by Arabella D’Havé, Laura Knüppel, Elizabeth North, Pavel Novotný
Background and Theory:
Football has long lived as the favored sport of countless countries. It provides a sense of community, hope, and encapsulates the spirit of human willpower which we know lives inside each of us. The sport has grown thanks to analytics, ever since the first football statistician Charles Reep in 1950, whose analysis is known to have indirectly birthed the long-ball movement in English football (Reep, 1968). Today’s analytics are a far cry from Reep’s hand counting, with sensing technologies at every professional match.
As these sensing technologies have been largely funded by private entities, sports analysis has been left to companies that fund the sensing technology (Pappalardo et al., 2019). We learned during our data search that football analytics is a largely private affair. However, we located two data sets — a robust public data set which logs multiple international competitions, and one which logs different player’s characteristics. These sets are where we will practice data visualization as we ask questions to the data, which can be found in the following sections of this report.
Research Questions:
How do teams differ in position play? (heat maps)
new
standard |xx — — — — -o — — -| novel
A visualization of how teams differ in position play has not been done in this specific way before. Similar things have however been done.
useful
not useful |xx — — — — o — — — | useful
Understanding how teams differ in position play has the power to change the entire dynamic of the world cup. Since the world cup is a competition which is played so infrequently, it is difficult to strategize exactly how each team will perform. If we can identify key traits among coaches or among teams, this could potentially change the entire way the cup is won.
feasible
simple |xxxxx — — — o — xxx| impossible
We are not totally clear on what will be the best way to visualize this yet, however we are keen to understand and try. We think it can be done.
How do players differ from each other? Do some players have more events than others? Can different players’ styles be identified, and if so, do they show a pattern across position or time? (sen key diagram)
new
standard |xx — — — — — o — — | novel
This analysis pairs well with our previous question, though we are not yet clear on how to accomplish it, it is certainly new.
useful
not useful |xx — — — — -o — — -| useful
One of the elements which makes the world cups so interesting is the bringing together of players which are not completely used to playing as a team. This dynamic is difficult to understand, especially for coaches. By identifying key consistencies in players styles, one could learn exactly what sort of instruction to give these players in order to help them quickly create a synergistic playing style.
feasible
simple |xxxxx — — — o — xxx| impossible
We are still figuring out exactly how to visualize the players. It is certainly an interesting question, but we are not entirely sure how feasible it is going to be. We could certainly create some simple pie charts of events per player, which is something you see quite often even during game play. However, we are hoping to discover a more dynamic visualization of the player, perhaps by even being able to characterize patterns among the players which show might be able to name certain styles, or identifying what players might tend to be more flexible than others.
