Performance Assessment in Water Polo Using Compositional Data Analysis

Area of Science:

• Sports Science
• Performance Analysis
• Water Polo Analytics

Background:

• Understanding offensive performance is crucial for success in water polo.
• Previous studies have analyzed individual performance metrics, but group-level analysis is less explored.

Purpose of the Study:

• To identify groups of offensive performance indicators that best discriminate between favorable, balanced, and unfavorable match scores in water polo.
• To analyze the impact of specific offensive actions on game outcomes.

Main Methods:

• Analysis of 88 Spanish Professional Water Polo League games (2011-2014).
• Clustering offensive indicators into five groups: attacks, shots, attack outcomes, shot origin, and technical execution.
• Application of additive log-ratio transformation for compositional data.
• Multivariate discriminant analyses to compare match scores.

Main Results:

• The "Attacks outcome" group showed the highest discrimination between match scores (60.4% classification accuracy).
• In games with penalties, goals (SC = .761), counterattack shots (SC = .541), and counterattacks (SC = .481) were primary discriminators.
• In games without penalties, goals remained the most significant discriminator (SC = .576).

Conclusions:

• Offensive performance groups, especially "Attacks outcome," are effective in discriminating water polo match scores.
• Specific offensive actions like goals and counterattacks play a critical role in determining game outcomes.
• This study offers a novel approach for evaluating the importance of offensive performance groups in water polo.