How much influence are you happy for AI to have?

[metalblade]

To start with I would think AI data will be used to find players who we can develop I doubt it will be much use for players we want to go in the 1st team this summer. Surely we aren't going to sack every scout we have looking for players. I would think we are using AI to build a data base of players who could do a job for us in the future, it will take time we have only taken a few steps down a new pathway, of course a lot of fans fear change. I'm looking at it as a tool to help our scouting system nothing more as yet, there will always be the human element to have the last say on if we sign someone or not.

 

[Tony_Kaufman]

It's based on something I've heard, which I can't go into, though I'm certain of its' validity.

But I'm not asking for you to believe me - you can just take it as a theoretical position against which we can discuss how far are we happy to let AI make decisions.

AI is not making decisions, humans are.

Imagine you have an excel spreadsheet with a list of strikers with their age in one column, their number of games played last season in another, assists, goals, contract expiry etc.

You set a filter and sort the data by goals scored. Then you filter the data by age, then contract expiry.

You look at the top three players and decide that you might consider signing one of them.

This is what AI will do for us. Humans will still look at the results and make the decision. They will also tell AI what they want it to look for.

AI is not deciding anything, it is helping. It's a filter, nothing more.

 

[PrestwichBlade]

I think in all honesty, I'm fairly comfortable with it being used. The owners have been fairly upfront from the start about using it more smartly, so of course they are going to push it more with the management/coaching team (I assume this is why they have gone for someone like Sellés?). I guess it will be done intelligently and through collaboration, with ultimately "human decisions" made, and if that's the case then they can crack on for me. There will be duffs I'm sure, but then when haven't there been duff signings?? The more it's used, the smarter it will become. Full disclosure though, my job is in 'data'!

 

[alcoblade]

AI is not making decisions, humans are.

Imagine you have an excel spreadsheet with a list of strikers with their age in one column, their number of games played last season in another, assists, goals, contract expiry etc.

You set a filter and sort the data by goals scored. Then you filter the data by age, then contract expiry.

You look at the top three players and decide that you might consider signing one of them.

This is what AI will do for us. Humans will still look at the results and make the decision. They will also tell AI what they want it to look for.

AI is not deciding anything, it is helping. It's a filter, nothing more.

In short, you're wrong

 

[HVR]

AI is not making decisions, humans are.

Imagine you have an excel spreadsheet with a list of strikers with their age in one column, their number of games played last season in another, assists, goals, contract expiry etc.

You set a filter and sort the data by goals scored. Then you filter the data by age, then contract expiry.

You look at the top three players and decide that you might consider signing one of them.

This is what AI will do for us. Humans will still look at the results and make the decision. They will also tell AI what they want it to look for.

AI is not deciding anything, it is helping. It's a filter, nothing more.

Yes, and a good example - but obviously there may come a point where AI learns a little bit more about how we set these filters, and decides the questions we are asking are not relevant - and starts to hide results it would have previously shown us. Due to the size and complexity of the data, we would have no idea whether the results are based on our queries - or results AI will give regardless of filters we use.

For example in your excel spreadsheet - if PLAYER A comes out on top - we could quickly check "Player A is 35 years old, that can't be right?" However, as the data becomes more complex, using millions of calculations - we can only trust it's the right answer, as we don't have the capacity to evaluate the results.

So we know 1+1 = 2, if a calculator states it's 3 we know it's wrong. At what point in mathematical computations do we simply trust the calculator to be right?

 

[Stretch]

This is what Chat GPT thinks...

"These are insightful and complex questions, especially in the context of how football clubs like Sheffield United are increasingly integrating AI into decision-making. Let's break each part down with a balanced view:

1. How much say should the new manager have?​

The manager must have significant say, especially regarding tactics, dressing room dynamics, and player morale. While AI can provide data-driven insights, it cannot replace firsthand player management or leadership.

However, the club likely wants to avoid returning to a scenario where a manager completely disregards the AI system. So the best model is collaborative:

• The manager should be required to engage with the data,
• But not dictated by it.
  For instance, he might need to justify decisions that diverge sharply from model recommendations, rather than have his autonomy overridden.

​

2. Should management be forced to pick Jefferson Cacares for the under-21s if AI recommends it?​

No—forced selection undermines coaching accountability.
However, AI suggesting Cacares should play might be used to prompt conversations within the development staff, such as:

Why is the model rating him so highly?

Are there training or off-field issues coaches are seeing that the model isn’t?

Ultimately, development squads are about nurturing potential, so it’s reasonable to expect more AI experimentation here—but it should inform, not dictate.


3. How much should AI influence first-team selection?

AI should inform first-team selection, not control it.
It can be extremely useful for:

Fatigue prediction

Injury risk

Tactical matchup analysis

Tracking off-ball performance

But football is a high-context sport. A manager might choose an “in-form” player over one the model rates more highly for intangible reasons: dressing room harmony, pressing intensity, or even a specific psychological edge.

So: AI should be part of the process, not the arbiter.


4. Should AI be used to justify much larger fees for players than the manager would himself sanction?

Here, it gets trickier. Clubs need to balance:

Long-term squad value (where AI models often excel),

Against the short-term competitive goals the manager is responsible for.

If a manager wants to spend £5m on a player but the AI says another at £10m has far greater resale and performance upside, it becomes a board-level decision—not just the manager's call."

 

[Davalon]

Does anyone actually have a Scooby how it's going to be used in detail?

Until we see some clear evidence over a notable period of time, it just feels like a lot of noise and people just sticking by their own agenda.

The three lads signed so far aren't conclusive evidence of anything yet.

 

[Sean Thornton]

There is certainly a lack of understanding of what is being used, so I’d guess that would be the first point to consider

My simple understanding is that the AI Model as you call it is a data based system to initially scout and potentially recruit players. Using attribute data which is available for every professional player can then allow you to tailor your recruitment based on the attributes you want for your players.

The AI part is programming a system to then find those players

The use of data in games and in preparation for games is not new. I think the first time we were wearing the catapult gps vests was possibly back under Clough, but Adkins was the first to make reference to it and using the data when we played Man U in the fa cup.

Adding, let’s say automation to a process by getting a computer to analyse data that the user requests is perhaps the part which is new and speeds up the process

As for recruitment and fees, I don’t think the manager has had control of the financial budget for many years when it comes to buying players. Generally the manager selects the players he wants and puts it to a committee, it’s then a group decision who gets bought

What’s this? We’ll have no common sense here, this is a forum for “local” people…..

 

[Bartleby]

So do you think, for example, Jefferson Caccares should be chosen based on his ability as seen by the management or by the fact that AI tells them he's a good choice?

We can all agree that we should be using data, and AI, to inform decision-making. The question is by how much?

I'd have thought it is more the discovery of affordable talent that we would never otherwise locate that AI is being used for. If they then don't perform in training I can't see AI telling them to play him.

I really think we should give it a go.

 

[Tony_Kaufman]

What’s this? We’ll have no common sense here, this is a forum for “local” people…..

 

[tincan]

The core remains who records the data that determines the performance of the players, their fitness, coaching, lifestyle, social background, mental welbeing and the cost of the "AI" model delivering the methods?
How do they check it remains fit for purpose, also when it is or isn't working . If we lose do we sack them or the Model?

 

[alcoblade]

I'd have thought it is more the discovery of affordable talent that we would never otherwise locate that AI is being used for. If they then don't perform in training I can't see AI telling them to play him.

I really think we should give it a go.

We are doing, and it's telling them to play him.

It's got to be very tough for the coaches to stand aside, and the players pushed out of position by someone who right now they all believe is inferior to them.

It's going to take some mindset shift, but that's the way it's going.

 

[The Crab]

So given that our old manager was ousted for a perceived lack of embracement of modern ways - particularly the new AI model - how much say should the new manager have?

For example, should management be forced to pick Jefferson Cacares for the under 21's?

How much should AI influence team selection for the first team?

Should AI be used to justify much larger fees for players than the manager would himself sanction?

A T-1000 replacing C3P0 at the back?

 

[jt64]

It is one of many tools that are available and should be utilised. Not utilising data to its fullest potential is setting money (and results) on fire.

 

[jt64]

A T-1000 replacing C3P0 at the back?

Hoof it up to ED-209!

 

[Ball_Sup (Phil)]

The Academy do NOT pick the best side for each U18/U21 game. We do not, necessarily, pick a side to win a particular game. An obvious example of that is when we drop a first teamer down to recover from injury, or coast through 45 to get something in the legs.

More generally, Academy sides are often picked based on the development needs of individual players. The point is, you often pick an Academy side based on wider, long term interests, rather than narrower, points in the bag interests.

I would hope that any recruitment model (AI or otherwise) would be able to identify POTENTIAL in an Academy Recruit, AND what development is necessary to realise that potential.

I find the situation with Jeff interesting. In old money, before recruitment, no one had seen him play the English game, the English way, against English teams. Shape discipline, intensity, tracking, finding space, etc.

You would want the recruitment process to identify he has the ability to develop those "English" skills, in order to progress.

I've seen Jeff a good few times. I've written about what I feel were his clear weaknesses, and where he needed to develop. And I've seen clear improvement in a "fast learner" way, as he's played more games.

I'm doing nothing more than musing on an ongoing AI role in Academy team selection, and in Shirecliffe Development Plans for individual players.

 

[Bartleby]

We are doing, and it's telling them to play him.

It's got to be very tough for the coaches to stand aside, and the players pushed out of position by someone who right now they all believe is inferior to them.

It's going to take some mindset shift, but that's the way it's going.

Doesn't appear to have done Brentford any harm

 

[PrestwichBlade]

We are doing, and it's telling them to play him.

It's got to be very tough for the coaches to stand aside, and the players pushed out of position by someone who right now they all believe is inferior to them.

It's going to take some mindset shift, but that's the way it's going.

And that's fine. It's a mindset change that's needed if it's any chance of working. Perhaps U21s is seen as a good place to test things out a bit? Is so, then meh, whatever

 

[GymClassHero]

It isn't. But the owners may well insist you carry out it's instructions to a level most people would be very surprised.

I'm very worried that the head coach is far from in charge now.

I think having a manager as opposed to a head coach has been to our detriment and it's probably the reason why we spunked so much money on poor signings.

I'm all for having a dedicated team and system to seek out the right players, as long as the head coach has full reign of teams selection and training etc.

 

[alcoblade]

Doesn't appear to have done Brentford any harm

Yep, everyone wants to be Brentford.

 

[alcoblade]

And that's fine. It's a mindset change that's needed if it's any chance of working. Perhaps U21s is seen as a good place to test things out a bit? Is so, then meh, whatever

I'm sure that'll be the current thinking regarding the U21's, and it's very reasonable.

I believe suggestions have been made for first team selection, and we're always rejected. I'm not so sure that'll be the case in future.

 

[Carlton Blade]

We are doing, and it's telling them to play him.

It's got to be very tough for the coaches to stand aside, and the players pushed out of position by someone who right now they all believe is inferior to them.

It's going to take some mindset shift, but that's the way it's going.

If the club believe Caceres has a better chance of being a regular 1st team player with the right development, but he's behind a different academy player now who they perceive to have a lower 'ceiling', of course he shoukd play.

I'm pretty sure there's nobody saying "we paid a fee so he plays in this game where the result is a lot less important than the development"

 

[alcoblade]

I think having a manager as opposed to a head coach has been to our detriment and it's probably the reason why we spunked so much money on poor signings.

I'm all for having a dedicated team and system to seek out the right players, as long as the head coach has full reign of teams selection and training etc.

Regarding the first paragraph, agreed. But the model failed on Tom Cannon, so isn't infallible.

Regarding the second paragraph. I think you'll be disappointed, though how much of it is ever disclosed we'll have to see.

 

[PrestwichBlade]

I'm sure that'll be the current thinking regarding the U21's, and it's very reasonable

I believe suggestions have been made for first team selection, and we're always rejected. I'm not so sure that'll be the case in future.

I'm sure discussions around future potential use of the system would have been had, in fact it would have been strange had they not. I really suspect though it will be more along the lines of "data is telling us such and such a player should start in this scenario, what we thinking?" Rather than "data is telling us Mr Selles, this player should be starting next week, you must play him".

 

[Tony_Kaufman]

Hoof it up to ED-209!

 

[The Crab]

Hoof it up to ED-209!

It’ll be like having Darius Henderson back…

 

[The Crab]

Regarding the first paragraph, agreed. But the model failed on Tom Cannon, so isn't infallible.

Regarding the second paragraph. I think you'll be disappointed, though how much of it is ever disclosed we'll have to see.

In line with the modern AI way of doing things, just turn Tom Cannon off then on again…

 

[Edgar Allan Poo]

It’ll be like having Darius Henderson back…

Darius's ability to back into defenders a give away a free kick, get himself booked or both was unrivalled as I remember.

 

[Ball_Sup (Phil)]

I do think some thought might need to be given to any impact on Academy Recruitment in the round.

At the moment, I feel we have a good sell to both local 10 year olds & their parents, Arblaster for eg, AND national 17 year olds searching for a second chance Academy, Peck for eg.

That sell is - a clear progession pathway and a track record of getting players who develop, and are good enough, into the First Team.

We wouldn't want a future Peck to say - why would I choose you? Isn't my path going to be crowded with some random player from Africa, currently playing in Bulgaria, I've heard you're more interested in turning a people trafficking profit, than in developing players into the EFL.

 

[alcoblade]

I do think some thought might need to be given to any impact on Academy Recruitment in the round.

At the moment, I feel we have a good sell to both local 10 year olds & their parents, Arblaster for eg, AND national 17 year olds searching for a second chance Academy, Peck for eg.

That sell is - a clear progession pathway and a track record of getting players who develop, and are good enough, into the First Team.

We wouldn't want a future Peck to say - why would I choose you? Isn't my path going to be crowded with some random player from Africa, currently playing in Bulgaria, I've heard you're more interested in turning a people trafficking profit, than in developing players into the EFL.

Yep, that's a big concern for me.