Football Betting Predictor with LSTM : Project “BetinAbi”

Introduction: Predicting sports match results has been an area of interest for many researchers to date. By using many different data sets and models, the percentage of correct predictions of these models has been tried to be increased, thus helping us to understand the parameters that give good results. After examining the studies that have been done so far, we aimed to bring a new breath to these studies with the machine learning methods developed over the years. We tried to predict the 1X2 results for football matches with the parameters we used for the match and the teams. We conducted a study to cover a total of 700 matches played in the 2022–2023 season in the football first leagues in England and Germany. Variables: In our ‘BetinAbi’ project, we used betting odds to understand the outcome and competition expectation in each match. We used 1X2 bets, total goals 2.5 under 2.5 over bets, and finally Asian Handicap 1 and 2 bets, subtracting the bookmaker’s profit from the odds proportionally. Thus, we added both the expected total number of goals for each match and the expectations about the match result to our model. <a href="https://medium.com/@gorkemturkk/football-betting-predictor-with-lstm-project-betinabi-18a59c6056d4">Read More</a>