Good piece by Martyn Samuel in todays Mail

[jeremyfulham]

He practically dismantles our method of buying players , the gist of it is data does not tell you anything about a players character , giving AK as a prime example whom he pegs as a wrong un . Hopefully the powers that be are listening !

[elgreenio]

link :

https://www.dailymail.co.uk/sport/football/article-6629271/What-stats-not-tell-Fulham-Aboubakar-Kamara-total-pest.html

The last song Eddie Cochran recorded before his death was the B-side to his single Three Steps To Heaven. It was called Cut Across Shorty and it's a classic.

There's a wonderful version by Rod Stewart on Gasoline Alley, his band, Faces, used to cover it live, Freddie and the Dreamers had a go, as did some fine country artists. Great song.

Cut Across Shorty is just a reworking of the tortoise and hare story, really. It tells of a foot race between 'a country boy named Shorty, and a city boy named Dan'. The prize for the winner is the hand in marriage of a certain 'Miss Lucy'.

Without wasting time on details, Miss Lucy is less of an unwitting stooge than she may appear, has a preference, and fixes the race for her favourite. That's how Shorty gets to cut across.

Why? Well, as the song's lyric explains: 'Now Dan had all the money/And he also had the looks/But Shorty must have had something, boys/ That can't be found in books...'

And we know, we get it. Marijohn Wilkin who, along with Wayne P Walker, wrote Cut Across Shorty, honed in on a universal truth. That there are some human qualities that defy logic and rational analysis; that cannot be found in learning; that are innate.

And so to Fulham and Roster Improvement Through Analysis (RITA).

Last summer, Fulham became the first promoted club to break the £100million transfer barrier. They were the third biggest spenders in the Premier League and, in 2018, the 13th biggest spenders in Europe, ahead of Bayern Munich, Borussia Dortmund and all of Serie A, bar Juventus and Roma.

Their reward, so far, is 19th place and, this week, a fracas at the training ground which saw striker Aboubakar Kamara arrested on suspicion of actual bodily harm and causing criminal damage. He had arrived in an attempt to sort out his future but instead sorted out a club security guard, inadvertently achieving his initial aim.

Kamara joined in 2017, by which time Fulham's RITA system was very much in place. Indeed, he is a typical RITA signing. Kamara started at Monaco, didn't make it and had an unmemorable year with Kortrijk in the Belgian league before signing for Amiens, a small club in France's Ligue 1.

After a single reasonable season, Fulham bought him for a fee in the region of £5.3m.

No doubt his numbers were excellent. Yet there is one aspect of a player's performance RITA cannot accurately espy: character. The warning signs were there in Kamara's first season in the Championship when he scored seven goals but drew more cards — seven yellow, one red.

This year, Kamara's flaws have been increasingly evident. He ignored his manager, Claudio Ranieri, and his team-mates to take a penalty ahead of Aleksandar Mitrovic — which he then missed.

He argued with Mitrovic again during a yoga session, for which he was dropped from the team and skipped training having been banished to work with the under-23 squad. Then, when he attempted to confront chief executive officer Alistair Mackintosh, a fight broke out at the training ground.

It appears Kamara, like Shorty, has something that cannot be found in books, or located via an analytics programme: he's a high-maintenance pest.

Also not showing up on computer screens were language skills, which is how Fulham came to sign two goalkeepers, Fabricio Agosto Ramirez and Sergio Rico, who barely spoke English.

Fabri, in particular, struggled immensely and could not understand his defenders or make himself understood.

He started the first two games of the season, let in five goals and has not been seen since, not even chanced in Carabao Cup matches with Exeter and Millwall or in the FA Cup with Oldham.

Fabri was a title winner with Besiktas but also the player who started to cry after letting in four first-half goals for the club, on the way to losing 6-0 against Dynamo Kiev in the Champions League.

'His scouting profile and data profile are both strong,' Fulham confirmed when Fabri arrived for a ball-park £5m.

Obviously, recruitment cannot just operate off a series of hunches. Analytics are a vital part, as are scouting and first-hand knowledge. One of Fabri's previous clubs was Deportivo La Coruna, where he worked with Fulham's goalkeeping coach Jose Sambade Carreira.

It is possible the same mistake could have been made by going the old-fashioned route of scouting research and recommendation. Yet Fulham's analytics department also rejected Glenn Murray, before he went to Brighton and Callum Wilson — now valued at £50m by Bournemouth — because their numbers did not add up.

The same numbers that cannot identify a suspect temperament, or a weakness of character, or incompatibility with team-mates, facets that are no less important.

In 1984, when Terry Venables considered taking Steve Archibald to Barcelona, he asked around.

Venables liked Archibald as a player, but didn't know him. He made discreet enquiries. What was he like? Quiet, everyone said, bit of a loner, doesn't really mix with the lads. At an English club that might have been a drawback — but Venables wanted him in Spain. A solitary man might as well be alone in Barcelona, as London, he reasoned. All the problems that might affect a British player abroad — language, being outside the group, inability to socialise — wouldn't matter.

Venables signed Archibald, who was a great success, and they won Barcelona's first league title in 11 years.

The need for character cannot be underestimated. It is what Maurizio Sarri is alluding to in his comments about Eden Hazard.

He plainly admires Hazard's abilities as a player but does not see him possessing the mentality to be a leader, as he should be at Chelsea.

'Wonderful, but an individual player,' was his description. 'He's more an individual than a leader.'

Sarri hasn't had everything right this season but he makes a point here. If one aspect of Hazard's make-up stops him being the best in the world it is his character.

He disappears for weeks, sometimes entire seasons, if circumstances are not to his liking, in a way Cristiano Ronaldo or Lionel Messi never do. He can turn matches in a minute, as Sarri identifies, but can also turn off — and while there will be a clue with a dip in his numbers, these alone can never tell the whole story.

Miss Lucy got it, Fulham didn't. That's the problem with analytics.

I think it's a pretty woefully thought out piece to be honest, but anything to waste a few minutes of a Friday shift at work

[Denver Fulham]

I'm so tired of this. WE HAVE SCOUTS. THEY WATCH THESE PLAYERS IN PERSON AND TALK TO PEOPLE ABOUT THEM BEFORE WE ELECT TO SIGN THEM.

(That's not to mention that data CAN determine personality traits; who knows if our setup does that, but it's absolutely possible)

If so many people on this board think everyone we sign is bad, where is the blame for Talbot and his crew? They're literally watching them play and determining that they are good.

[Woolly Mammoth]

A good piece by Martin Samuel, he hit the nail on the head.

[YankeeJim]

I don't have a problem with the use of stats as a tool to find players but they should be scouted so that what we yanks call "intangibles" can be sorted. As the article indicates important things don't show up. They also don't show how a player is being used. If all you saw of Sess was his Prem play at left back would you splash 50 mil for him? So, go ahead use stats but use them to better control what your scouts look at. Then the scouts can see if that player is still running and battling when the team is down by two and if he has legs in the last minutes. Heart and fitness go a long way.

[Denver Fulham]

I don't have a problem with the use of stats as a tool to find players but they should be scouted so that what we yanks call "intangibles" can be sorted. As the article indicates important things don't show up. They also don't show how a player is being used. If all you saw of Sess was his Prem play at left back would you splash 50 mil for him? So, go ahead use stats but use them to better control what your scouts look at. Then the scouts can see if that player is still running and battling when the team is down by two and if he has legs in the last minutes. Heart and fitness go a long way.

"Stats" absolutely show how a player is being used. You can parse their data any of a billion ways with a proper system.

[Mince n Tatties]

I'm so tired of this. WE HAVE SCOUTS. THEY WATCH THESE PLAYERS IN PERSON AND TALK TO PEOPLE ABOUT THEM BEFORE WE ELECT TO SIGN THEM.

(That's not to mention that data CAN determine personality traits; who knows if our setup does that, but it's absolutely possible)

If so many people on this board think everyone we sign is bad, where is the blame for Talbot and his crew? They're literally watching them play and determining that they are good.

Do you actually think Talbot and his crew are out in all these different countries,watching all these players.
I'll tell you where he is sitting at home eating fish n chips with the wife,and looking at you tube highlights of some Belgian league 3 player....Wow he looks good,what's his agents number.😎

[Deeping_white]

It's an awful article, incredibly lazily written and parts of it are conjecture or made up. It only really serves the purpose of bashing recruitment that uses stats even though every single premier league club will be using stats to sign players. We're just an easy target as the club have been open about signing players using statistical analysis. Ironically he could use much better examples than Kamara of potential stats failures but instead tries to make a tedious link around someone's personality as a means to bash stats based signings which is quite stupid really when you think of the fact that Chelsea spent £40m buying Bakayoko who flopped beyond belief and we spent £5m on a player who helped get us out of the championship and in all honesty we wouldn't struggle to recoup that figure either. Martyn Samuel is just part of the old fashioned group of journalists that don't like to see the football world evolving, but then he is a Daily Mail journo so that should sum up the state of what he's going to publish before it's even out, because they don't like to let facts get in the way of writing some utter drivel.

[Nero]

I'm so tired of this. WE HAVE SCOUTS. THEY WATCH THESE PLAYERS IN PERSON AND TALK TO PEOPLE ABOUT THEM BEFORE WE ELECT TO SIGN THEM.

(That's not to mention that data CAN determine personality traits; who knows if our setup does that, but it's absolutely possible)

If so many people on this board think everyone we sign is bad, where is the blame for Talbot and his crew? They're literally watching them play and determining that they are good.

But are our Scouts any good is the question? If they've watch some of these players it worrying

[Statto]

(That's not to mention that data CAN determine personality traits; who knows if our setup does that, but it's absolutely possible)

How?

I agree aspects of the article are terrible but he makes a good point for example about Steve Archibald suiting the culture/lifestyle of a foreigner in Barcelona better than he suited being 'one of the lads' in an English team. What data can you "parse" to show that? Is there a stat for the number of pints they have in the pub each week?

[Denver Fulham]

(That's not to mention that data CAN determine personality traits; who knows if our setup does that, but it's absolutely possible)

How?

I agree aspects of the article are terrible but he makes a good point for example about Steve Archibald suiting the culture/lifestyle of a foreigner in Barcelona better than he suited being 'one of the lads' in an English team. What data can you "parse" to show that? Is there a stat for the number of pints they have in the pub each week?

Modern data analysis absolutely can account for/benefit from knowing personality traits. The simplest example is something like Facebook, where users actively participate in the data submission and then get served ads accordingly based on their preferences.

For footballers, you can do personality profile testing and see what qualities translate within your team, your league, that position etc. And that doesn't even approach using artificial intelligence, which can suss out your personality: https://www.cnet.com/news/artificial-intelligence-can-predict-your-personality-by-looking-at-your-eyes/

No one assessment is going to be perfect, but it definitely can be accounted for between data, scouting eyeballs and conversations/testing.

[Statto]

(That's not to mention that data CAN determine personality traits; who knows if our setup does that, but it's absolutely possible)

How?

I agree aspects of the article are terrible but he makes a good point for example about Steve Archibald suiting the culture/lifestyle of a foreigner in Barcelona better than he suited being 'one of the lads' in an English team. What data can you "parse" to show that? Is there a stat for the number of pints they have in the pub each week?

Modern data analysis absolutely can account for/benefit from knowing personality traits. The simplest example is something like Facebook, where users actively participate in the data submission and then get served ads accordingly based on their preferences.

For footballers, you can do personality profile testing and see what qualities translate within your team, your league, that position etc. And that doesn't even approach using artificial intelligence, which can suss out your personality: https://www.cnet.com/news/artificial-intelligence-can-predict-your-personality-by-looking-at-your-eyes/

No one assessment is going to be perfect, but it definitely can be accounted for between data, scouting eyeballs and conversations/testing.

The suggestion that we can tell how good a player will be by scanning their eyes has just replaced the Magath cheese story as the most ridiculous Fulham-related comment I've ever heard.

The other two suggestions, Facebook and personality tests, might be reasonable in some other context but I can't see how they'd work in this context. Would that mean we can only sign players who are highly active Facebook users, with a public profile? And that when we're interested in a player, we fax their agent a personality test and ask them to complete and return it before we decide whether to put in a bid with their club? 

[filham]

A very good article.
I would just add that the proof of the pudding is in the eating.
If that £100m summer spending was based on our Stats System then the system is not very god and needs dropping PDQ.

[Denver Fulham]

(That's not to mention that data CAN determine personality traits; who knows if our setup does that, but it's absolutely possible)

How?

I agree aspects of the article are terrible but he makes a good point for example about Steve Archibald suiting the culture/lifestyle of a foreigner in Barcelona better than he suited being 'one of the lads' in an English team. What data can you "parse" to show that? Is there a stat for the number of pints they have in the pub each week?

Modern data analysis absolutely can account for/benefit from knowing personality traits. The simplest example is something like Facebook, where users actively participate in the data submission and then get served ads accordingly based on their preferences.

For footballers, you can do personality profile testing and see what qualities translate within your team, your league, that position etc. And that doesn't even approach using artificial intelligence, which can suss out your personality: https://www.cnet.com/news/artificial-intelligence-can-predict-your-personality-by-looking-at-your-eyes/

No one assessment is going to be perfect, but it definitely can be accounted for between data, scouting eyeballs and conversations/testing.

The suggestion that we can tell how good a player will be by scanning their eyes has just replaced the Magath cheese story as the most ridiculous Fulham-related comment I've ever heard.

The other two suggestions, Facebook and personality tests, might be reasonable in some other context but I can't see how they'd work in this context. Would that mean we can only sign players who are highly active Facebook users, with a public profile? And that when we're interested in a player, we fax their agent a personality test and ask them to complete and return it before we decide whether to put in a bid with their club? 

I was just using Facebook as an example of an enterprise that can turn personal likes/dislikes into useful information.

American football franchises give prospects all sorts of personality tests and quizzes as part of the evaluation process. Whether it's used well/properly depends on the team's investment in said analysis.

[Sting of the North]

How can anyone think that this is a good article? Although it is quite well written from a stylistic point of view, it is basically a pile of thrash from a journalistic perspective. It sets out a false (or at least questionable) premise (that we rely solely on stats and therefore completely ignore all other aspects), and then goes on to prove nothing with a few mind-numbingly shallow examples. Even without any sort of research he should have been able to write something better than this.

Shows the state of journalism today if people believe that this is anything worth any praise.

Can anyone point to anything in the article that actually adds anything else than "you cannot rely solely on stats as they don't give the full picture"? Because surely that is no news to anyone?

Sure, the Archibald part was interesting, but there is no connection to our current process. This has probably to do with the fact that no one outside the club knows exactly how it looks, but still.

[bill taylors apprentice]

I'm so tired of this. WE HAVE SCOUTS. THEY WATCH THESE PLAYERS IN PERSON AND TALK TO PEOPLE ABOUT THEM BEFORE WE ELECT TO SIGN THEM.

Do they?
If so, they are clearly not watching closely enough or talking to the right people!

[David I]

A good piece by Martin Samuel, he hit the nail on the head.

I would much prefer someone would hit TK on the head 😉

[Statto]

I was just using Facebook as an example of an enterprise that can turn personal likes/dislikes into useful information.

American football franchises give prospects all sorts of personality tests and quizzes as part of the evaluation process. Whether it's used well/properly depends on the team's investment in said analysis.

Still not sure how that would work in a football transfer. I suppose at the same time as a new signing has their medical, we'd make them sit a psychometric test. I will admit that would at least be logistically viable. But it would have to be part of the late stage due diligence, when negotiations are already at an advanced stage, so it wouldn't help us with the initial search for potential targets. Also a fundamental problem is that psychometric tests and such like are still nigh on meaningless. I've done several for work, generally answered them lazily and randomly, and been shown to have totally different personality types from one test to the next.

[Lyle from Hangeland]

I'm so tired of this. WE HAVE SCOUTS. THEY WATCH THESE PLAYERS IN PERSON AND TALK TO PEOPLE ABOUT THEM BEFORE WE ELECT TO SIGN THEM.

(That's not to mention that data CAN determine personality traits; who knows if our setup does that, but it's absolutely possible)

If so many people on this board think everyone we sign is bad, where is the blame for Talbot and his crew? They're literally watching them play and determining that they are good.

Yeah, I agree. Martin Samuel doesn't know what he's talking about. Fulham does employ scouts who actually go and look at players. Major League baseball teams knee deep in big data analysis employ dozens of scouts to literally watch a game a day during the baseball season. No doubt they also try to measure to some degree the immeasurable like character, ability to take instruction, etc, by getting their scouts to ask around and observe. Martin is trolling folks for clicks.

[jeremyfulham]

It's an awful article, incredibly lazily written and parts of it are conjecture or made up. It only really serves the purpose of bashing recruitment that uses stats even though every single premier league club will be using stats to sign players. We're just an easy target as the club have been open about signing players using statistical analysis. Ironically he could use much better examples than Kamara of potential stats failures but instead tries to make a tedious link around someone's personality as a means to bash stats based signings which is quite stupid really when you think of the fact that Chelsea spent £40m buying Bakayoko who flopped beyond belief and we spent £5m on a player who helped get us out of the championship and in all honesty we wouldn't struggle to recoup that figure either. Martyn Samuel is just part of the old fashioned group of journalists that don't like to see the football world evolving, but then he is a Daily Mail journo so that should sum up the state of what he's going to publish before it's even out, because they don't like to let facts get in the way of writing some utter drivel.

You really think we will get our money back on him ??? Your criticism of the article couldn't be because you hate the paper it was in could it ?