Available are the STATA dataset and all accompanying STATA do-files to generate the tables and graphs in the article entitled "Non-transitive patterns in sports match outcomes: a profitable anomay". This article is forthcoming (or published) in Empirical Economics. Abstract of the article: Sports betting markets share similarities with traditional financial markets. However, they are more accessible for empirical research thanks to availability of high-quality data. As a result, they often serve as a field laboratory for studying financial markets. This study examines 25 seasons of English Premier League matches, revealing consistent non-transitive patterns in match outcomes among various triads (groups of three clubs). These empirical findings are difficult to rationalize and represent a notable anomaly. Bookmakers ignore the non-transitive patterns when setting odds. Faced with a trade-off between efficiency using historical information and maintaining consistency when setting their odds, they prioritize consistency. The dataset has the following information: hometeam = team playing home matchawayteam = team playing away matchfthg = full-time home goalsftag = full-time away goalsb365h, b365d, b365a = betting odds b365 home-draw-awayiwh, iwd,iwa = betting odds iw home-draw-awaywhh, whd, wha = betting odds wh home-draw-awaySeizoen = season