�ݺ�ߣ

Editor's Notes

• #2: Very happy to be given the opportunity to talk here at EGOS about my research and hopefully get a lot of good and nice feedback on what I am doingFrom pictures, it should be apparent that I am going to talk about football, and more specifically about professional football in Europe. I am going to focus on the careers of professional football players and how organizational status can act as a bridge – both upward and downward - to later career attainment in professional footballDescribe careers in terms of the temporal order and status position of organizations through which these careers develop
• #3: While the interaction between players and teams, and their use of the labor market is the key element in my thesis, my goal in this study is much more modest.I am looking to explore and describe careers of players, in terms of the temporal order of teams, and the status positions which teams take within the marketBased on two observations: one on the market and one in the literature:First observation is how ubiquitous transfers are in European football, in my complete dataset I see more than 450,000 transfers in 9 years time. If I look at a much more restricted set including 184 teams which played in the highest two leagues in the top 7 countries in European football for all years between 2003 and 2011, I see 5,000 players who amassed a total of more than 11,000 transfers. Hence, in 9 years time, on average each players moved 2 times.Such numbers are also backed up by FIFA (more than 11,000 international transfers in 2012)A transfer often involves “the fish pond metaphor”Player has to choose to be either a big fish in a small pond (important player in a smaller team) or a small fish in a big pond (regular player in a bigger team). This is also in some sense what Phillips calls the promotion paradox, in which he showed that Silicon Valley lawyers are promoted to partner earlier in small firms, while remaining longer in lower positions in big firms
• #4: The fish pond dilemma over the career is my lens, which helps me focus on certain career paths.Basically, what I am arguing is that the fish-pond dilemma can be solved in two different ways over the career, and I will try to examine how common these “ideal types” are within professional footballI am exploring or describing how organizations of different status positions are temporally ordered in the careers of professional football playersI argue that such temporal ordering of organizations can take place in two ways:Early selection: players are a small fish, and move to a big pond at a very early age. In this big pond, they are hoping to grow into a bigger fishEarly movements to the big teams, grow slowly into a bigger player. Growing can take place in two ways: sponsored mobility & tournament mobilitySponsored: train and groomTournament: selection through contestsLate selection: players start out in a smaller pond, in which they are a relatively big fish. They hope to move into a larger pond later on in their careerStay in smaller teams at the start of their career to build experience and capabilitiesMove to bigger teams later on
• #5: Now lets talk a bit about my dataThe source for my data is transfermarkt.de, which is a very comprehensive online database, recording information on various aspects of players and teams. As I am interested in professional football in Europe, I limit my analysis to players who at some point in their career played for a team which was active in the highest two divisions of the 7 best countries. These countries include Spain, England, Italy, France, Germany, Portugal, and the Netherlands.This leads to a total of almost 14,000 players -> randomly selected 1,000 of themOf these 1,000 players, I recorded the team for which they played for each month in their career. This leads to 191,240 player-months with team affiliations.I am interested in how players move between bigger and smaller teams, and I use team status to express what is a big and what is a small team:Team status is based on the performance of teams in the past five yearsSimply stated, performance = rank in the league weighted by the strength of that league in EuropeThis leads to a precise status ranking, in which each team has its own spot. For the analysis, I simplified this to 5 status groups, each encompassing 20% of the ranking, from bottom to top. Group 1 is the lowest status group, and 5 the highestI used this information on the careers of players to perform a sequence analysis to get a feel of the overall temporal structure in the careers of playersExamine temporal order of organizations with different status positions in the careers of professional football playersTransformation costs are based on transition rates between status groupsClustering was stopped after I obtained five clusters -> large increase in cluster variance & number which I can handle
• #6: These five groups can be characterized in terms of the early and late selection model:Early selection with tournament mobility is applicable to 45% of players. Moreover, there is a clear difference between the winners and the losers of contestsThe careers of a further 15% of players is better characterized by the late selection modelFinally, there is a relatively big group of 40% of the players which I find difficult to classify. This really shows that what I am doing is still very much work in progressI will look at early selection, late selection, and difficult to classify separately using the sequence index plots
• #7: First lets turn to the early selection with tournament mobilityTo help interpreting the sequence index plots:Plots the sequences ordered on their length and the time they spend in the different status groupsOn the vertical axis are the number of sequences in this groupOn the horizontal axis are the number of months after the 16th birthday of a playerAdded reference lines at 48, 96, 192, and 240 months. This corresponds to age 20, age 24, age 32, and age 36Legend is in the top corner: basically the darker the color the higher the status categoryThe early selection with tournament mobility encompasses 45% of the players in the sample:We see that they move into football at a relatively young age: most have entered before their 20th birthdayMany of these early moves are into the high status teams signaled by the darker colors at the start of the careers of playersHowever, the sequence index plot can be divided into two groups of players:Top 80% of the sequences, these are the losers of tournament mobilityAround 80% of their time in professional football is spend in the lower status categories (3.5 year) – 20% in higher status categories (1 year)Even though they move into a high position early on, but most exit in their early twenties. By 48 months many have exited from the higher status groupsDarker at the start than towards the end of their careerA smaller group of players, around 20% of the players under early selection can be seen as the winners in the contest for the high status positions.They move early in the high status positionsThey maintain their position in the high status teams, and only start exiting after their 24th birthday -> continuous selectionSpend around 90% of their time in professional football in the higher status groups (9 years), while spending 2 years in the lower status groupsOther difference: these players have relatively long careers in professional football (11 years compared to 5 years)
• #8: Besides this large groups of players who experience early selection, there is also a smaller group experiencing late selection encompassing 15% of the players:Start at around the same age as the early selection group (well before 20th birthday), but they start out at lower level teamsThe sequences turn darker over time, signaling that these players move upwards as their career developsUpward mobility visible from early twentiesLate selection players have careers which are comparable in length to the winners of tournament mobility (10/11 years)However, spend more time in the middle of the status ranking, rather than at the top: around half to two third of their career is spend in the middle of the ranking
• #9: Then there is also a third group of 40% of the players which I find difficult to classify in terms of early and late selection:They spend most of their career in non-professional teamsHowever, non-pro category encompasses teams from the lower leagues and teams from abroadDifficult to know the meaning of this category, and the direction of movementsSpend very short spells in professional football -> again meaning is unclearBrazilian player who cannot get accustomed to European footballLower league player who is given an opportunity at a higher ranked team, but doesn’t make itYouth player breaking into the first team, but than exiting professional football rapidlyWhat is clear though is that they enter football at a relatively late stage in their career -> this might be a sign of late selectionObviously, the large size of this group is something with which I am unhappy at this pointShows that it is very much work in progress rather than a finished project
• #10: To sum up, in this presentation I was concerned with exploring and describing careers of professional football playersInterested in the overall structure of the data, rather than on individual transitionsTalked much about the fish-pond dilemma; be a big player in a small team, or a small player in a big teamTwo models of career development have focused the exploration. Main results are as follows:Early selection model with tournament mobility describes the careers of a large group of playersLate selection better describes the careers of a smaller group of playersUnfortunately, there is also a big group which is harder to classifySome questions which come out of this analysis:We see that much happens and is decided when players are in their early twenties. But what is the exact tipping point?Which teams are able to place players on upward or downward moving career trajectories?Can players come up after they went down?Second question is more a normative question: what are the implications of the continuous filtering of playersIn my study, we saw players moving down after they have been filtered outHowever, I miss filtering out at the youth levelQuestion is how well the big teams or national associations are able to identify talents at a young age, and also when someone is not a talentQuestion is what happens before players are filtered out: what are their alternatives after being filtered out?