Akademik Veri Yönetim Sistemi
INTERNATIONAL JOURNAL OF APPLIED EXERCISE PHYSIOLOGY, cilt.8, sa.4, ss.69-77, 2019 (Hakemli Dergi)
Application of network-based algorithms to the sports metrics is addressed to several disciplines. PageRank algorithm has been used for both team sports and individual players performance assessments whereas there is no study to handball team rankings. Purpose of this study is to investigate the differences between the traditional ranking system and PR algorithm in handball. Total 306 game results, scores made, conceding goals, offense and defense-related match statistics that played in Bundesliga 2017-18 season were collected from DKB Bundesliga official web page. We estimated four different PR values by name PageRankstraight, PageRankoffence, PageRankdefense, and PageRankdifference. The official points and normalized PR scores and provide the team standings changed by using the PR algorithm. A positive correlation was found between official system ranking and PageRankstraight, PageRankoffence, PageRankdefense, and PageRankdifference respectively r=0,932, p=0,000; r=0,915, p=0,000; r=0,711, p=0,001; and r=0,926, p=0,000. In this study, the PR algorithm presented a novel approach to ranking handball teams. Further researches needed to investigate alternative ranking methods which allow objective evaluation of a team's performance, not just game results. Finally, we do not conceive the substitute traditional system within the PR algorithm for league standings, but it could be regarded as a method of assessment team market values and distributions of income.