Data driven technology for player recruitment

[cubehindthegoal]

In the sevco discussion, this got me thinking on a separate theme:

“49ers cutting edge, data driven technology for player recruitment” …

.. so, that seems to be a thing just now. Brighton and Bournemouth spring to mind for two … but one thing that nags at me is … if they ALL have versions of it, then what does that mean ? Does it all just cancel out ? We all admire how Bournemouth and Brighton have recruited, and perhaps over-achieved, and many put it down to their recruitment software systems … but surely if they are so good, then Liverpool, Arsenal, Man City etc will just buy them out or find or create better systems at some point ? And then what ?

And .. here’s a thought … when AI kicks in … what then ?

(hmm if AI grows arms and legs .., will it then replace players too ?)

Ok, last line was flippant, but the rest of it intrigues me … it’s as clear as mud who and how are the leaders in this, and who are up and coming, and does anyone actually know the full picture ? And … I keep thinking other things round this lol … are the 49ers going to be better than the rest with their system in the US ? They aren’t at the moment.

Ok, I’ll stop there. That’s info overload, sorry folks. Welcome to my brain 🙄😖

[Bridge hibs]

In the sevco discussion, this got me thinking on a separate theme:

“49ers cutting edge, data driven technology for player recruitment” …

.. so, that seems to be a thing just now. Brighton and Bournemouth spring to mind for two … but one thing that nags at me is … if they ALL have versions of it, then what does that mean ? Does it all just cancel out ? We all admire how Bournemouth and Brighton have recruited, and perhaps over-achieved, and many put it down to their recruitment software systems … but surely if they are so good, then Liverpool, Arsenal, Man City etc will just buy them out or find or create better systems at some point ? And then what ?

And .. here’s a thought … when AI kicks in … what then ?

(hmm if AI grows arms and legs .., will it then replace players too ?)

Ok, last line was flippant, but the rest of it intrigues me … it’s as clear as mud who and how are the leaders in this, and who are up and coming, and does anyone actually know the full picture ? And … I keep thinking other things round this lol … are the 49ers going to be better than the rest with their system in the US ? They aren’t at the moment.My knowledge of this is near to zero but I would think as well as having the up to date software and data you would need the very best people to operate it, thats what probably separates the wheat from the chaf.

On top of that you would have club stature and finances, that alone would boot a good few teams into touch. Tech has a massive part to play in recruitment nowadays and I think gone are the days of club scouts pacing the touchlines or sitting in stands looking to grab the next rising star.

Thats just my very limited and probable narrow minded opinion on this.

[cubehindthegoal]

My knowledge of this is near to zero but I would think as well as having the up to date software and data you would need the very best people to operate it, thats what probably separates the wheat from the chaf.

On top of that you would have club stature and finances, that alone would boot a good few teams into touch. Tech has a massive part to play in recruitment nowadays and I think gone are the days of club scouts pacing the touchlines or sitting in stands looking to grab the next rising star.

Thats just my very limited and probable narrow minded opinion on this.

It’s interesting actually who does actually gather and input the data, and does this rule out expert people/opinion ultimately. And presumably, AI may mean a camera could film a whole match, and an AI programme could gather that footage and create data that AI then processes quicker and better than people ever could … where does it end? What does it mean for “competition” in the future? More questions than answers in both my posts and also in what’s happening in football and where it goes …

[wookie70]

Im not sure it will make much difference. Graham Taylor used data pretty successfully. Get the ball in the box as most goals are scored there at a basic level. I can see how it can be used Moneyball style in Baseball as getting on base is teh equivelant of ball in box. However I'm struggling with such a dynamic sport as football. Is there a measure for character like Rocky has shown or bravery going into 50:50s. So many of these things actually lead to team spirit and are just as valuable as speed, technique and tackles won. You only have to look at Xg stats when Aberdeen were unbeaten to see how stats can tell a story but actually the results are very different. Over time the stats start to become more like reality but player performance often changes team to team and is very much about team mates and environment too. As long as it is just a part of player recruitment and used to confirm what scouts have witnessed then I like the idea. No convinced otherwise

[The Spaceman]

I think the OP summarises why I found it very funny how Hearts and BBC thrashed themselves silly when it was announced that they were going to subscribe to Jamestown Analytics. They aren’t suddenly the only team going big on data, every single team in the Premier League and UCL will be doing exactly the same as the money is there to justify the investment. Hibs will be benefitting from the Black Knights Group network as well. Like all tech races, there will be risers and fallers along the way.

[matty_f]

In the sevco discussion, this got me thinking on a separate theme:

“49ers cutting edge, data driven technology for player recruitment” …

.. so, that seems to be a thing just now. Brighton and Bournemouth spring to mind for two … but one thing that nags at me is … if they ALL have versions of it, then what does that mean ? Does it all just cancel out ? We all admire how Bournemouth and Brighton have recruited, and perhaps over-achieved, and many put it down to their recruitment software systems … but surely if they are so good, then Liverpool, Arsenal, Man City etc will just buy them out or find or create better systems at some point ? And then what ?

And .. here’s a thought … when AI kicks in … what then ?

(hmm if AI grows arms and legs .., will it then replace players too ?)

Ok, last line was flippant, but the rest of it intrigues me … it’s as clear as mud who and how are the leaders in this, and who are up and coming, and does anyone actually know the full picture ? And … I keep thinking other things round this lol … are the 49ers going to be better than the rest with their system in the US ? They aren’t at the moment.

Ok, I’ll stop there. That’s info overload, sorry folks. Welcome to my brain 🙄😖

It depends on how good the data is, how good the data model is, and how good the data analysts are.

For example

One club might have a player's xG stats for a match
One might have it for a season.
Another might have it for the player's full career.

They all have data, it's all the same measure, but they will likely tell you very different stories.

The first club might have a data model that considers xG as a shot on target
The second might look at xG as a measure of how good the chance was based on where the shot was taken
The third might look at it considering if it was a cross, a pass, a shot, a low or high shot, a header, if there was a defender blocking, and if it was on the player's weak or strong foot.

The first club's analyst might look at their data and think they've got a great idea of how good that player is.

The second club's analyst might look at it and think they've got a better idea than the first club has.

The third club's analyst will be looking at what the data doesn't tell them and figuring out how to get a model that mitigates that lack of information, as well as being able to tell a story based on what they know.

Data itself is just a tool, you need good analysts to understand the usefulness and limitations of the tool in order to get the most out of it.

[PatHead]

It all depends on how they interpret the data surely.

In addition surely the character of the player is equally important.

It's only part of the process.

[cubehindthegoal]

Im not sure it will make much difference. Graham Taylor used data pretty successfully. Get the ball in the box as most goals are scored there at a basic level. I can see how it can be used Moneyball style in Baseball as getting on base is teh equivelant of ball in box. However I'm struggling with such a dynamic sport as football. Is there a measure for character like Rocky has shown or bravery going into 50:50s. So many of these things actually lead to team spirit and are just as valuable as speed, technique and tackles won. You only have to look at Xg stats when Aberdeen were unbeaten to see how stats can tell a story but actually the results are very different. Over time the stats start to become more like reality but player performance often changes team to team and is very much about team mates and environment too. As long as it is just a part of player recruitment and used to confirm what scouts have witnessed then I like the idea. No convinced otherwise

Yeh get what you say, but also, isn’t tech and AI capable of doing what we can do better than we can, and can learn more and learn quicker and process quicker, that’s what I seem to keep hearing, and so will replace much of what we do and think is complicated and needs experience ? I’m just bouncing thoughts here, but can’t help thinking these are what we will be talking about some time soon anyway. Many are already accepting software can do better than people, and ones teams software might be better than another’s is going to make the difference … eg Brighton or Bournemouth, and there are much more I’m sure, but I’m a Hibs fan so that’s what I’ve noticed in posts here, as brighton are linked to our rivals I guess.

[cubehindthegoal]

I think the OP summarises why I found it very funny how Hearts and BBC thrashed themselves silly when it was announced that they were going to subscribe to Jamestown Analytics. They aren’t suddenly the only team going big on data, every single team in the Premier League and UCL will be doing exactly the same as the money is there to justify the investment. Hibs will be benefitting from the Black Knights Group network as well. Like all tech races, there will be risers and fallers along the way.

I can’t help thinking the teams with most money are already on this … and perhaps the fact Brighton and Bournemouth aren’t top of the league, but are maybe becoming clubs that are recruiting the odd player who will make their team better, and more you can do that better you get, but ultimately you will sell on for a big profit those same players, when Man City or suchlike decide they’re better than what they have.

Wonder what that all means for us I guess, for Hibs, that’s what I care about. Where does it lead to for us.

[CB Hibs 68]

I think the OP summarises why I found it very funny how Hearts and BBC thrashed themselves silly when it was announced that they were going to subscribe to Jamestown Analytics. They aren’t suddenly the only team going big on data, every single team in the Premier League and UCL will be doing exactly the same as the money is there to justify the investment. Hibs will be benefitting from the Black Knights Group network as well. Like all tech races, there will be risers and fallers along the way.
Of course but Jamestown analytics seem to hit the a mark.Brighton spot a player .Spend a few million quid on them .Eventually Chelsea on Liverpool will spend avast amount on the same player..That’s how it works

[cubehindthegoal]

It depends on how good the data is, how good the data model is, and how good the data analysts are.

For example

One club might have a player's xG stats for a match
One might have it for a season.
Another might have it for the player's full career.

They all have data, it's all the same measure, but they will likely tell you very different stories.

The first club might have a data model that considers xG as a shot on target
The second might look at xG as a measure of how good the chance was based on where the shot was taken
The third might look at it considering if it was a cross, a pass, a shot, a low or high shot, a header, if there was a defender blocking, and if it was on the player's weak or strong foot.

The first club's analyst might look at their data and think they've got a great idea of how good that player is.

The second club's analyst might look at it and think they've got a better idea than the first club has.

The third club's analyst will be looking at what the data doesn't tell them and figuring out how to get a model that mitigates that lack of information, as well as being able to tell a story based on what they know.

Data itself is just a tool, you need good analysts to understand the usefulness and limitations of the tool in order to get the most out of it.

Can tech progress, and using AI, perhaps replace even the analysts sooner than later too ? I’ve no idea myself, but it seems that’s what we hear now. So, whatever the analysts do now, AI learns and does easier and quicker .. and most of all - for some - cheaper.

Not sure it matters in many ways, from point of view of what club does better than others, but there’s a lot that’s going to change I guess in next few years.

[cubehindthegoal]

It all depends on how they interpret the data surely.

In addition surely the character of the player is equally important.

It's only part of the process.

I get that. And yes … it’s only recruitment .. the character of the player, the circumstances around the club, the players own personal circumstances, their mood and drives etc … but recruitment is what this is about. Presumably percentages and probability ? Ok .. now I’m getting out my depth lol .. but hope you see what I mean …

[matty_f]

Can tech progress, and using AI, perhaps replace even the analysts sooner than later too ? I’ve no idea myself, but it seems that’s what we hear now. So, whatever the analysts do now, AI learns and does easier and quicker .. and most of all - for some - cheaper.

Not sure it matters in many ways, from point of view of what club does better than others, but there’s a lot that’s going to change I guess in next few years.

AI will play a part but I think there will still be a high degree of human validation needed. It's certainly an interesting point - it would be something i would want to learn more about, definitely.

[Viva_Palmeiras]

Can AI predict the outcome on a player who gets pelters for having the occasional lapse in concentration?

Or predict whether a player with fantastic “stats”
has the baws to ignore history and go on to make their own and bust 114 years jinx to lift the Holy Grail?

What’s our Holy Grail now btw - still the Scottish Cup?

[GreenCastle]

It all depends on how they interpret the data surely.

In addition surely the character of the player is equally important.

It's only part of the process.

Exactly..

It’s only as good as the staff using / understanding it plus linked to things like wage structure / club funds etc.

Data stuff has been around for a while and clubs all use it differently.

Regarding AI - clubs have already been using it for scouting players and various other predictions like how a player will fit into a team and things like if certain players will play better under a different coach with different style etc.

AI isn’t the sole way to sign players but clubs like Celtic have openly admitted using it to help look at players coming into the club.

[wookie70]

Yeh get what you say, but also, isn’t tech and AI capable of doing what we can do better than we can, and can learn more and learn quicker and process quicker, that’s what I seem to keep hearing, and so will replace much of what we do and think is complicated and needs experience ? I’m just bouncing thoughts here, but can’t help thinking these are what we will be talking about some time soon anyway. Many are already accepting software can do better than people, and ones teams software might be better than another’s is going to make the difference … eg Brighton or Bournemouth, and there are much more I’m sure, but I’m a Hibs fan so that’s what I’ve noticed in posts here, as brighton are linked to our rivals I guess.

If you can make something a bit of comparable data then maybe but we are miles off AI determining character or bravery imo. I still think AI is only as clever as the people that create the algorithms. It will certainly be able to crunch much more data from many more sources but I think the Intelligence part is overstated at least for now. If AI is clever enough to pick the best players then surely it could ref a match. Far too many grey areas for it to do that imo as there are in determining if a player is the correct signing to blend with the group, fit in with the style of football in the country, teh way we play, bounce over other players etc etc. All those things are hard to define so how can you tell AI what to think. Tackles won etc are fine but is teh data complete enough to say if every tackle was massively in their favour. No chance we will get data that detailed for our level.

I could maybe see AI playing a part in formations as that may be easier to use data for. ie look at the team you are playing against various formations and ask AI where all the chances were cred for that team and against, what side, long pass or short passing etc.

Frankly I'd rather they concentrated use on where it makes things better. I'd rather see it banned from football than embraced unless it was for something like generating subtitles or sign language on the fly, even then I'd rather a human done that job

[cubehindthegoal]

Can AI predict the outcome on a player who gets pelters for having the occasional lapse in concentration?

Or predict whether a player with fantastic “stats”
has the baws to ignore history and go on to make their own and bust 114 years jinx to lift the Holy Grail?

What’s our Holy Grail now btw - still the Scottish Cup?

But don’t the folk running our clubs, managing them, do that? Maybe AI could too? Hate the thought in many ways tbh, but it’s something that’s rolling as far as I see it. No expert though so happy to be shown wrong.

[Viva_Palmeiras]

I suspect that recruitment patterns and data and success will need to reveal over time. Not many players from Lapland making it in the Premiership just yet…

Not sure how extensive the data is and coverage relevant to Scotland. Was and is Chris (Owl man) Mueller US international? Rangers has US capped players come over and make a difference.Chris just didn’t work out. His self motivational quotes were ridiculous tho.

Pre AI - Jonatan Johansson came to Hibs with pedigree in our division - left without scoring a single goal - could AI have predicted?

Any which way you look at data - it needs to have quality, relevance and for it to be useful, suitable interpreted and analysed - or is that deferred to a machine?

[cubehindthegoal]

If you can make something a bit of comparable data then maybe but we are miles off AI determining character or bravery imo. I still think AI is only as clever as the people that create the algorithms. It will certainly be able to crunch much more data from many more sources but I think the Intelligence part is overstated at least for now. If AI is clever enough to pick the best players then surely it could ref a match. Far too many grey areas for it to do that imo as there are in determining if a player is the correct signing to blend with the group, fit in with the style of football in the country, teh way we play, bounce over other players etc etc. All those things are hard to define so how can you tell AI what to think. Tackles won etc are fine but is teh data complete enough to say if every tackle was massively in their favour. No chance we will get data that detailed for our level.

I could maybe see AI playing a part in formations as that may be easier to use data for. ie look at the team you are playing against various formations and ask AI where all the chances were cred for that team and against, what side, long pass or short passing etc.

Frankly I'd rather they concentrated use on where it makes things better. I'd rather see it banned from football than embraced unless it was for something like generating subtitles or sign language on the fly, even then I'd rather a human done that job

Personally I’d prefer us recognising our human abilities and expanding that, but some will say AI is an example of that. I’d discuss that 😉 But it’s not that easy either .. you said you think AI is only as clever as the people that create the algorithms .. but what happens when AI creates the algorithms ? In fact, that must be happening already.

Hmm. Sorry folks in a way, this is brain breaking when you think about it. But also it’s real. Hey … we are Hibs..

Europe
Floodlights
etc

Maybe AI next 😉

[CB Hibs 68]

Of course but Jamestown analytics seem to hit the a mark.Brighton spot a player .Spend a few million quid on them .Eventually Chelsea on Liverpool will spend avast amount on the same player..That’s how it works

Don’t want to be boring but there is absolute no doubt that Jamestown analytics are head and shoulders above any other team.Whether that transfers to Hearts we will see..

[cubehindthegoal]

Pre AI - Jonatan Johansson came to Hibs with pedigree in our division - left without scoring a single goal - could AI have predicted?

No being cheeky, honest, but could AI have done any worse there ?

[cubehindthegoal]

Don’t want to be boring but there is absolute no doubt that Jamestown analytics are head and shoulders above any other team.Whether that transfers to Hearts we will see..

What are the other analytics companies/software that other teams use ? We need to compare to know that I think. For example, do Man City or Barcelona or Bayern Munich or other teams have similar systems and use them, and how do they compare ? I think we need to look at what the teams above Brighton are using, and why they are still above them (is it just money?) … what you say doesn’t present any evidence, no offence intended.

[cubehindthegoal]

My brain can’t cope any more tonight contributing to my post here … not anyone else’s fault, and I apologise … but hopefully it’s been interesting and useful .. and I do believe our club can and have been leaders in a good way without credit over the years .. so .. maybe this is something we can embrace and find our way to add a lead to somehow … just a thought

.. and my last thought for now …I need to rest !

[wookie70]

Personally I’d prefer us recognising our human abilities and expanding that, but some will say AI is an example of that. I’d discuss that 😉 But it’s not that easy either .. you said you think AI is only as clever as the people that create the algorithms .. but what happens when AI creates the algorithms ? In fact, that must be happening already.

Hmm. Sorry folks in a way, this is brain breaking when you think about it. But also it’s real. Hey … we are Hibs..

Europe
Floodlights
etc

Maybe AI next 😉 It is a bit like where does the sky stop. Once it is a big part of life and you could argue it already is then it probably can't be stopped. There is an old way of looking at knowledge called the wisdom of crowds. In my head that is what AI is at the moment. The difference is it is a fairly knowledgeable crowd to start with. Saying that look at the absolute half-wits that Musk has brought in who are supposedly whizz kids. I'm not sure I would trust them to be able to work out how to write BOOBLESS on a Texas Instrument calculator. I'm so glad I am old and can for the most part ignore technology I have no interest in. Interesting discussion though

[matty_f]

I wouldn't be surprised if there isn't already an AI model either in use or in development that could take data from players' training and match performances, figure out strengths and weaknesses and who's most likely to have a good game with which other players, to the extent where it could suggest a team with a breakdown of why it's suggesting it.

[007]

I wouldn't be surprised if there isn't already an AI model either in use or in development that could take data from players' training and match performances, figure out strengths and weaknesses and who's most likely to have a good game with which other players, to the extent where it could suggest a team with a breakdown of why it's suggesting it.

Maybe I need to upgrade to ChatGPT Pro.

Who should Hibs sign as their next striker?

https://i.ibb.co/QhnZxfs/Screenshot-20250221-010650-Chat-GPT.jpg (https://ibb.co/mdtmPGQ)

[Unseen work]

Data is all well and good.

But nothing meets the classic ‘eye test’ for me.

[007]

Data is all well and good.

But nothing meets the classic ‘eye test’ for me.

The data should be used to narrow down who should be watched.

[HoboHarry]

On the subject of AI, I've wondered a few times if Erling Haaland was created by AI to show it could be done and then to prove AI has a sense of humour it made him Norwegian so we'd never see him at a major tournament. :greengrin

[Yorkshire HFC]

Data is all well and good.

But nothing meets the classic ‘eye test’ for me.

How does the data deal with a player who goes out drinking the three nights a week, who's dog just died, who's going bankrupt, who wants to leave because of rubbish weather in Scotland? How does it deal with human emotions?

[Callum_62]

How does the data deal with a player who goes out drinking the three nights a week, who's dog just died, who's going bankrupt, who wants to leave because of rubbish weather in Scotland? How does it deal with human emotions?Possibly better than some humans? [emoji6]

Sent from my Pixel 7 Pro using Tapatalk

[Centre Hawf]

Something that has always bothered me with 'data-driven' recruitment thing is that a lot people will tend to reference Moneyball. But for all the lovely stories about underdogs, and Brad Pits acting, the reality is it doesn't really work the way people think it does. Especially when, as the OP alluded to, the big dogs start doing it too.

Not to bore anyone with baseball history but while the Oakland A's had a great season when using it, something the film and story neglects to mention is they already had some star players at that time and never actually came close to winning the World Series and were actually flung out the post season without much of a fight. They did do well in their free agency recruitment and found diamonds in the rough definitely. But they didn't go from the team that was under '50 feet of crap' to topple the big teams etc.

When the rest of the league was put on alert on ways to find some sort of hidden value in some players it was already game over for 'Moneyball' as the spending power has just toppled any tiny gains given to you by it. Now you have teams like the Dodgers/Mets/Yankees handing out half a billion dollar salaries for 10 year contracts.

In terms of football, yes you can have all the data you want and even add in AI to try and get it to collect or be analysed faster, but as others said you still need a person at the end to make sense of it all. There will be a team one day within the next 10/15 years that will have a lovely BBC or Athletic article about how AI decides their transfer targets to squeeze better value out of their business and compete at a higher level, it will be heralded as some sort of new era thinking, and in time it will fail like the rest of them do, because ultimately more money with better staff that are good at their jobs will always win over a spreadsheet.

[matty_f]

Data is all well and good.

But nothing meets the classic ‘eye test’ for me.

You need both, imho - we're at a point where anyone relying solely on one or the other is at a disadvantage.

[A Hi-Bee]

How does the data deal with a player who goes out drinking the three nights a week, who's dog just died, who's going bankrupt, who wants to leave because of rubbish weather in Scotland? How does it deal with human emotions?

They make a country & western song about it, no problem.:aok: