Technische Universität Wien
> Zum Inhalt
2018-12-04 [

Florian Aigner

 | Press Release 106/2018 ]

The Science of Team Sports

Joint successes in the past increase the chances of winning. This has now been statistically proven in a variety of different team sports.

What makes a team successful? Joint successes in the past are a good predictor of future victories.

Julia Neidhardt

What makes a team successful? This is not only a crucial question for football coaches, it plays a role in almost all areas of life, from corporate management to politics. It goes without saying that a team can only win if the team members have the necessary skills. But there is another important element: joint successes in the past increase the chances of winning. This effect shows up in a similar way in completely different team sports.

A research team from TU Wien (Vienna), Northwestern University (Evanston, USA), and the Indian Institute of Management (Udaipur) were able to statistically prove this phenomenon by analyzing large amounts of data in physical sports (football, baseball, cricket and basketball), and also in e-sports (namely the multiplayer online game "Dota 2"). The results have now been published in the renowned journal "Nature Human Behaviour".

Skills are not everything

The research team collected extensive data on numerous teams from several sports. The strength of individual players was quantified using different parameters - for example in basketball, the number of points scored and the number of assists was taken into account. The strength of the team can then be calculated as the average strength of the players.

"This gives us a value that can predict the outcome of a game reasonably well," says Julia Neidhardt of the E-Commerce research unit (Institute for Information Systems Engineering, TU Wien, Vienna). She conducts research in the areas of team performance, user modeling and recommender systems. She does not only consider individuals, but also models their relationships, for example with the help of social network analysis. "Teams with better individual players have of course a higher chance of winning - but that's not the end of the story," says Neidhardt.

The team effect
In all the sports studied, the actual results of the games can be predicted even better by not only considering the average strength of the team members, but also taking into account how often they have been victorious together in the past. It is therefore not only important to bring the best possible stars to the field, they also have to gain experience together as a team by celebrating joint victories.

Especially in elite sports, where the skills of all involved professionals are extremely high, individual differences do not necessarily play the key role. As the differences in the skill levels decrease, common experience becomes more important.
It is particularly interesting that the effect was to be seen in very different sports: In football or in the e-sport "Dota 2", the team members permanently depend on each other. Most actions are performed by several players at the same time. In baseball, on the other hand, throwing and hitting the ball are individual actions that have nothing to do with the rest of the team. Nevertheless, the team effect can be seen in all these sports.

Robust result
There are different possible explanations for this: By training and playing together for a long time, the players become better at coordinating their actions and predicting their teammates’ reactions, but there may also be strong psychological effects, when there is a strong emotional bond between the team players. The statistical data cannot conclusively answer the question which effect is more important.

"We can see clearly that in the case of similar skill levels, prior shared success is a good predictor of which team is going to win", says Julia Neidhardt. "This effect is very robust, in a variety of sports. This leads us to suspect that similar effects also occur in other areas."

Original publication: Mukherjee et al., Prior shared success predicts victory in team competitions, Nature Human Behaviour (2018). DOI: 10.1038/s41562-018-0460-y

Contact:
Dr. Julia Neidhardt
Institute for Information Systems Engineering
TU Wien
T: +43-1-58801-188300
julia.neidhardt@tuwien.ac.at
julia.neidhardt@tuwien.ac.at