An A.I. idea for talent I.D.
Problem: Too many players with unique athletic potential spend their lives 'undiscovered' across the world (for the purpose of this article I will stick to the theme of football which I am most interested in).
One solution I appreciate: The Tonnser App which allows football players to easily advertise their stats to prospective scouts and coaches. In Tonnser's own words "Europe's leading football performance app is connecting more than 800,000 players with coaches and scouts."
Two issues I recognize: Stats don't tell the 'full story' and the App is presently only available in certain countries. I believe it would need to be available in all countries to make a dent in the problem I've presented.
My idea: To further the ability a player has to promote themselves by creating an App which scans the player (likely moving with and without the ball) and compares the player's 'movement signature' combined with body type with those of professional players and identifies their pro football player 'doppelganger' (if they have one).
The so called "eye test" seems to be losing relevence with the popularity of analytics. I understand to a degree, however I think that the 'eye test' from a reliable "eye" is becoming underrated. Imagine that someday an Inuit player in Northern Quebec has a Tonsser profile that allows him to show professional coach's that he's leading the local U18 elite Futsal scoring charts by a wide margin. Due to the remoteness of the player's location I don't think this stat would 'catch the eye' of a prospective coach, however what if the coach's were also informed that he had a body type like Ronaldo Luís Nazário de Lima and was one of very few players in the world with a similar movement signature?
Concerns and conclusion: The possiblity of the idea becoming part of the solution hinges first on the A.I. technology which seems far off (despite reading about scientists in Finland using machine learning to study movement signature) and secondly whether having a doppelganger in this mode could become relevant to scouts and coaches in the real world. My final concern would be that if it did work there would be some players with great potential but due to the uniqueness of their movment signature (Erling Haaland comes to mind) a doppelganger is not registered and this reduces their prospects. I wouldn't want players to become
Nonetheless I have great belief in a players body type and movement signature as highly relevant information, especially considering this has been the most important formula for me as a scout. I have emperical evidence due to my experience with talent identification but also as an eventor at an analytics company where we watched pro matches in slow motion. I came to realize that the teams of higher quality had many more players of unique body type and movement than those of lower quality, which often had players with a stocky body type and generic movements.
Building something exciting at the intersection of football, AI & gaming ????
3 年Fascinating piece Andrew. Aside from some of the issues you mentioned, you would need the opt-in of virtually every top player. As far as I can tell, the players (or their clubs) agreeing to have their movement signature measured is the only way to generate the exhaustive database needed to make the doppelg?nger scouting model viable - otherwise a young prospect with say Messi's movement sig would not be recognised as being special if Messi himself didn't opt-in. Your point about Haaland is an excellent one and that would be my worry as well. Nevertheless, it's at the minimum a fascinating thought experiment.
I work with professional players, coaches, and CEOs because they also need to understand life and how life creates success
3 年Andrew Jefferson that is why I teach payers about the mind. Any player 6 months could play for the team of their choosing by understanding how to create with their minds. I know for most this sounds like balderdash. But I have done it myself on a personal basis and to players as well that were struggling in no leagues to become the best player in the team in the top league. Every player can learn how to create with the mind---not mindset work----Creating, but they have to be willing to look at something that is not excepted that wide because of the old-fashioned views of things that most people do not understand.
Head of Player Recruitment & Talent ID. Previously at Manchester Utd, Brighton & Hove Albion, Liverpool FC, Charlton Ath, Millwall FC.
3 年Good article. But I have the key questions: we all spot good players or potentials correct and some have longer experiences than others. Does the App allow all the other relevant factors and information?: Education, social skills, life skills, mentality, preparation, ability, availability, foundations, background, family history, culture, environment or is it just the best clips, current clips and footage or just the best bits?. I have not seen the app so excuse me if this is all covered and equally do you think its relevant? Does a coach and recruiter see and look for the same thing, or is it win win for the coach and development for the scout/recruiter? Mark.