Context

This dataset contains a large number of tennis matches from ATP and ITF leagues. Specifically, matches are available for ATP majors, masters, 500, 250, and challengers, as well as ITF procircuit and juniors. Doubles matches are also available (playerid field will have both names sorted and joined by an underscore (''). For example, a doubles team of Roger Federer and Novak Djokovic would appear as novak-djokovicroger-feder in the playerid column of the all_matches.csv file. Location of players is also recorded, and standardized in a 3 character code. A country join table is also provided for convenience.

Betting data comes in 3 flavors:
Money Line - predict winner of match
Spread - predict whether the difference in games (or sometimes sets) between the players is greater (or less) than a certain number
Totals - predict whether the total number of games (or sometimes sets) between the players is greater (or less) than a certain number

Acknowledgements

Pulled from ITF and ATP sources.