In this bachelor work I choose to analyze the creation of sports odds in bookmaking, from betting shop's point of view and look for any suspicious values, values which are significantly different from values estimated by a model, and then I will try to explain these values. Firstly is made regression model via method of ordinary least squares, which is later tested and corrected to a version, fulfilling Gauss-Markov Theorem. After that the analysis of residuals is performed and suspicious values... show full abstractIn this bachelor work I choose to analyze the creation of sports odds in bookmaking, from betting shop's point of view and look for any suspicious values, values which are significantly different from values estimated by a model, and then I will try to explain these values. Firstly is made regression model via method of ordinary least squares, which is later tested and corrected to a version, fulfilling Gauss-Markov Theorem. After that the analysis of residuals is performed and suspicious values are found out. All of this is firstly made for 10 rounds of Premier League and then for 10 rounds of Czech league. After all, these 2 leagues are compared to each other. All functions are made in program gretl. As data were used odds from Czech betting shops Tipsport and Fortuna. The way, how the bookmakers make their odds was explained and we also explained, which values in our models are suspicious and also from which reason. In the comparison of leagues, Premier League was better, due to lesser manipulation with odds. |