In the other football, there are some established best practices for NFL teams trying to acquire new players. Teams don’t necessarily abide by them, but there’s enough data and history to figure out what typically works and what doesn’t.

In the college draft, you only want to pick premium-position players early on, i.e. the players at the positions that have the highest average salaries in the league: quarterback, offensive tackle, and pass rusher. And unless you’re trying to acquire a quarterback, you don’t want to trade away multiple draft picks to acquire a higher draft pick. Instead, you want to trade down to acquire more picks because over the long run, no one is better at drafting players than anyone else. The best drafting teams are the teams that give themselves more chances to draft.

The clear positional delineations, plus established salary and roster rules, make these kinds of distinctions much easier in American football. In soccer, though, everyone plays offense and defense, there are barely any scoring plays, salary info isn’t publicly available, there are hundreds of professional leagues, and budgets across and within leagues are drastically different.

But what if we looked at the top 100 transfers in global soccer from the past four years — the 25 highest fees paid in each season — to see if there were any trends? What were the worst and best transfers? And what can they tell us about which moves are likely to succeed and doomed to fail?


How to measure a successful transfer

There are all kinds of definitions of success when it comes to a transfer: minutes played, trophies won, goals scored, shirts sold, Instagram followers accrued, etc. But none of those take into account the fact that every club — including even those run by sovereign wealth funds — have a limited amount of money they’re capable of spending within the vague, ever-changing rules instituted by the sport’s various governing bodies.

If you spend $100 million on a player, it means that you didn’t spend $100 million on two or three players. There are opportunity costs to every move, and while it’s a somewhat cold distillation of how the sport works, there’s a simple formula that determines team success: money spent on players, multiplied by the value provided for every dollar (or euro or pound) spent.

The more you spend, the more inefficient you can be. The less you spend, the more efficient you have to be. But even the richest clubs benefit from being efficient with where their money goes. Real Madrid’s new era of dominance has been driven largely by a new, efficient approach in the transfer market. Manchester City rarely make truly disastrous transfer moves. And Liverpool’s run of success under manager Jurgen Klopp was largely driven by a sensational stretch of player acquisition.

To define efficiency, we’re going to compare the fee a team paid for a player with the player’s estimated transfer value via Transfermarkt a year after the player was acquired, and the more efficient a move, the better it will rank.

This doesn’t account for wages — a large blind spot in the process — and factors other than “how good the player is” determine transfer values, but there should still be plenty of insight from a large sample of transfer fees. The crowd-sourced fees on Transfermarkt tend to correlate pretty well with actual transfer fees, but the farther away you get from an actual transfer, the more estimated and inaccurate the market values are likely to be.

So, if a team signed a player for a transfer fee of, say €40 million and his performance or larger economic factors pushed his market value to €70 million a year later, we’re grading that as a fantastic move. Credit to NFL analyst Kevin Cole, who helped me come up with this framework. All transfer fees mentioned come from Transfermarkt.


What were the worst and the best moves?

We’ll start with the top five:

1) Erling Haaland, 21 years old, Manchester City
€60m from Borussia Dortmund, €180m market value a year later: 200% increase

2) Jude Bellingham, 17 years old, Borussia Dortmund
€30.15m from Birmingham City, €70m market value a year later: 132% increase

3) Nicolò Barella, 23 years old, Inter Milan
€32.5m from Cagliari, €70m market value a year later: 115% increase

4) Eduardo Camavinga, 18 years old, Real Madrid
€31m from Rennes, €60m market value a year later: 93.5% increase

5) Enzo Fernández, 21 years old, Benfica
€44.25m from River Plate, €80m market value a year later: 80.8% increase

What do all of these players have in common? Well, four of them are central midfielders, but all five of them were acquired before they entered their prime years (24 to 28).

Haaland’s low fee was driven by a release clause, and according to estimates from the site FBref, he’s the second-highest-paid player in the Premier League (about €23 million per year). Despite all of that, the deal has been a bargain. Haaland is arguably the most valuable player in the world, and he’s already given Man City two of the greatest-ever seasons by a Premier League attacker.

If I had to theorize about the others, I would say that midfielders who play at a young age are an especially good bet to become stars. There’s more certainty, perhaps, because midfielders touch the ball so much. You get a larger sample size of actions to judge a player on, and for that reason, a younger midfielder could more negatively affect his team’s performance than players at other positions, so the guys who are playing are probably pretty good. Managers can risk a talented but inconsistent young forward or winger because they don’t touch the ball as much, but a midfielder who can’t pass or keep possession will tank his team’s performance.

As for the worst moves, here’s the bottom five in descending order:

5) Wesley Fofana, 21 years old, Chelsea
€80.4m from Leicester City, €40m market value a year later: 50.2% decrease

4) Randal Kolo Muani, 24 years old, PSG
€95m from Eintracht Frankfurt, €45m market value a year later: 52.6% decrease

3) Robert Lewandowski, 33 years old, Barcelona
€45m from Bayern Munich, €20m market value a year later: 55.6% decrease

2) Marc Cucurella, 24 years old, Chelsea
€65.3m from Brighton, €28m market value a year later: 57.1% decrease

1) Casemiro, 30 years old, Manchester United
€76.5m from Real Madrid, €30m market value a year later: 60.8% decrease

What these moves have in common is that most of the fees are massive, and most of the players are either already in their primes or well beyond them. With Fofana, Chelsea paid a fee that meant he would have to become one of the handful of best center-backs in the world for the market value to match up. Same goes for Cucurella among full-backs — and for Kolo Muani, among forwards. None of these players were close to sure-thing stars, but the fees all suggested they were.

Lewandowski and Casemiro, meanwhile, involved large fees for older players well beyond their primes. Neither move had any real chance of aging well, and neither one has.


What can we learn about how transfers work?

On average, these 100 moves on the most-expensive list have seen a 1.8% depreciation a year after the transfers were made.

So, if you’re beating that number with a transfer, you’re beating the field among premium transfers. Forty-three of these moves either maintained their value a year later or saw an increase. Liverpool and Manchester City lead the way with four each: Luis Díaz, Diogo Jota, Dominik Szoboszlai, and Ibrahima Konaté for the former, and Haaland, Ferran Torres, Jérémy Doku, and Rúben Dias for the latter.

However, City have a bunch of other negative-value moves, and Liverpool lost a bunch of value (25%) from the Darwin Núñez transfer. Real Madrid, meanwhile, have three top-100 moves over the past four years, and they’re all largely positive. The acquisitions of Camavinga, Bellingham, and Aurélien Tchouaméni saw an average value increase of 60.2%.

At the other end of the spectrum, Manchester United have made nine moves in this top 10 and only one of them (Raphaël Varane, from Real Madrid) led to an increase in value after a year. Taken together, the nine acquisitions lost an average of 26.7% in value after a year. Chelsea, meanwhile, lead the way with 13 top-100 moves. Despite a more-than-70% gain from acquiring Cole Palmer last summer, they’ve still averaged a 19.4% loss on those deals.

Could Man United tempt Alphonso Davies to join the club?

Gab Marcotti debates the future of Alphonso Davies if he doesn’t join Real Madrid this summer.

But what actually drives these values? Are there any factors that make a transfer more likely to succeed? And can teams target players that check certain boxes in order to increase their chances of success?

It would appear that central midfielders are more likely to be successful transfers than any other position. These players have seen a rise of 8.8% in value a year after joining their new clubs. Center-backs saw a 0.4% decrease — a decrease to be sure, but still better than the average we alluded to earlier. Every other position declined by a higher amount than the 1.8% baseline.

However, the statistical relationship between the year-later value and a player’s position isn’t strong enough for us to be confident that it’s not just random noise. I like the theory I posed earlier about midfielder touches, but I wouldn’t be willing to bet millions of dollars on it — and that’s what these teams are doing every time they dip into the transfer market. Instead, the main drivers appear to be age, fee paid, and the player’s market value on Transfermarkt at the time of the deal.

With age, players 21 and under saw a 17.2% increase in value, while players older than 21 saw an 8.1% decrease. Framing it a little more equally, players 24 and under had values that increased by 2.8% while players 25 and older decreased by 13.2%

The relationship to fee is pretty simple: the more you pay, the less likely the deal is to age well. For deals more than €50 million we have seen an average 12.4% decrease in value. For deals of €50 million or less: an 8.3% increase in value.

The opposite is true with market value. The players valued at more than €50 million saw a 6.4% increase in value, while those at €50 million or below saw a 5.8% decrease. That might seem confusing, but the players with a higher market value are more likely to be an under-pay by our definition (fee below market value a year after the deal), while the opposite is true for players with a lower market value.

We can then use these little insights to project how likely a move is to work out. I won’t bore you with the formula here, but we’ll use both of Bayern Munich’s moves this summer as an example:

Michael Olise: 22, €55m market value, €53m fee
João Palhinha: 29, €55m market value, €51m fee

The Olise deal is projected to see a 20% increase in value, while the Palhinha move is projected to decline by 18.4%. These are just estimates, but you can see that while both moves are slightly below the market values, the age factor massively outweighs those minor differences.

Real Madrid know this, and so do Liverpool. If you want to win the transfer market, sign young players with sizable professional track records for below-premium prices. It sounds easy, but if it really was, wouldn’t everyone else already be doing it?