Data sources
In this research, the data used are soccer players’ source from the International Journal of Computer Science in Sport and Ivo Hendricks, modeling the transfer of football players, which gives the details on the wages of the players as well as their skills in the field. The data used entails the player’s gross salary as agreed and signed between the club and the player’s agent. As of late, there are laid down stringent regulations by UEFA and FIFA on matters of football player’s wages, which stipulates that each football club is required to show the correct wages for its players. Football clubs that violated FIFA and UEFA regulations have, in the past, suffer because it violates regulations (Yaldo & Shamir, 2017). In that matter, the data shown by the clubs can be dependable. This report will also look into data of football players` from Der Standard, which gives the details of the global sports salaries, but more focus will be on football as a sport.
Definition of variables
Players` wages vary with the leagues, whether in La Liga, Premier League, Bundesliga, Serie A, among others. Players have a right to move from one league to another league of his choice thanks to the Bosman ruling of freedom in the year 1990 (Yaldo & Shamir, 2017). The dataset incorporates football players across different leagues.
| Variable | Description |
| Name of the player | Entails the names of the specific player in full |
| Weekly pay | The amount of money the player will be paid every week |
| position | This is the position which the player plays while on the field |
| Date of birth | The year, month and date when the player was born |
| height | The elevation of the player in centimeters (cm) |
| weight | The heaviness of the player measured in kilograms (kg) |
| Foot preference | The boot which the players use to pass the ball either left or right |
| Crossing | The capability of the player to pass a long ball probably from the wing of the field to a teammate in the penalty box |
| Finishing | Ability to score a goal |
| Accuracy of heading | The ability of a player to pass the ball by heading accurately |
| Short-passing | The ability of a player to short-pass the ball faster and precisely to a target |
| volleys | Defines the ability of a player to boot the ball precisely and accurately when it is high |
| Free kick | The accuracy that a ball will get to the opponent’s goal |
| Long passing | The ability of the player to pass long balls faster and accurately to a target |
| Ball control | States the ability of the player to control the ball whenever he comes into contact with the ball |
| Acceleration | The capability of the player to speed quickly |
| Sprint speed | Is the maximum speed that a player can run when he is in his full speed |
| Agility | Change of direction as fast as possible |
| Reactions | Defines the time the players take to adjust to a change |
| Balance | The ability of the player to stay on his feet when tackled by an opponent |
| Jumping | The highness of the player when he jumps to reach the ball |
| Stamina | Defines the distance a player can run before slowing down |
| Strength | In a bodily challenge, the capacity of a player to overwhelm a rival is defined by his strength |
| Shot-power | The ability of the player to shoot the ball with strength |
| Long shot | The capacity of a player to pass a long shot from the outside the penalty area |
| Penalties | The ability of a player to drive the ball into the net through a penalty |
| Aggression | Use of strength to win in tackling |
| Positioning | Identifying unmarked spaces in the opponents` side |
| Vision | Ability to see the positions of teammates in the field. |
| Marking | Following of opponent to ensure that they don’t get the ball |
| Standing tackle | Winning balls with no foul |
| Sliding tackle | Winning balls through tackling by sliding |
| Interception | Identification of a pass and ability to intercept the ball |
Table 1: Variables and their descriptions
These skills and capabilities are what determines the value and market price of the player. Highest paid players are playing for bigger clubs such as Barcelona, Real Madrid, Manchester United, and Juventus, among others. It is no wonder these clubs are the biggest spenders in terms of player wages, as shown by the global sports salaries report. Below is a table extract of the report. Table 2 shows the average annual and average weekly wages of the players of the selected clubs in the year 2017 (Sporting Intelligence. (n.d.)).
| Team | League | Average annual | Average weekly |
| Barcelona | La Liga | £10,454,259 | £201,043 |
| Real Madrid | La Liga | £8,089,582 | £155,569 |
| Juventus | Serie A | £6,726,615 | £129,358 |
| Manchester United | English Premier league | £6,534,654 | £125,666 |
| Cardiff | EPL | £957,471 | £18,413 |
Table 2: Average wages annually and weekly in 2017
Econometric modeling.
Definition of data.
We are focused on predicting the earnings of famous and non-famous players in the next window of transfer.to model this, then we need to find a way to see how y varies with a change in x. y denotes the price while x denotes the interest variables (Hendriks, I., 2017)
(1)
(2)
is the parameter of our interest since it holds the strength and the course of the correlation of x
(3)
Having a detail on x and y, the method of least squares gives that it reduces the total squared errors in its estimation (Hendriks, I., 2017).
Residual sum of squares thus,
RSS= (4)
Re-writing the above equation we get,
RSS= – )2 (5)
Characterize the solutions s β0 and βi in minimization of the problem,
argminb (6)
and are the arguments for the minimization problem,
This results in Least Squares estimate βbLS (Hendriks, I., 2017)
Results and discussion
When modeling, there are challenges in the heterogeneity of the players and the heterogeneity of the club. One has to further distinguish a player with his abilities compared to the other. Also, one has to observe the difference between the clubs. There is variance in terms of revenues among the different clubs. Additional variables may be included in determining the price of the player in terms of fame. In making the above modeling, there are assumptions put in place. One is that the above model is linear parametrically βi and flexible since we can alter y and xi. Another hypothesis is that the data modeled represents a valid population. Also, independent variables do not vary, and constant exists no correlation between the independent variables
Yaldo, L., & Shamir, L. (2017). Computational Estimation of Football Player Wages. International Journal of Computer Science in Sport, 16(1), 18-38. doi:10.1515/ijcss-2017-0002
Sporting Intelligence. (n.d.). Retrieved May 8, 2020, from https://www.globalsportssalaries.com/
Hendriks, I. (2017). Modeling the transfer prices of football players (Unpublished master’s thesis). Thesis / Dissertation ETD.