Can we use artificial intelligence for Football Awards?
Can we use artificial intelligence for Football Awards?
I am not an expert in artificial intelligence. Though I know its basic potential. I.e. I know the types of things that it can do, because I have watched classes from MIT, I know how they weight the models etc.
I also know its limitations. I know it cannot do everything (as of now).
But with football awards, what are we judging?
We want to know who contributes the best from an aggregate of matches, judged in comparison to other players. That is really within the scope of PRESENT artificial intelligence. Things like chance creation, goals, defensive interventions such as tackles, clearances, saves, it goes on and on. Even more nuanced things like 'vision', I think is within scope.
The main thing I see where there might be some difficulty might be the 'intangibles'. But then I think on it, what does the fact that some would prefer Federer over Djokovic (due to the intangibles) or the fact that I prefer Brazilian football to European football due to the intangibles have to do with anything? The thing that fans love the most about Messi is not actually his end product but the intangibles, the beauty of his game. How he carves his way through teams. But that subjectivity can simply be contained to fandom. The fact that you think that Ronaldo is more mechanical is neither here nor there. Subjectivity does not need to be incorporated into awards. That is your own personal thing that you can tell your grandma about and your gateman.
All we need to know is who is doing something useful on the pitch. Does Messi get penalised for walking on the pitch (distance covered?) when he is actually scanning the pitch? Is that a useful intangible? How many datapoints do we want?
Based on everything that I know about artificial intelligence, I feel like it can do a better job than humans currently can, and that is the important point. Human beings CANNOT best artificial intelligence in deciding the best footballers in awards.
I thinks this notion that captains and coaches play the game and they know the nuances, well that only goes so far as how they can conceptualise the awards. They frame the awards from their specific vantages. It becomes incredibly subjective. Not to mention that gam theory comes into play. They gamify the awards based on what they might gain, or how they might appear.
I know a lot of people will rubbish this topic, but I honestly do not see why artificial intelligence in 2024 cannot do a better job. Of course the model will need to be very sophisticated, but we CAN build sophisticated models in 2024.
I also know its limitations. I know it cannot do everything (as of now).
But with football awards, what are we judging?
We want to know who contributes the best from an aggregate of matches, judged in comparison to other players. That is really within the scope of PRESENT artificial intelligence. Things like chance creation, goals, defensive interventions such as tackles, clearances, saves, it goes on and on. Even more nuanced things like 'vision', I think is within scope.
The main thing I see where there might be some difficulty might be the 'intangibles'. But then I think on it, what does the fact that some would prefer Federer over Djokovic (due to the intangibles) or the fact that I prefer Brazilian football to European football due to the intangibles have to do with anything? The thing that fans love the most about Messi is not actually his end product but the intangibles, the beauty of his game. How he carves his way through teams. But that subjectivity can simply be contained to fandom. The fact that you think that Ronaldo is more mechanical is neither here nor there. Subjectivity does not need to be incorporated into awards. That is your own personal thing that you can tell your grandma about and your gateman.
All we need to know is who is doing something useful on the pitch. Does Messi get penalised for walking on the pitch (distance covered?) when he is actually scanning the pitch? Is that a useful intangible? How many datapoints do we want?
Based on everything that I know about artificial intelligence, I feel like it can do a better job than humans currently can, and that is the important point. Human beings CANNOT best artificial intelligence in deciding the best footballers in awards.
I thinks this notion that captains and coaches play the game and they know the nuances, well that only goes so far as how they can conceptualise the awards. They frame the awards from their specific vantages. It becomes incredibly subjective. Not to mention that gam theory comes into play. They gamify the awards based on what they might gain, or how they might appear.
I know a lot of people will rubbish this topic, but I honestly do not see why artificial intelligence in 2024 cannot do a better job. Of course the model will need to be very sophisticated, but we CAN build sophisticated models in 2024.
Buhari, whose two terms thankfully ground to a constitutional halt in May. (One thing both democracies have going for them is that their leaders, however bad, have only two terms to swing the wrecking ball.) Under Buhari, growth per head also plunged to 0. An economic agenda drawn from the dusty pages of a 1970s protectionist handbook failed to do the trick. Despite Buhari’s promise to tame terrorism and criminality, violence flourished. Despite his reputation for probity, corruption swirled. FT
Re: Can we use artificial intelligence for Football Awards?
The other thing about artificial intelligence is that the model should learn from data, supervision will be the least important aspect. What types of contributions actually impact games the most. This we do not actually know perfectly. No football coach, player, fan knows this perfectly. We have rough ideas on impact positions. We think the #10 for example is the engine, we think that we live and die by the striker etc. but we do not quantify these things.
One of the problems with football awards is we just tally goals and assists. Artificial intelligence can and will go deeper to understand the impact before the assist, the contributions that prevent the goal and how important they are specifically.
We may be surprised with the results.
One of the problems with football awards is we just tally goals and assists. Artificial intelligence can and will go deeper to understand the impact before the assist, the contributions that prevent the goal and how important they are specifically.
We may be surprised with the results.
Buhari, whose two terms thankfully ground to a constitutional halt in May. (One thing both democracies have going for them is that their leaders, however bad, have only two terms to swing the wrecking ball.) Under Buhari, growth per head also plunged to 0. An economic agenda drawn from the dusty pages of a 1970s protectionist handbook failed to do the trick. Despite Buhari’s promise to tame terrorism and criminality, violence flourished. Despite his reputation for probity, corruption swirled. FT