Talent.AI · chairmen and academies

A potential index, and everything it does not tell you.

Talent.AI places a player against the standards of his own division, the one above and the one below. It is a tool for projection, not valuation. This page exists because we chose to take it out of the coaching plans — and it seemed more honest to explain why than to sell it quietly.

In none of the coaching plans

What the index says

The level a player has the means to play at.

What it measures

The level a player has the means to play at

The index places a player against the standards of his current division, the one above and the one below. A positive score indicates headroom at the level above; a negative one, a profile better suited to the level below.

What it is calculated on

Instrumented matches, not the whole season

It is only calculated on matches where the platform has the detail: exact positions, actual minutes, validated events. In the example report, 14 matches out of 23 — and the page says so.

How often

Recalculated after every match

Progress shows before it appears in the results. That is what makes the index useful to an academy: spotting a movement, not recording a trophy cabinet.

What it does not say

Four limits, written down before you find them.

A potential index sells easily and turns on you quickly. Here is what this one does not cover — and what, in our view, has to stay your judgement.

01

It is not a transfer value

The index says nothing about a price, a market or a capital gain. It factors in neither demand, nor contract, nor age as a financial variable. Any monetary reading is an interpretation that we do not supply and do not endorse.

02

It is not a career prediction

A high index at 21 says the player is producing above the standards of his division in the matches measured. It does not say he will play at the level above: injury, personal trajectory, the club that takes him and plain luck weigh more than the data.

03

It cannot see anything off the pitch

Attendance, behaviour in the group, schooling, family circumstances: none of it enters the calculation. For an academy player, that is often what decides it.

04

It depends on events entered by a human

Goals, passes and assists are validated by the staff. An index that depends on data entry can be influenced by that data entry. We say so here because a coach put it to us, and he was right.

The method

25 criteria, worked through in a CIFRE doctorate.

A CIFRE doctorate carried out with the URePSSS laboratory at the University of Lille is working through the 25 criteria that feed the indices. The arrangement is formally agreed and the protocol documented. The work is ongoing: we claim no established validation, and this page presents no success rate.

What we cannot prove yet

An index without a denominator is worth nothing.

Anyone can show you three players they spotted correctly. The question that matters is the opposite one: out of how many players tracked, how many did it get wrong? Until those three numbers are published, Talent.AI has to be read as an aid to spotting players, not as a reliable measurement. They are missing today, and we would rather write that down.

The cohortHow many players tracked, across which divisions and over what period.figure to be supplied
False positivesHow many players projected to the level above never played there.figure to be supplied
False negativesHow many players stepped up without the index seeing it coming.figure to be supplied

One report, read

Six players at a French regional club, 2025-26 season.

Taken from the anonymised end-of-season report published on this site. Indices calculated on the season's 14 instrumented matches, not on all 23 matchdays. R1 is the top French regional division; N3 the national tier above it. Names have been changed; the values are real.

PlayerPositionAgeIndexProjection
LucasCentre-forward21+0.641R1 → N3
RomainCentral midfielder30+0.541R1 → N3
JulienRight winger30+0.513R1 → N3
HugoCentre-back21+0.508R1 → N3
NathanCentre-back21+0.473R1 → N3
MaximeRight-back22+0.351R1 → N3

← Drag the table to see every column

These six are the players whose index passes +0.35 in this squad. What the table does not show, and which counts just as much: the other fourteen players in the group, whose index places them at their current level or below. Read the full report →

Why it is not sold to coaches

Because the coach's interest is not the index's interest.

We pulled it from the coaching plans after user testing.

A head coach wants to keep his senior players. Handing him a score that identifies the players ready for the level above means selling him a tool that works against him — and that is exactly what coaches told us: “that is an agent's argument, not a coach's”.

The index keeps its value, but for other people: a chairman preparing a close season, an academy deciding who moves up an age group. That is who this page is for, and nobody else. What the coach gets inside MyMister is the SFI — what the player produced — and nothing that invites him to let go of anyone.

Do you run a club or an academy?

Talent.AI is discussed case by case, with your data and your age groups. We would rather have a conversation than a price list, for as long as the cohort remains unpublished.