FOLLOW

FOLLOW

SHARE

Analysing The Transfer Window With Data

Big players cost big money - but is it time to say enough is enough

26Sep

Fuelled by speculation, it seems as if the transfer window never closes. Players are often ‘courted’ months beforehand – since Luis Suarez’s transfer to Barcelona from Liverpool, a number of reputable sources have claimed that the transfer had been wrapped up for months. The transfer window does however only open twice a year - before the start of the season and in January.

Many commentators have suggested that the summer window should close before the first match of the season, but the reality is that for the first fortnight of the league, managers have one eye on their team on the pitch and the other on the potential recruits that are going to be needed in order to ensure that the team performs well. There’s now little room for managers looking to build a team over a three-year period and this mind-set has coincided with a gung-ho approach to the market, with fans demanding that their club attracts the biggest names.

You might think that managers making transfers at the last minute is a bit irresponsible, but every year there’s at least one major transfer on the final day. Last year it was Marouane Fellani, this year it was Radamel Falcao, the Colombian striker who transferred to Manchester United on a loan from Monaco. It’s a deal which in reality is probably going to cost Manchester United over £50,000,000, but with FIFA’s Financial Fair Play restrictions now enforceable, loan deals are quickly becoming a loop-hole for some of Europe’s more affluent clubs.

The madness that surrounds transfer deadline day is just one of the hurdles that databases like CIES have to overcome. As many clubs tend to leave the majority of their business to the last forays of the market, there is a tendency to leave themselves at the peril of clubs who quite frankly, are out to squeeze every last penny from their rivals. It’s situations like this, which make it so difficult for companies like CIES to correctly foresee the valuation of a player.

When we looked at the use of data in the CIES Football Observatory a couple of editions back, it became clear that the transfer market had been spiralling out of control for some time, and like a leaking roof, it was always going to get worse - £835 million was spent by English Premier League clubs and Angel Di Maria became the most expensive British import since the ill-fated Fernando Torres signed for Chelsea in 2012 – it was a partnership that ironically came to a painful end the same week his own record was smashed. He left Chelsea for AC Milan in a 2-year loan deal, after being widely perceived as one the biggest flops of all time.

Just as a recap, the CIES Football Observatory uses a statistical model to assess the value of a player – it uses information from a sample of 1,500 players transferred for a fee from the top-5 European Leagues. It takes into account a number of variables such as the player’s age, the years he has remaining on his contract, international caps and position. Lionel Messi is the world’s most valuable player, worth almost twice as much as his nearest rival, Cristiano Ronaldo.

It was very interesting to see what their statistical model had made of the latest transfer window, and whether it could determine which clubs had over-spent on players.

The biggest spenders in the English Premier League were Manchester United, who spent in the region of £150 million on 5 players. According to the CIES model, they over-spent on three of their players – Luke Shaw, Ander Herrera and Angel Di Maria, a player who according to CIES came at 30 million euro premium. But what does this mean for efficacy of this model? Were Manchester United just out-negotiated by Real Madrid, or alternatively does it just mean that data can’t keep up with the intricacies of the market?

It’s all about the market place, supply and demand, if you will. Manchester United were in desperate need of a world class talent, Real Madrid had one in Angel Di Maria, and conveniently it was a player they were more than willing to let go. This was a chain of events that put Real Madrid firmly in the driving seat – if Manchester United weren’t willing to spend the required money, Real Madrid would have wagered that one of Europe’s elite, whether it be Bayern Munich, Chelsea or Manchester City would have succumbed to temptation sooner rather than later. It’s difficult for data to predict this, and makes the concept of ‘over paying’ somewhat redundant.

Timing also plays an important role. Take David Luiz, a player that CIES feel PSG have spent 29 million euros more on than they should have. He had a pretty solid season with Chelsea and was one of the faces of the World Cup, and although his price tag at the time raised a few eyebrows it wasn’t met with the ridicule that it is now. After his woeful showing against Germany, were Brazil were humiliated 7-1, there’s no chance PSG would have spent that kind of money on him, as he is, in effect, damaged goods. Thankfully for Chelsea, PSG bid whilst his value was at its highest. Again, the sophistication of the CIES model has not yet reached the level where these fluctuations in value can be reflected in the data. On the other side of the coin, James Rodriguez, a Colombian, attacking midfielder who recently signed for Real Madrid, would never had been valued at £60 million had he not put together a string of impressive performances in Brazil, something which arguably inflated his value by 25 million euros.

Of course, it’s not always the buyers that lose out, sellers do sometimes as well. CIES valued Diego Costa as having an upper limit of 54.5 million euros at the end of last season- Chelsea got him for around 40 million euros, triggering his buyout clause. This also marks a limitation for CIES, surely a player’s value can only be as high as his buy out clause? It’s mandatory for players to have one in Spain and although Atletico Madrid could have effectively put a £500 million pound clause in Diego Costa’s contract, he would have been unlikely to sign as it would have inhibited him from moving to a more lucrative club in the future.

Even the players CIES claim have gone for the right price are a little questionable – the general consensus is that Toni Kroos arrived at Real Madrid as a bargain, a player whose performances in the Champions League and throughout the World Cup have seen him heralded as one of the world’s most established midfielders – at £24 million, £5 million less than what Manchester United paid Bilbao for Andrea Herrera, an up and coming talent, who’s probably unlikely to reach the same level as Toni Kroos regardless of well he progresses. This however is a victory for CIES and shows how their model incorporates the time left on a contract - if Kroos hadn’t been so close to having his contract expire, he would never have been so cheap, but rightfully, with only one year left, the player’s valuation was also going to be lower than if he had just signed a new contract.

If I were a developer at CIES, I would say unconditionally that the majority of Europe’s top clubs have, by and large, overpaid for most of their recent acquisitions. It could be argued that Barcelona’s capture of Luis Suarez was something of a coup, but the majority of the window’s major signings have not be in line with their valuations through data. It seems as if you want to attract the best players these days you have to pay through the nose for it. I wouldn’t say that CIES’s algorithms are defunct; they just need to develop significantly if they’re to actually reflect the intricacies and fluctuations of the transfer market.

Comments

comments powered byDisqus
Tennisraq

Read next:

Presenting a Diorama of Player Movement in Sport

i