What Determines the Market Value for a Winger in Football?

Introduction

In the last decade, the footballing industry has evolved its approach towards the signing of players. Traditionally, football clubs allocated their control with the manager who would supervise the signings of players; however, clubs have begun by enlisting roles such as sporting directors, who oversee the structure of clubs with specialisation of data (Parnell et al., 2022). Data has been recognised as an integral part of a club’s structure and provides a compelling and nuanced perspective to a team’s success (Memmert and Raabe, 2018).

This research attempts to target empirical analysis specifically of a winger—one who plays in wide areas of the pitch, dribbles past players and crosses the ball into the box (Aalbers and Van Haaren, 2019). This approach has been influenced by alternative sporting advancement, particularly baseball, through sabermetrics, pioneered by Bill James and widely adopted by the Oakland Athletics in the early 2000s.



Methodology and Data

Data Sources

  • Transfermarkt: Used for player market values and general player data.
  • Fbref: Provides in-depth statistics like goals, assists, and defensive metrics.
  • Opta League Rankings: Determines the competitiveness of different leagues.

Multiple Linear Regression (MLR)

A multiple linear regression (MLR) was conducted to understand the impact of different key performance indicators (KPIs) on a winger’s market value. The regression equation used:

Market Value = β0 + β1(Age^2) + β2(Goals & Assists) + β3(Expected Values) + ... + Error

Regression Assumptions Checked

  • Linearity (scatterplots were examined)
  • Homoscedasticity (residual vs. fitted plot)
  • Normality (Q-Q plot of residuals)
  • Multicollinearity (Variance Inflation Factor analysis)



Results

Once having a finalised model with all variables, interpretation of the regression model can now be initiated. As shown below, there is a combination of positive and negative coefficients associated with the market value. Additionally, this consisted of several statistically insignificant variables with “high” p-values. This also contained variables deemed as statistically significant with some as partially significant. The correlation between market value and the independent variables is also included to understand simple relationships between the inputs.

Estimate Std. Error p-value
(Intercept) -190.1 91.53
I(age^2) -0.036 0.02
g_and_a 34.45 17.68
npxgxa 16.12 29.58
pass 0.65 0.55
cross -6.092 13.25
take_on 5.76 2.97
def -4.3 2.23
I(minutes^2) 0.000005 0.0000009
lrank 1.84 0.83



Results from regression model

As shown in the table above, both minutes played and league ranking were deemed as statistically significant, both with positive relationships. Whilst both of these characteristics were positive, the estimates for the number of minutes played (0.000005) was considerably lower than league rankings (1.84), suggesting that the difficulty in the league has a much greater effect on market value than playing time. Other characteristics such as goals and assists, take-ons, and defending were all slightly higher than the significance level. These had p-values just greater than 0.05, but less than 0.06, allowing conclusions to still be made. As shown above in Table 4, goals and assists had the greatest impact on market value with an estimate of 34.45. This suggests there is an expected change of €34.45 million for a one-step increase in goal contributions. This reflects the importance of this performance metric when analysing the market value. Figure 2 below represents the positive impact this has, with a strong positive correlation (0.59).



Key Findings

  • Goals & Assists: Strongest positive correlation with market value (+€34.45M per additional contribution per 90 mins).
  • Take-ons (Dribbles): Moderately positive correlation (+€5.76M per successful dribble per 90 mins).
  • Defensive Actions: Negatively correlated with market value (-€4.30M per defensive contribution per 90 mins).
  • Expected Goals & Assists: Statistically insignificant.
  • League Ranking: Strong positive correlation, indicating playing in a tougher league increases market value.



Regression Output

Variable Estimate Std. Error p-value
(Intercept) -190.1 91.53 0.042
Age^2 -0.036 0.02 0.072
Goals & Assists 34.45 17.68 0.056
Expected Values 16.12 29.58 0.59
Take-ons 5.76 2.97 0.057
Defending -4.3 2.23 0.059



Visualisations

Market Value Analysis for Wingers

Goals & Assists vs. Market Value

Defensive Actions vs. Market Value

Take-ons vs. Market Value

Discussion

Comparing Findings with Literature

  • Confirms traditional KPIs for wingers (goal contributions, dribbling).
  • Challenges contemporary claims on defending as an asset for market value.
  • Expected Goals (xG) remains underutilised in market valuations.

Broader Implications for Football Clubs

The results indicate that clubs continue to prioritise attacking output when determining market value, even as modern football places growing emphasis on tactical flexibility and pressing. While defensive contributions appear to have a negative correlation with market value, this could suggest that defensive work is either undervalued for wingers or expected as part of their role without explicitly influencing transfer fees. Additionally, the insignificance of xG and xAG in predicting market value raises questions about whether these newer metrics are fully integrated into transfer negotiations or if clubs still lean on traditional output measures like goals and assists.

Another key takeaway is the impact of league ranking. Playing in a stronger league significantly increases a player’s market value, reinforcing the importance of competition level in evaluating player worth. This suggests that clubs looking for undervalued talent might find opportunities in lesser-ranked leagues where high-performing players could be acquired at a lower cost.

Furthermore, while traditional metrics remain dominant in valuation, there is a growing shift towards deeper statistical analysis. Clubs increasingly consider factors such as player versatility, pressing intensity, and tactical adaptability, which may not yet be fully reflected in transfer fees but are becoming crucial in player assessments. This suggests that future player valuations may evolve to incorporate a more holistic statistical approach, rather than relying predominantly on goal contributions and league strength.

Additionally, age remains a crucial factor, with younger players generally commanding higher values due to potential resale opportunities and longer peak performance periods. However, this model suggests that a player’s technical output is a stronger determinant than age alone, challenging the notion that younger players are always the best financial investment.



Conclusion

This research highlights the evolving role of data in football transfers, validating sabermetric principles by demonstrating the strong influence of goal contributions, dribbling ability, and league strength on market value. While clubs increasingly use advanced metrics in scouting and recruitment, traditional stats like goals and assists remain dominant factors in market valuation.

Future research could expand by incorporating additional contextual metrics, such as tactical role, pressing efficiency, and team playing style, to provide a more nuanced understanding of player valuation. Additionally, investigating longitudinal trends could determine whether newer metrics like xG/xAG gain greater influence in transfer valuations over time. By continuing to refine player evaluation methods, football clubs can optimise their recruitment strategies and maintain a competitive edge in an increasingly data-driven sport.

Furthermore, this study suggests that clubs may benefit from incorporating alternative financial models that assess a player’s value not just in performance terms but also in potential marketing, sponsorship, and long-term club branding. Given the rapid evolution of data analytics in football, ensuring that recruitment strategies adapt to both on-field performance and commercial potential will be key to sustained success.



This document is formatted for Markdown, making it easy to upload to GitHub. Let me know if you want any additional formatting, edits, or explanations!