Analysis Of Professional Soccer Player Transfer Market Based On Complex Network Theory

The development of professional football leagues has attracted the attention of researchers from various field around the world.Since 1969,many scholars have turned their attention to the huge economic benefits of professional football competition.However,due to the influence of some historical and political factors,the data concerning the transfer fee of the players in the professional football league has not been publicized.As a result,current research on professional football leagues is often limited to certain player's own skills,cooperation among various players and the club's investment.In addition,In the field of intense competition among major clubs,each club still decides which players to buy or sell based on the intelligence of football agents and scouts when making player transactions.As well as related player valuations,there is a lack of a method to quantify player value and club needs,research in related fields needs to be carried out urgently.According to the complex network theory,this paper builds a data set based on the transfer data of professional football players crawled from the Transfermarkt website,and conducts indepth analysis and data mining of the transfer rules of professional soccer players.Through the use of statistical analysis,network measurement analysis and other data mining techniques,the hidden rules were dug out from the records of professional football players in the past years,which could be used as the decision support for the clubs who traded players in the transfer market and the players that took career plans.The research of this paper is divided into two parts: the analysis of the cross-border transfer network of football players and the estimation of the transfer fee of football players based on transfer network features.In this thesis,the analysis of the transnational transfer network of football players is divided into two perspectives: space and time.From the perspective of space,in order to explore the impact of geographical space on club transfer,this paper builds a transfer network for a total of 72,603 transfer data of 32 countries and 1,367 clubs from 2011 to 2015.Furthermore,through the Java-based gephi platform,the construction of the transfer network has been visualized.By defining and calculating the network efficiency among the subnets of the transfer network,this study finds that the efficiency between two subnets in the transfer network is inversely proportional to the geographical distance between countries,which means in the global professional football transfer market,the more geographical distances between the two countries,the lower the probability of player transfer.From the perspective of time,in order to explore the impact of transfer policies on the transfer market,based on the clubs of the first-tier leagues of 12 countries,this paper calculates the average shortest path,modularity,network density and other transfer network measures of the transfer network from 1990 to 2015 with gephi-toolkit,and correlates the key inflection p ? oints of these network measures over time with the transfer policies of each season.The experimental results show that the change in the measure presented by the transfer network constructed based on the transfer records of each season can reflect the actual impact of the policy that takes effect this season.In the study of the prediction of football player transfer fee based on transfer network features,in order to overcome the limitations,such as the lack of objectivity and low efficiency,of Transfermarkt's player transfer fee estimation methods based on collective wisdom theory,this article designs the clubs' network measurement features such as degree and centrality measures,constructs sorts of machine learning models such as the Ridge regression model and Gradient Boosting regression model,and trains models based on the transfer data from Transfermarkt website and the players scores data from the sofifa website after preprocessing,it proved that the accuracy of the assessment method based on collective intelligence theory has been improved.Moreover,this paper digs out the features that have a significant impact on player transfer fees and analyzed the underlying reasons.