Jump to content

Rangers using Artificial intelligence to win the league


Junior Soprano

Recommended Posts

11 minutes ago, Redwhiteandblue said:

Can you not copy and paste? I don't want to click the link.

STEP AHEAD Rangers ‘using AI technology’ to predict when players could suffer injuries

Rangers have kept one step ahead of injuries this season by using AI technology to keep their players' fitness at its optimum.

Gers have partnered with Zone7, a company who use Artificial Intelligence and Machine Learning, to forecast injury-risks within the club.

The brand have been working with Spanish side Getafe since 2017-18 who have noted a vast reduction in the amount of muscular injuries since then.

Zone7 were founded in 2017 by Tal Brown and Eyal Eliakim who have both served in Israel’s defence intelligence community.

Gers have been fortunate with injuries this season with boss Steven Gerrard having a full squad to pick from for the majority of the campaign.

And Brown is delighted to have the Light Blues on board with the company now working with 40 clubs across the globe.

The Ibrox side have had no real long-term injuries this season

The Zone7 co-founder and chief executive told the Athletic: “We basically put out some thoughts on our website and our blog around injury risk analysis and what’s the sweet spot in terms of managing players and risk.

“This led to about a dozen teams reaching out to us and I’d say that the common ground was that they felt the typical weekly cadence of training and adjusting of workloads needed revision during this COVID time.

“We were fortunate that Rangers were one of them. John Colquhoun (the former Hearts and Scotland forward) is an investor in Zone7 as he knew some people to help us get there.

“Going back two years, there was a huge pushback about applying AI to medical or performance data.

“It was perceived to be a project that had no chance of working but we were able to demonstrate early on that there was potential and the first investor was Jordi Cruyff (the former Manchester United and Barcelona midfielder, and son of football legend Johan), which led to a bunch of pilots.

"There’s thousands of parameters coming in every day for 25 individuals.

“Our approach is to use a machine that can process unlimited amounts of data and be able to apply complex equations which identify patterns that have been found in other data sets preceding injuries.

“Look at chess. People study it for a long time, they read the books, learn the plays and some have an incredible intuition for the game. But technology has become better and better at creating chess programmes and it now has the opportunity to analyse every chess game ever recorded by mankind and apply these learnings into a large dataset and generate these patterns and strategies from it.

“Over time, we’ve seen that a chess programme can outperform pretty much everyone at a human level. The reason is that AI nowadays is able to harness this data in a powerful way. If you apply that to what we do, we are not using one team’s data. We baseline that data using the 10,000 players and 150 to 200 million hours of football performance we’ve recorded and analysed.

“That covers a lot more than teams can get access to and allows us to apply much more sophisticated AI tools or deep learning tools.

“For someone like Jordan (Milsom, head of performance at Rangers) to learn to trust an interpretation tool, they need to feel that it’s accurate, whether that be, ‘Player X is at risk’ or, ‘We should tweak the training’. So we offer the chance to analyse their historical data.

“We can go and look at their data from last season and produce a report saying, ‘This is what we could have advised if we were active’. It’s not as strong as a live trial but it’s pretty strong proof. We have about 15 of those case studies so it’s not just about team X having high detection rates for injuries. Our technology is able to detect 70 to 75 per cent of injury incidents going back a year or two.

“That leads to what we call an activation, where the engine starts to run live and surface the insights as to who is a risk and who may need adjustments to their workload — what the optimal recovery and workload protocol is for each player every day.

“That’s what we provide but it’s important to note that Rangers have a fantastic staff and have an existing apparatus so we were able to plug into that. They’re phenomenal to work with; very innovative and very data-oriented.

“An organisation like Rangers has an existing apparatus. They are not blind but it requires analysis and interpretation. At times, we help clients triangulate an existing suspicion and sometimes we help them cover a new blind spot.

“It’s not always about resting players, it’s about being aware of the risk and being able to work that into your decision making. Sometimes we suggest limiting game minutes, sometimes we say player x should do 80 per cent of running, 75 per cent of sprinting and 120 per cent of something else. That’s what we help calibrate so that they can be where their team wants them to be on a Saturday.

“We’ve found that a one-week detection window is appropriate for a football environment because it gives you enough time before your next game to intervene. In baseball (where games are more frequent), it’s five days. Could we look at a 30-day predictive window? We could, but it’s less actionable.

“I’m not saying this is the only tool used by any team or is 100 per cent every time. This is about people like Jordan being able to intelligently integrate the data they have in front of them with an interpretation tool that offers accurate, consistent, risk analysis and being able to bring it to the attention of the manager effectively.

“Generally speaking, the more clients are using the product the more raw data we have to validate it and improve it.

“The more data we have, we can do things like predicting the injury severity more accurately: there’s risk, but is it risk for a five-day injury or a 50-day injury?

“We’re not a diagnostic tool but if we can compare a player to 10,000 players we can, with a high degree of certainty, issue an alert — and not just an alert, a hamstring alert.

“The other thing is that the more data we have from a specific team, the more the system understands their coaching schedule, philosophy and individual players. After three years, we’ll have analysed over 500 training sessions so we know what kind of impact their drills have on their players.”

 

Link to post
Share on other sites

38 minutes ago, Moody Blue Legend said:

Aye I should read these articles in full tbh.

Just carry on like I do m8, read half a paragraph and then burst in with an opinion which is either right or wrong, I'm usually wrong which doesn't bother me at all, when I'm right though the entire internet will hear about it. 

Link to post
Share on other sites

1 minute ago, Scottywellhousetb said:

Just carry on like I do m8, read half a paragraph and then burst in with an opinion which is either right or wrong, I'm usually wrong which doesn't bother me at all, when I'm right though the entire internet will hear about it. 

I only read the bit under one of the pictures, but that's all I need to know I'm right Scotty.  

Link to post
Share on other sites

2 hours ago, plumbGER said:

Easy, remarkably if a player has an off game for the beggars they seem to have injuries for months after it.

Pretty predictable imo.

Spend £5M on a shit goalie, next thing you know hes injured and last years 3rd choice gets the gloves.

I know people think sports science is a load of pish, but times moved on and im glad were doing everything we can to manage players fitness/injury prevention.

Link to post
Share on other sites

Archived

This topic is now archived and is closed to further replies.

  • Upcoming Events

    • 21 April 2024 14:00 Until 16:00
      0  
      Rangers v Hearts
      Hampden Park
      Scottish Cup
×
×
  • Create New...