Mexico. Its Fuerzas Básicas has produced players who went on to play in Liga MX and for the national team. But the scouting model of Mexican football in general still runs the old way: the human eye, relationships, and a notebook. A scout attends a local league, takes notes, files a report, and the decision rests on an impression plus the credibility of the person who filed it.

The problem with that model is that it leaves no audit trail. When a signing fails, nobody can trace why that name entered the list. Every contract is an investigation. Every signature is a clue. Without records there is no investigation, and without investigation every mistake gets renamed a "learning experience."

The financial context makes this project more interesting still. Liga MX does not have MLS money, let alone European money. Mexican clubs generally choose two paths: buy proven South American players at moderate cost, or push academy graduates through. The third path, using data to find undervalued players, has appeared only sporadically.

The second notable point is geography. Pineda's mention of the United States is not a minor detail. Every year a large pool of Mexican-heritage players are born and raised in the US, developed inside MLS academy systems, and sit outside the field of vision of traditional Mexican scouts. This group is routinely undervalued for two reasons: they do not appear in Mexican youth competitions, and they have no agent inflating prices in the domestic market. That is empty ground, and empty ground is always where data is worth the most.

Three layers. That is how I read this system, and also how I test it.

The first layer is the filter. Analytics and AI sweep raw data, minutes played, position, action metrics, physical data, to cut thousands of names down to a few dozen. This is the cheapest part and the easiest part to talk about. A club can buy data access without hiring a single extra person. A player's true value is not in the contract, but in the numbers nobody bothered to count — and the filter exists to find those numbers again.

The second layer is verification. Pineda is explicit: Analytics filters, the person watching live decides. Methodologically, this is the part I rate highly. A purely algorithmic model in youth football goes wrong easily, because historical youth data is thin and distorted. A 17-year-old with 800 minutes in a lower division does not generate enough sample for a model to conclude anything with certainty. Keeping a human as the final arbiter is the right way to reduce risk in principle.

The third layer is the pipeline. Fuerzas Básicas sits inside the project's scope, and U-19 and U-21 players are already under constant observation by the first-team staff. Pineda says the club currently has great talents. If true, this is the fastest-measurable part: how many academy players get promoted to the first team each season, how many minutes they actually play, and what they sell for later.

The fourth layer sits outside the model but inside the project: geography. Expanding the network into the United States means more scouts, more travel costs, and more paperwork regarding nationality and academy contracts. This is the most expensive layer and the least discussed.

Now the audit.

Based on my experience watching Liga MX matches, I always ask four numbers before believing any technology project: budget, data vendor, timeline, and performance metric. Across everything published, all four are absent. No vendor named. No budget. No timeline. No KPI.

That absence is itself data. A project already running has a vendor contract, invoices, and a person initialing every page. A project still in the "exploration" phase has a press release. Numbers do not lie, but the people supplying the numbers do. And in this case, the people supplying the numbers have not even been named.

That does not mean the project is theatre. It means there is nothing yet to verify.

The reasonable part of the sceptics' case is simple. "AI" in football, in most cases, is a trade name for a database plus a spreadsheet plus a few statistical models. Any club can stick that label on its analysis room. When Santos Laguna declines to publish a vendor and a budget, people are entitled to suspect this is a public relations release wrapped in technological vocabulary.

But the reasonable part of the supporters' case is just as strong. A data filter costs far less than a bad signing. If it removes even two unsuitable names out of ten per transfer window, it has paid for itself. Keeping scouts in the final decision seat shows the club understands the limits of the model it is building.

The real risk of this project lies elsewhere. It is not that the algorithm fails. It is that the algorithm becomes a shield. When a signing flops, the club can say the model picked him. Responsibility shifts from a named person to an unnamed system. Nobody gets asked. Nobody gets reviewed. And the scouting report, the very document that should be archived, becomes a data file nobody opens.

The second risk is single-point dependency. Pineda is the only public voice on the project. No sporting director is named, no head of analysis is named. If he leaves the club, the project leaves with him. A scouting system cannot survive on one coach's reputation.

The metric is not how many players the AI finds. That is the prettiest and most meaningless number in any report. The real metric is: how many of those players get promoted to the first team, how many minutes they play, and how much they sell for. Three numbers. No more.

I do not write from emotion. I write from minutes, bank statements, and the things people try to hide. Santos Laguna's record currently contains three names and two countries. It will need three more numbers before it becomes a file worth reading.