Wonderkids: 2021 Outlook

Nothing captures the imagination of a football fan quite like the emergence of wonderkids, players surrounded by both high hopes and expectations, and significant uncertainty. Predicting the career trajectory of a young footballer is an extremely complex exercise, bearing in mind the many variables that can suddenly derail even the most obvious of career paths. At Soccerment we try to manage such complexity by building data-driven models and – as described in this analysis – applying them to football players born in 2001 or later. We then rank these players for their expected performance over the next 12 months, with our ultimate goal being to offer insights into the most exciting young players in Europe.

Does football hold any bigger thrill than a wonderkid?

As rightly pointed out by Ryan Baldi in his recent book ‘The Next Big Thing‘, “nothing captures the imagination of the football fan in quite the same way as the emergence of a prodigious young player“.

At the end of 2019, Erling Braut Haaland was a promising young Norwegian forward playing in the Austrian league. Fast forward twelve months, and Haaland is regarded by the football world as an infallible scoring machine (and rightly so, having scored 35 goals in 34 appearances for Borussia Dortmund), has won the Golden Boy award, is the object of the biggest club’s most heartfelt (transfer!) desires, and has millions of social media followers. Haaland revealed himself as a “wonderkid” and since then just hasn’t stopped beating the higher and higher expectations he created for himself.

‘Wonderkid‘ is a term that we derived from the videogame Football Manager, one of our key sources of inspiration. It identifies young players with high potential. We like the concept so much that in our algorithmically-derived player descriptions, we have found space for wonderkids. In our Player Smart Search (click here for more information), the user can easily select the category “wonderkids” and deep-dive into the shortlist (spoiler: at the time of writing, Haaland is on the list).

Unlike Haaland’s experience, Ryan Baldi’s aforementioned book focuses on the personal stories of fifteen wonderkids who did not make it, young players whose destiny was to “burn brightly and fade quickly“. The author explores each case separately and explains all the things that can go wrong in the early career of a new football star: injuries, excess, bad relationships with managers and bad transfer choices are just a few of the things that can derail a wonderkid from the path to success and glory. This fascinating book exposes just how difficult it is to analyse the potential of these players.

We try to embrace the extreme complexity of predicting the career trajectory of a young footballer by building data-driven models. In the following analysis, we apply these models to football players born in 2001 or later, and rank them for their expected performance over the next 12 months. Just to clarify, with this exercise and the resulting ranking, we are not trying to predict which young footballers will have the brightest careers: the forecast is indeed limited to the calendar year ahead.

[ec_product productid=”12″ style=”3″]

The project

Our ‘Wonderkids: 2021 Outlook‘ started out as an internal game: the initial idea was to try to forecast the 50 players on the shortlist of the Golden Boy, an award given by sports journalists to the young footballer playing in Europe considered to have been the most impressive during the calendar year (two halves of two separate seasons). The prize, established by Italian sports newspaper ‘Tuttosport’ in 2003, includes jurors from ‘Bild’ (Germany), ‘Blick’ (Switzerland), ‘A Bola’ (Portugal), ‘l’Équipe’ (France), ‘France Football’ (France), ‘Marca’ (Spain), ‘Mundo Deportivo’ (Spain), ‘Ta Nea’ (Greece), ‘Sport Express’ (Russia), ‘De Telegraaf’ (Netherlands), and ‘The Times’ (United Kingdom). All nominees for the Golden Boy must be under the age of 21 and play in a European nation’s top tier league.

What was intended to be a “game”, became motivation to develop our models that predict football players’ performances over a certain time frame.

Developing models like this is not an easy task. Developing them so that they are useful when applied to young footballers is even more difficult. There are two important reasons for this:

the first one, already mentioned, has to do with the very high volatility that young players’ performances are subject to.
the second reason is that, in the case of Under 20 players, the historical stats available for analysis is very limited as the data for footballers before they reach the top tiers of professional football is scarce (this is a problem we are trying to solve by developing scalable, affordable and user-friendly football data-gathering solutions, such as this).

Nonetheless, we have built a simplified version of our model and applied it to the players in our database who were born in 2001 or later. This simple model was developed to form the ranking of the top 50 wonderkids for 2021.

Later in the analysis we list the top wonderkids, from 50th to 1st, depending on the value of their Expected Soccerment Performance Rating (‘xSPR’).

At the end of the year we will be able to check whether:

there is any correspondence between our list and 2021 Golden Boy shortlist;
the Expected Soccerment Performance Ratings (‘xSPR’) that we use to rank the players in this document is more or less in line with the players’ SPR at the end of the year.

As you might have already guessed, our ‘Wonderkids: 2021 Outlook’ will be followed by a ‘Wonderkids: 2021 Review” sometime in December 2021.

The methodology

[Click here to skip the methodology and jump to the final ranking]

Initial list of around 100 players

We took into consideration all players who were born on 1st January 2001 or after and who – as of 31st December 2020 – had played at least 500 minutes in total and at least 270 minutes in the current season, in one of the fourteen leagues we cover (Tiers 1 and 2 in England, Spain, Italy, Germany and France, and Tier 1 in the Netherlands, Portugal, Belgium and Turkey).

The playing time constraints, which ensure statistical relevance, drastically reduced the number of players that we could include in our Outlook, which means that we had to exclude some very interesting prospects. Here below is a partial list of some of the main names we were forced to leave out.

Player	Club	Reason for exclusion
Youssoufa Moukoko	BVB	< 270 mins in 2020/21 & < 500 mins tot
Amad Diallo	MUN	< 270 mins in 2020/21 & < 500 mins tot
Joshua Zirkzee	BAY	< 270 mins in 2020/21 & < 500 mins tot
Xavi Simons	PSG	< 270 mins in 2020/21 & < 500 mins tot
Billy Gilmour	CHE	< 270 mins in 2020/21 & < 500 mins tot
William Saliba	ARS/NIC	< 270 mins in 2020/21
Eric Garcia	MCI	< 270 mins in 2020/21
Eduardo Quaresma	SCP	< 270 mins in 2020/21
Rayan Cherki	LYO	< 500 mins in total
Jamal Musiala	BAY	< 500 mins in total
Fabio Silva	WOL	< 500 mins in total

List of some of the most interesting U-19s that were excluded from our Outlook, due to time constraints | Data related to the leagues we cover, as of 31.12.2020

Ranking methodology

The methodology by which we arrived at the wonderkids’ final ranking was divided into two stages.

First step: selecting the best 50

The first stage consisted in ranking the initial list of around 100 players by using our Soccerment Performance Rating (‘SPR’), the algorithm we have developed to evaluate the performances of football players.

Soccerment Performance Rating

The SPR takes into account every match event, weighted through ad-hoc coefficients.

The various events (rebased to ‘per 90 minutes’) form the contribution of a player to the defensive, buildup and attacking phase of her/his team. These contributions are then weighted according to the player’s role (for example, the attacking contribution will weigh more for a forward and less for a defender).

In line with what we do for the overall stats, we show the SPR of a player when she/he has accrued at least 270 minutes of playing time.

In order to reward consistency of performance and limit short-termist hype, the SPR is adjusted by taking playing time into consideration. At 270 minutes of playing time the adjustment is 84.7%. It reaches 100% when the footballer has played at least 1,800 minutes in the league (in most leagues, this equals half a season, plus one match). Therefore, every additional minute above 270 and up to 1800 minutes is worth 0.01% of SPR.

It is worth mentioning that the five players with the highest SPR at the moment are (in terms of total SPR, i.e. the weighted-average SPR for all the seasons available in our database):
➜ Lionel Messi, with 84
➜ Neymar Jr, with 76
➜ Cristiano Ronaldo, with 75
➜ Robert Lewandowski, with 70
➜ Kylian Mbappé, with 70

A total SPR of 70 is the threshold that we use to define the so called “world-class” players, while for the “elite players” the threshold was set at 60.

It is also worth mentioning that the maximum level of a seasonal SPR was set at 100 and corresponds to the performances reached by Lionel Messi in the 2011/12 La Liga: 50 goals (plus 8 shots against the woodwork), 16 assists (from 92 chances created), 176 successful dribbles, almost 2000 accurate passes, plus other kind-of-unreal stats.

We thus easily filtered the top 50. Here are some statistics regarding the top 50 list:

the average age is 18.6 years;
the most represented leagues, with 8 players each, are: La Liga, Ligue 1, and the Jupiler Pro League;
the most represented nationality is French (8 players);
the most represented clubs are Rennes and Club Brugge (3 players each).

Second step: ranking the top 50

The second step in our analysis was to find a quantitative methodology to rank the top 50 players, excluding any subjective inputs.

In order to do that, we built a simple model, which included all the key variables and indicators at our disposal to allow the assessment of both the “internal” and “external” factors that influence player development, especially when compared to players’ typical performance curves over their entire careers. These factors combine to define the outperformance or the underperformance of a player’s career trajectory, when compared to her/his peers. But first, let’s define this trajectory, or what we call the ‘performance curve’.

Performance curves

Footballers’ performance output grows substantially over the beginning of their careers, until they peak at around 28-30 years of age and then start to gradually decline.

Graphically, it looks something like the chart here below, which we originally published on our Instagram account on 23rd December 2020, where we plot players’ performance output (as expressed by our Soccerment Performance Rating) and their age. In the chart we highlight the names of the main outliers by age group.

The performance curves, however, differ substantially depending on the playing roles. It is pretty intuitive to consider, for instance, how goalkeepers can play at a high level when they are 35 years old or older, while the same cannot usually be said about central midfielders (there are obviously exceptions to the rule, like these 33 players – to view the full list a free signup is needed).

[ec_product productid=”12″ style=””]

The differences in annual growth rates for performance output, as expressed by our Soccerment Performance Rating (‘SPR’), for players who are younger than 20 years old and depending on their role is shown in the matrix below. The matrix was derived by looking at the historical stats in our database and calculating the average yearly growth rates for each age group depending on the players’ roles.

Age	Defenders	Midfielders	Forwards
16 yo	+8%	+13%	+10%
17 yo	+6%	+12%	+9%
18 yo	+5%	+10%	+7%
19 yo	+4%	+9%	+6%
20 yo	+3%	+8%	+5%

Matrix with the yearly average growth rates of the players’ SPR, broken down by role and age (GK included in Defenders)

For a quick explanation of how to read the matrix, here’s an example: on average, a midfielder who is today 18 years old and shows a SPR of 50, will experience a 10% improvement over the next 12 months, when the player’s SPR will thus reach 55. An average player who is 18 too and shows a SPR of 50 as well, but who is a defender, will see her/his SPR grow by 5% and the SPR will reach 52.5 in the next 12 months.

Factors leading to under- or outperformance of the average growth rates

Our model, as mentioned above, considers a number of factors that can influence relative improvements compared to the role average.

When it comes to “player-specific” factors, we use the following variables:

SPR’s recent trajectory. This gives an indication of how the player is reacting to the increased level of attention and scrutiny. In some cases, young players are able to continue meeting the higher expectations. In other cases, especially following a breakout season, the players experience a performance decline, which needs to be carefully monitored and managed.
UCL exposure. Playing in the Uefa Champions League, in our view, is not only an achievement for a young footballer, it is first and foremost an occasion to grow and improve by experiencing the highest level of football, which gives them valuable experience. We take it into consideration in our model.
Body Mass Index (BMI). We use this as an indicator of a player’s current muscularity and its potential growth over the following few years. When a young player approaches senior football and the training sessions that come with it, muscular growth is often impressive and with the added strength and explosiveness performances normally improve. Some noteworthy examples are the transformations experienced by the likes of Adama Traoré and Leon Goretzka.

As for external factors, i.e. the so called ‘context’, we took into consideration the following:

The player’s league level (average SPR for the league) and the player’s team level (the team’s SPR). These are useful indicators of how brightly the player’s star can shine, when taken either singularly or in combination (it is easier to do well and gain confidence when playing for team that finishes up the table).
Playing time as a percentage of availability (adjusted for the games missed due to injury). This is useful to understand whether the player fits well into a team and the team’s manager appreciates their contribution.
Percentage of times in the starting lineup. Again, a useful indicator of how well the manager considers the player.
Percentage of times in the correct role. This mainly tells whether the footballer is playing in a team that uses a suitable formation and/or fits well in the squad.

Final ranking and Top 11

The final ranking of Under 19 players, as well as the U-19’s Top 11 selection, is based on the Expected SPR in twelve months, that we call xSPR.

Final ranking

Here below the top 50 players, ranked by the fiftieth to the first. Click here to skip the detailed list and jump to the summary table.

50th. Andrea Papetti (18 | Italy | Brescia Calcio | xSPR: 37.7)

Andrea Papetti is a 18-year-old centre-back who plays for Brescia Calcio in the Italian Serie B. He’s 189 cm tall and weighs 84 kg. Key skill: Heading (he has already scored 2 headers in 22 senior career appearances).

Rank	Player	Age	Nat.	Role	Club	SPR	xSPR
1	A. Fati	18	🇪🇸	FW	FCB	49	58.5
2	B. Badiashile	19	🇫🇷	CB	AMO	53	58.0
3	R. Gravenberch	18	🇳🇱	CM	AJX	47	57.7
4	L. Traoré	20	🇧🇫	FW	AJX	47	53.7
5	C. Jones	19	🏴󠁧󠁢󠁥󠁮󠁧󠁿	CM	LIV	49	53.6
6	E. Camavinga	18	🇫🇷	CM	REN	45	52.6
7	F. Wirtz	17	🇩🇪	CAM	B04	44	51.8
8	O. Kossounou	20	🇨🇮	CB	BRU	45	51.5
9	J. Doku	18	🇧🇪	WAM	REN	42	51.3
10	F. Daniliuc	19	🇦🇹	CB	NIC	48	51.2
11	G. Reyna	18	🇺🇸	CM	BVB	42	51.1
12	B. Saka	19	🏴󠁧󠁢󠁥󠁮󠁧󠁿	WAM	ARS	42	49.4
13	M. Boadu	20	🇳🇱	FW	AZA	49	49.2
14	N. Madueke	18	🏴󠁧󠁢󠁥󠁮󠁧󠁿	FW	PSV	44	49.2
15	R. Aït-Nouri	19	🇫🇷	FB	WOL	47	48.4
16	M. Ihattaren	18	🇳🇱	CAM	PSV	45	47.9
17	M. Greenwood	19	🏴󠁧󠁢󠁥󠁮󠁧󠁿	FW	MUN	46	47.8
18	J. Bellingham	17	🏴󠁧󠁢󠁥󠁮󠁧󠁿	CM	BVB	36	47.8
19	A. Truffert	19	🇫🇷	FB	REN	44	47.3
20	Pedri	18	🇪🇸	CM	FCB	38	46.8
21	N. Collins	19	🇮🇪	CB	STK	40	46.3
22	Rodrygo	20	🇧🇷	WAM	RMA	43	46.3
23	N. Mendes	18	🇵🇹	FB	SCP	41	46.2
24	H. Elliott	17	🏴󠁧󠁢󠁥󠁮󠁧󠁿	WAM	BLK	41	46.0
25	A. Aouchiche	18	🇫🇷	CAM	ASSE	41	45.8
26	B. Gil	19	🇪🇸	WAM	EIB	40	45.7
27	N. Raskin	19	🇧🇪	CM	STA	41	45.4
28	A. Hickey	18	🏴󠁧󠁢󠁳󠁣󠁴󠁿	FB	BOL	42	45
29	N. Rovella	19	🇮🇹	CM	GEN	39	43.7
30	M. Thiaw	19	🇫🇮	CB	S04	41	43.2
31	I. Ilic	19	🇷🇸	CM	VER	39	43.1
32	A. Broja	19	🇦🇱	FW	VIT	40	43.1
33	K. Sardella	18	🇧🇪	CB	AND	44	42.8
34	E. Destanoglu	20	🇹🇷	GK	BES	42	42.8
35	T. Coulibaly	19	🇫🇷	WAM	VFB	39	42.6
36	C. de Ketelaere	19	🇧🇪	FW	BRU	35	42.4
37	T. Kubo	19	🇯🇵	WAM	GET	44	42.4
38	C. Devenish	19	🇨🇴	CB	BOA	40	41.9
39	E. Bilal Touré	19	🇲🇱	FW	REI	40	41.6
40	Y. Verschaeren	19	🇧🇪	CM	AND	38	41.2
41	B. Diakite	20	🇫🇷	CB	TOU	44	41.1
42	I. Kabore	19	🇧🇫	FB	MEC	40	41.0
43	L. Kang-In	19	🇰🇷	WAM	VAL	37	40.8
44	N. Katterbach	19	🇩🇪	FB	KOL	42	38.8
45	Y. Badij	19	🇸🇳	FW	BRU	36	38.7
46	Y. Musah	18	🇺🇸	WAM	VAL	35	38.7
47	A. Vranckx	18	🇧🇪	CM	MEC	34	38.7
48	J. Chotard	19	🇫🇷	DMC	MPL	40	38.6
49	A. Barrenetxea	19	🇪🇸	FW	RSO	37	38.4
50	A. Papetti	18	🇮🇹	CB	BRE	39	37.7

Does football hold any bigger thrill than a wonderkid?

The project

The methodology

Initial list of around 100 players

Ranking methodology

First step: selecting the best 50

Second step: ranking the top 50

Performance curves

Factors leading to under- or outperformance of the average growth rates

Final ranking and Top 11

Final ranking

50th. Andrea Papetti (18 | Italy | Brescia Calcio | xSPR: 37.7)

49th. Ander Barrenetxea (19 | Spain | Real Sociedad | xSPR: 38.4)

48th. Joris Chotard (19 | France | Montpellier | xSPR: 38.6)

47th. Aster Vranckx (18 | Belgium | Mechelen | xSPR: 38.7)

46th. Yunus Musah (18 | USA | Valencia CF | xSPR: 38.7)

45th. Youssouph Badij (19 | Senegal | Club Brugge | xSPR: 38.7)

44th. Noah Katterbach (19 | Germany | 1 FC Köln | xSPR: 38.8)

43rd. Kang-In Lee (19 | Korea Republic | Valencia CF | xSPR: 40.8)

42nd. Issa Kabore (19 | Burkina Faso | Mechelen | xSPR: 41.0)

41st. Bafode Diakite (20 | France | Toulouse | xSPR: 41.1)

40th. Yari Verschaeren (19 | Belgium | Anderlecht | xSPR: 41.2)

39th. El Bilal Touré (19 | Mali | Reims | xSPR: 41.6)

38th. Cristian Devenish (19 | Colombia | Boavista | xSPR: 41.9)

37th. Takefusa Kubo (19 | Japan | Getafe | xSPR: 42.4)

36th. Charles de Ketelaere (19 | Belgium | Club Brugge | xSPR: 42.4)

35th. Tanguy Coulibaly (19 | France | VfB Stuttgart | xSPR: 42.6)

34th. Ersin Destanoglu (20 | Turkey | Besiktas | xSPR: 42.8)

33rd. Killian Sardella (18 | Belgium | Anderlecht | xSPR: 42.8)

32nd. Armando Broja (19 | Albania | Vitesse | xSPR: 43.1)

31st. Ivan Ilic (19 | Serbia | Hellas Verona | xSPR: 43.1)

30th. Malick Thiaw (19 | Finland | Schalke 04 | xSPR: 43.2)

29th. Nicolò Rovella (19 | Italy | Genoa | xSPR: 43.7)

28th. Aaron Hickey (18 | Scotland | Bologna | xSPR: 45.0)

27th. Nicolas Raskin (19 | Belgium | Standard | xSPR: 45.4)

26th. Bryan Gil (19 | Spain | SD Eibar | xSPR: 45.7)

25th. Adil Aouchiche (18 | France | Saint-Étienne | xSPR: 45.8)

24th. Harvey Elliott (17 | England | Blackburn | xSPR: 46.0)

23rd. Nuno Mendes (18 | Portugal | Sporting | xSPR: 46.2)

22nd. Rodrygo (20 | Brazil | Real Madrid | xSPR: 46.3)

21st. Nathan Collins (19 | Republic of Ireland | Stoke | xSPR: 46.3)

20th. Pedri (18 | Spain | FC Barcelona | xSPR: 46.8)

19th. Adrien Truffert (19 | France | Rennes | xSPR: 47.3)

18th. Jude Bellingham (17 | England | Borussia Dortmund | xSPR: 47.8)

17th. Mason Greenwood (19 | England | Man Utd | xSPR: 47.8)

16th. Mohamed Ihattaren (18 | Netherlands | PSV | xSPR: 47.9)

15th. Rayan Aït-Nouri (19 | France | Wolves | xSPR: 48.4)

14th. Noni Madueke (18 | England | PSV | xSPR: 49.2)

13th. Myron Boadu (20 | Netherlands | AZ | xSPR: 49.2)

12th. Bukayo Saka (19 | England | Arsenal | xSPR: 49.4)

11th. Giovanni Reyna (18 | USA | Borussia Dortmund | xSPR: 51.1)

10th. Flavius Daniliuc (19 | Austria | Nice | xSPR: 51.2)

9th. Jeremy Doku (18 | Belgium | Rennes | xSPR: 51.3)

8th. Odilon Kossounou (20 | Côte d’Ivoire | Club Brugge | xSPR: 51.5)

7th. Florian Wirtz (17 | Germany | Bayer 04 | xSPR: 51.8)

6th. Eduardo Camavinga (18 | France | Rennes | xSPR: 52.6)

5th. Curtis Jones (19 | England | Liverpool | xSPR: 53.6)

4th. Lassina Traoré (20 | Burkina Faso | Ajax | xSPR: 53.7)

3rd. Ryan Gravenberch (18 | Netherlands | Ajax | xSPR: 57.7)

2nd. Benoit Badiashile (19 | France | Monaco | xSPR: 58.0)

1st. Ansu Fati (18 | Spain | FC Barcelona | xSPR: 58.5)

Summary table

Top 11: using a 3-4-1-2 formation

Conclusions