[Paper Review] A statistical view on team handball results: home advantage, team fitness and prediction of match outcomes

Jens Smiatek, Andreas Heuer|arXiv (Cornell University)|Jul 3, 2012

Sports Analytics and Performance1 references3 citations

TL;DR

This study applies model-free time series analysis to 10 seasons of German Handball Bundesliga data, revealing a nearly negligible home advantage (≈2 goals) and a seasonal decay in team fitness despite stable long-term correlations. The dominance of top teams is explained by rising total goals per match and stronger offensive strategies, enabling a prediction method with 74% accuracy for match winners.

ABSTRACT

We analyze the results of the German Team Handball Bundesliga for ten seasons in a model-free statistical time series approach. We will show that the home advantage is nearly negligible compared to the total sum of goals. Specific interest has been spent on the time evolution of the team fitness expressed in terms of the goal difference. In contrast to soccer, our results indicate a decay of the team fitness values over a season while the long time correlation behavior over years is nearly comparable. We are able to explain the dominance of a few teams by the large value for the total number of goals in a match. A method for the prediction of match winners is presented in good accuracy with the real results. We analyze the properties of promoted teams and indicate drastic level changes between the Bundesliga and the second league. Our findings reflect in good agreement recent discussions on modern successful attack strategies.

Motivation & Objective

• To analyze team fitness, home advantage, and goal distribution in German Handball Bundesliga using a model-free statistical approach.
• To investigate the time evolution of team performance and compare it with findings from soccer leagues.
• To explain the sustained dominance of top teams through increased goal totals and offensive efficiency.
• To develop a prediction method for match outcomes based on statistical trends in goal difference.
• To assess the performance gap between Bundesliga and second-division teams, especially for promoted sides.

Proposed method

• Conducted model-free time series analysis on 3,060 match results from 10 seasons (2001/2002–2010/2011) of the German Handball Bundesliga.
• Used goal difference as a proxy for team fitness, tracking its evolution across match days and over multiple seasons.
• Applied linear regression to quantify the performance gap between promoted teams in the second league and their Bundesliga performance.
• Analyzed the distribution of goals per team per match, finding a binomial distribution rather than Poissonian, indicating higher offensive variability.
• Calculated win probabilities and draw frequencies to assess stochastic contributions to match outcomes.
• Proposed a prediction model based on early-season goal difference trends, achieving 74% accuracy in identifying match winners.

Experimental results

Research questions

• RQ1 How does team fitness, measured by goal difference, evolve over a single handball season compared to soccer?
• RQ2 What is the magnitude and significance of home advantage in team handball relative to total goals scored?
• RQ3 Why do a few teams dominate the Bundesliga, and is this due to improved offensive performance or other factors?
• RQ4 How do promoted teams from the second league perform in the Bundesliga, and what explains their initial struggle?
• RQ5 Can early-season goal difference trends be used to predict match outcomes with high accuracy?

Key findings

• The home advantage in handball is minimal, amounting to only about 2 goals per match, which is negligible compared to the average total of 55.5 goals per match.
• Team fitness, measured by goal difference, shows a short-term decay within 5 match days and a slight overall decrease over the season, contrasting with the nearly constant fitness in soccer.
• The dominance of top teams like THW Kiel and SG Flensburg-Handewitt is primarily due to a significant increase in total goals per match, not superior defensive performance.
• The distribution of goals per team per match follows a binomial distribution, indicating higher offensive variability compared to the Poissonian distribution seen in soccer.
• Promoted teams from the second league exhibit a strong negative goal difference in their first Bundesliga season, with a linear regression intercept of approximately -325 goals, indicating a substantial performance gap.
• A prediction model based on early-season goal difference achieved 74% accuracy in identifying the winning team, demonstrating the method's reliability for match outcome forecasting.