Related Posts
More Posts
Additional Posts in Data & Analytics Consultants
New to Fishbowl?
Download the Fishbowl app to
unlock all discussions on Fishbowl.
unlock all discussions on Fishbowl.
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.
Download the Fishbowl app to unlock all discussions on Fishbowl.
Copy and paste embed code on your site

Scan your QR code to download
Fishbowl app on your mobile

You have my attention…
Also, YNWA!!!
Haha…how you feeling now, Man U fan?
Tell me more...
DM me
Coach
I really wanted to see if we could use image recognition to help scouting. I wonder if it would be possible to get a player’s tendencies (think of football manager attributes) through an algorithm
That would be helpful but I think exists. For example, a heat map of where players touch the ball and percentage of completed passes, ball retention, etc in specific areas and compare that to a club manager’s desired style of playing.
For example, Pellegrini likes to play with overlapping fullbacks, but for that to be effective you’ll need wingers who don’t lose the ball when they receive it near the corners of the box because they enjoy cutting inside or prefer playing the ball through channels to the striker. Ideally, in addition to those traits above, he’ll want a winger that 1. has a trait (shooting from range, incisive cross, etc) that forces the opponent fullback to play close otherwise they can lay off and wait to cover the overlapping fullback and 2. not always look for the ‘magic’ ball since they’ll look to play the striker first instead of the space his fullback runs into. All of these stats can be tracked and give you an idea of how a player can perform in a given system.
I’m in
Can you please DM me?
Interested
DM me
Love me some footy, what’s on your mind
Footystats 😂
@All - At this point, I only player and match related data. Recently the FIFA 22 player ratings were abysmal and I thought if I can find any correlation. Thought of doing regression, will be starting this weekend or may be next. Then will take it up a notch using predefined ML models in SPSS modeler.
Mostly focused on PL, Bundesliga, Serie A and La Liga. And may be PSG just. 😂
Open to other ideas.
Think the idea is really fun but you have to consider that FIFA is a video game first, and soccer related second. Keeping that in mind, the first priority is to adjust stats to make the game playable which may not be reflective of reality. Example below:
1. Pace: Defenders have a significant bump in pace than their real life counterparts. This is to limit the effectiveness of spamming long balls to pacey forwards. While it’s still an effective strategy, it would be broken further if defenders didn’t have the bump in pace as you could put guys like Deandre Yedlin at striker and cheese all game.
FM stats would be much better as they have a legitimate scouting network that evaluates players where FIFA seemingly goes by “how good do we think this guy is and how do we balance the stats to reflect that in the way we built our game”.