Alex Stewart assesses what patterns in transfers between Ligue 1 and the Premier League can tell us about whether clubs follow trends in player acquisition
I recently wrote a piece for Analytics FC‘s LinkedIn page on French centre-backs – you can read it here should you feel so inclined. The piece was a extension of work done in an article by The Athletic which contained a line that got me thinking:
“Football is also a human market of bias and expectations. In recent seasons, several expensive defensive signings from Ligue 1 have signed and prospered….Every successful arrival is proof of concept and the Premier League’s reliance on Ligue 1 is likely to stay.”
On the face of it, this is a very sensible observation. Arsenal especially have done well: William Saliba and Gabriel are two elite defenders who made the journey across the Channel, with both making last year’s PFA Team of the Year. French players currently make up the second biggest cohort of Premier League players after English (although not all French players have been bought from Ligue 1 and not all Ligue 1 transfers are of French players).
I thought it would be interesting to look at the last 25 years of signings from Ligue 1 to see if there were any trends. I picked this time period because a) 25 years is a pretty robust sample and b) I expect that there is greater uniformity in access to markets and funds now, as well as more widespread use of good scouting methods outside the UK. The 10 years before that, post-Wenger essentially, was interesting, though. Ginola joined Newcastle in 1995/96, Franck Leboeuf and Nicolas Anelka the following year, while (pub quiz nerds take note), the most expensive French signing in 1997/98 was neither Manu Petit or Gilles Grimandi, but Valérien Ismaël joining Palace. The spread of clubs was fun, too, with Newcastle, West Ham, and Liverpool joining usual suspects Arsenal in raiding Ligue 1.
Anyway. Using data from Transfermarkt, I noted down transfers from Ligue 1 to the Premier League by position to see if the observation that successful transfers offered some kind of proof of concept that then showed up in the data. Given that only around half of signings are a success, it was hard to expect anything specific but my working assumption going in was that certain transfers would surely move the needle (looking at you, N’Golo).
Firstly, the total number of signings, and here we can see probably more stability over the last 25 years than expected.
The downward trend is mild and a handful of signings in the last five years would probably keep the line stable. This suggests that Ligue 1 overall is very much not a whim and that the market has been of interest to Premier League clubs consistently over the last 25 years.
Is there anything to the peaks? If we follow the logic of The Athletic piece, the year or two before a peak should show some successes that encouraged an upswing in interest. Although not in the data, World Cup winning year 1998 should obviously be just such a peak: Leboeuf was already at Chelsea, and Petit at Arsenal, but Marcel Desailly joined his centre-back partner at Chelsea (from Serie A, though), while Stephane Guivarc’h did join Newcastle in the summer of ’98. But while ’98/99 saw 14 transfers, ’99/00 saw a below average nine.
The first real peak in our dataset is 2002/03. The French were eliminated from the World Cup in the groups, so that likely had no positive impact (maybe it reduced costs?). But ’00/01 saw the arrivals of Sylvian Wiltord, Fabien Barthez, and Robert Pirès, all of whom played with distinction, while Olivier Dacourt was also a big name signing. The following window 01/02 was also important: Fulham spent heavily with deals for Steve Marlet, Steed Malbranque (who racked up 55 Premier League assists), and Sylvain Legwinski, probably because their Head Coach was Frenchman Jean Tigana. William Gallas also joined Chelsea, while Jon Arne Riise moved from Monaco and Laurent Robert from PSG. Interesting, the big selling clubs were Lyon (finished 2nd in Ligue 1 in 2000/01), Bordeaux (4th), and PSG (9th), but champions Nantes, 3rd place Lille, and 5th place Sedan did not sell anyone to a PL club. These transfers probably did create a sense that good players could be found relatively at value in Ligue 1. But that picture changed the next window, our first peak.
While 2002/03 saw many players move from Ligue 1, most were a failure. Peripatetic Anelka joined Manchester City from PSG, although he had been on loan at Liverpool the season before, and promptly netted 14 league goals. But the two most successful moves after him were arguably Michael Svesson’s €3m move from Troyes to Southampton (he went on to make 87 appearances for the club) and Jay-Jay Okocha’s free from PSG to Bolton. The Nigerian only scored 14 and assisted 11 in 124 PL games but became a firm favourite, both of Bolton fans and neutrals, as one of the most skilful and imaginative players to put on a PL shirt. Marc-Vivien Foé, who sadly died in 2003, would certainly have gone on to play more, having made 35 appearances on loan at City. But other players who moved, including El-Hadji Diouf, Bruno Cheyrou, David Sommeil, and Salif Diao never really justified the prices paid. That could well explain the dip after the high, with a decline after clubs spent €63.5m in total for not a whole lot back. Perhaps it’s easier to argue that poor transfers impact the market more than good ones?
Let’s turn now to proportions. If the trend for Ligue 1 players is stable overall, maybe we can see swings with regards to player type? Below is a graph, using positions at time of transfer as delineated by TransferMarkt, showing percentages of transfers by defence (FB or CB), midfield (DM or CM), or attack (winger or AM or striker). I excluded goalkeepers from the graph as only 25 have made the move (7.2% of total transfers, but with no real trend except four moving in 2020/21 after three years of none – no, I don’t know either).
There are a few interesting things to note here. The first is that the trend for midfielders is consistent, at least compared with the two other units. We’ve always liked a French CM/DM; where Rémi Garde led, many others followed, and the trend remains positive. The other two positional groups are more volatile.
Let’s look at two periods: 2005/06 and 2006/07 for a defensive proportion peak in 2007/08 and 2012/13 and 2013/14 for an attacking proportion peak in 2014/15.
In 2005/06, Patrice Evra was an obvious and immediate success, and Abdoulaye Faye (loan) was a very solid pro, but Philippe Christanval, Christian Bassila, and Kelly Youga made very little impression. In 2006/07, Faye made his loan move permanent (it’s worth saying we count this twice, because we are looking at frequency and whether that suggests trends), while Benoît Assou-Ekotto (Lens to Tottenham) would go on to make 155 PL appearances. Abdoulaye Méïté was a smart acquisition by West Brom and stayed with the club through relegation and return to the PL, before going on to a rather random selection of lower level clubs. Souleymane Diawara, though, lasted a season at Charlton before leaving for France again (and making a €1.5m loss in fees). None of these, except Evra and perhaps Assou-Ekotto and Faye, immediately suggest precipitating a defensive gold rush, though. Of the defenders who moved in 2007/08, Younès Kaboul and Bacary Sagna were the outstanding successes, with 464 PL appearances between them, but the rest generally missed the mark (Habib Bèye was fine, on balance).
Turning now to 2014/15’s attacking peak: in 2012/13, we got one of the best single windows for French or France-based players. It’s a thing of beauty. Hazard, Lloris, Giroud, Azpilicueta, Debuchy, and Sissoko were unqualified successes. Loïc Rémy scored 28 PL goals in 79 appearances, while even Stéphane Mbia’s one season saw 2420 mins played. Not every transfer landed but you can see why Hazard, Giroud, and Rémy got people looking at Ligue 1 attacking talent, especially while Alan Carr’s dad was turning Newcastle into Paris-sur-Tyne.
The following season, though, was all about defence: Zouma, Sakho, Capoue, Lovren…you would expect this to preempt a surge of centre-back signings, rather than attackers. The only attackers who moved in this window were Yannick Sagbo (two goals in 15-0 minutes) and loans for FM legend Lacina Traoré (1 minute played, no goals) and Morgan Amalfitano (seven goal involvements in 28 games). Nothing about that suggests anything like the previous window, although Amalfitano did sign the next window for West Ham and contributed three goals and one assist.
But in 2014/15 (and also in 2015/16, by pure volume in the joint-third biggest window by volume for Ligue 1 players overall) the PL went big on Ligue 1 attackers, while 2015/16 arguably rivalled 2012/13 for quality, including Kanté, Payet, Cabaye, Ayew, and Gueye. Not many attackers made too much of an impression, though, although Payet obviously excelled creatively, especially from set-pieces, Origi scored crucial goals for Liverpool, and Falcao managed a healthy nine goal involvements in 1,300 league minutes for United; his Chelsea spell was not a success.
You can see a little more easily how, especially because of 2012/13, there was an increased interest in attacking players, but again, it’s probably easier to explain why the proportion of Ligue 1 attackers relative to Ligue 1 signings overall declined after this window rather than why it increased beforehand.
In the last six or seven years, Edinson Cavani has arguably been the only genuinely good Ligue 1 striker signing to join a PL club. There have been some good wingers, but also some notable disappointments. Most of the more successful players have been defenders and midfielders. It’s also worth saying that generally the market values attacking players more, and the flipside to Ligue 1 being a buyer’s market, which can reduce price, is that clubs (and players) can generate more interest and thus fees because Ligue 1 is an established destination for scouts. This possibly creates the kind of situation where Lille can extract €80m from Arsenal for Nicolas Pepé.
So, what can we conclude from all this? If we recall the quotation at the beginning of this article, we are looking for evidence that:
“Football is also a human market of bias and expectations. In recent seasons, several expensive defensive signings from Ligue 1 have signed and prospered….Every successful arrival is proof of concept and the Premier League’s reliance on Ligue 1 is likely to stay.”
Firstly, Ligue 1 was and continues to supply an average of 14 players per season to the Premier League and this trend is pretty stable. The conclusion to the quotation is therefore very likely true, but also less a reflection of recent successes and more just a stable, long-term association.
Secondly, the proportion of midfielders has also been positively trending upwards, unlike that of attackers or defenders; the former seems to be a stable group in proportional popularity, while the latter is actually on the decline, which runs counter to the prevailing narrative. Even if we break the defenders group in to centre-backs and full-backs, the trend is downward for both (you’ll have to take my word for it, because frankly there have been a lot of graphs already and there are more to come!). That suggests that maybe there is a little, understandable recency bias in the observation around defenders from Ligue 1, but again, this could also continue and disrupt the trend. Lastly, we can see that while the overall trend for attackers is stable, the up and down is greater and, as you will see below, the trend by volume is declining.
It is hard to know whether “bias and expectation” plays a part in all this – my instinct is that it does, but also, if a league demonstrates an aptitude for producing a certain kind of player then it’s not exactly bias to assume there are more examples to be found. The bias may be more around skipping some due diligence, although in this age of data scouting that seems unlikely.
I suspect too that poor performers, especially expensive ones, have a greater impact on expectations and biases than good ones – for most clubs, prioritising an aversion to risk makes more sense, especially when big sums are concerned – but that this is probably also localised a bit to clubs rather than leagues. This type of negative expectation probably explains the decrease in the proportion of attacking players signed since the high of 2015/16, especially when that group costs more than the more successful midfield and defensive groups.
My guess overall, though, is that league-wide fads are reserved for leagues that previously have been of less interest (an example would be the increased interest in J-League players following successful signings by Celtic and Brighton, as well as Belgian clubs) and that it’s less a fad than a proof of concept – this kind of player can work here, so now we should look at the league as a whole. This makes sense as a data point in itself because while you can compare league strength or physicality for example, it is hard to know whether players will settle abroad and being able to point to one or two who have is helpful. Incidentally, it’s arguably more interesting to look at where this hasn’t happened, for example the limited number of transfers from the Czech top tier despite the success of Coufal and Soucek. But Ligue 1 is so well-established, and the pathway so well-trodden by both successes and failures, that such proof of concept is unnecessary.
The last point to make is that football is obviously an industry that works on short-termism (although, as a data consultancy, we’d love it if that were not the case) and so the change season-on-season may have a greater relevance in some ways than overall trends – this would align with the quotation’s observation, but again, it’s hard to see in the data. But it’s also worth saying (and not critically – as a former sports journalist I totally get it) that this idea of bias and expectation may be as much a reflection of coverage and interest by the media as it is a real factor.
By which I mean, articles on French CBs are hot right now because several expensive defensive signings from Ligue 1 have signed and prospered, but that this reflects more a desire to find interesting things to talk about (and talk about well, to be clear – again, this is not a criticism of the piece at all) than an actual reflection of the thinking within clubs which, as evidenced by output, either suggests consistency or a slight decline in interest where defenders are concerned (and why, by the way, could that be happening? Maybe the increased interest pushes up prices which in turn depresses demand?).
As a company working in the space, we have to try to avoid any kind of bias or expectation, and solely to follow where the data or our own experience of the industry leads us. That means leaning on helpful information – players from France can clearly settle and succeed, but that doesn’t mean they will; Japanese players can do the same, so let’s explore that region more now – but without having preconceptions. Ultimately, that is hopefully the main takeaway from this piece – some transfers work and many don’t, and looking at each player on his or her merits with as much contextual information as possible is what makes it more likely a move will work, not simply because they come from a specific league or team or play a certain position.
For interest of those who have got this far, below are the position specific data by season with trend lines.
Header image copyright IMAGO / News Images / Gareth Evans
