Multiple Regression Models Were Used to Predict Match Outcome Through Football Team Match Data

With the increasing prosperity of football, the result prediction of football matches has become a hot spot in the commercial operation of sports, and also an important issue studied by the academic circle. Research about the results of football prediction, most research scholars from the factors of the results, such as the team strength, the weather, the team ranking, team status, coach, team home and away combat ability, but a large number of historical game data collection is more difficult, and part of the political factors cannot be quantified. A former study found that gambling companies mainly analyze the data of football games, while the team mainly focuses on the presence of the players, so as to analyze the situation before and after the game. The analysis of the influencing factors of competition results is mostly the way to calculate the complete influencing factors by traditional methods to realize the purpose of competition analysis and prediction. This paper takes the game data in the football website of the scout network as the data source, and captures the historical data of all the recent two seasons of the English Premier League through the web crawler technology. The collected data were cleaned in detail, and the football history data were analyzed by multiple factors. Multiple regression models were used to predict the outcome of football matches, and the influencing factors that were valuable for the analysis and prediction of football matches were found.