Football continues to evolve thanks to new technology and the use of data has emerged over the last decade giving clubs more power in player signings and squad selections.
Now, computer scientists England’s Loughborough University have developed Al algorithm enabling football coaches to better identify potential players that could be top players. According to Dr. Baihai Li, the head of the project, the technology may allow a club to identify the right recruit through data.
The new technology makes use of computer vision, deep learning, and AI to:
1. Detect a player’s body pose and limbs to identify actions during play
2. Track players to get individual performance data throughout a match
3. Camera stitch (using two cameras to record half of the football field to get a full picture of the performance
At the moment, player performance analysis is a long, time-consuming process in which an individual watches many hours of video recordings of a player’s matches. Due to a reliance on personal judgement, the process can be filled with errors and bias.
The new project from Loughborough uses a hybrid system in which human data entry is supplemented by camera-based automated methods.
It is claimed that the new system will help to get the data needed for accurate player performance analysis and talent identification. Other sports may also be able to use the technology to identify and recruit players.
Over the last decade, more and more technology has been introduced to the game of football to help teams better judge player performance. Data has been used by a number of top clubs including Liverpool and Manchester City to recruit players and to justify the transfer prices they have paid.
Loughborough has conducted a number of projects involving sports including a 2019 initiative to help ex-prisoners find work through sports.
The football technology is still under development but could be used by football clubs in the near future.