English premier league(EPL) is one of the most popular professional soccer leagues around the world. It caught out interest that high-market-value players can achieve astonishing commercial value by increasing TV ratings of soccer game they play. So, we wanted to use the dataset of EPL players to group an optimal soccer team with highest total market value. There are two steps of our analysis. The first one is to predict players’ market values using multiple regression in python. In this process, the noteworthy things are that we split the data into train data and test data, used backward selection to select features that are most correlated to market value and successfully built a regression model to calculate market value accurately. Then comes to optimization and We used Excel Solver as our optimizing tool. Variables are to decide whether to select players or not, constraints comply to the rules of English Fantasy Premier League and the objective is to maximum the total market value of the soccer team. There is one thing notable is that due to the limit of 200 variable in Excel Solver, to achieve the optimal solution, we arranged the players’ market values in descending order and took the first 200 players as our variables. The result of our optimization is that, with the budget of 100 million-euro dollars, the team that we group can achieve market value up to 443.63 and consists of 4 attacker players, 8 defender players and 3 goal keepers.