Predictive game patterns in World Rugby Sevens Series games using Markov chain analysis

Abstract

The purpose of this research was to identify game patterns in rugby sevens, and utilise a Markov chain, to detect how scoring plays are likely to develop. Using notational analysis, a total of 5413 phases were coded from 117 Men’s World Rugby Sevens Series games. Results of a Fisher’s exact test identified significant differences between game patterns in pool games and finals games (p < 0.01) and category 1 games (top 4 ranked vs top 4 ranked teams) and category 2 games (all other games) (p < 0.01). Markov chain analysis revealed that some scoring phases were a result of the previous phase actions. These findings met the Markov assumption that the probability of transition to the next state depends only on the current state. Scoring phases that met the Markov assumption tended to relate to transitional plays resulting from turnovers in possession. The findings suggest that rugby sevens teams should incorporate unstructured practise that involving transitioning from defence to attack, attack to defence and structured defence to unstructured defence into their training.

Keywords:

• Performance analysis
• rugby sevens
• Markov chain
• game patterns