Use common opponent analysis to give yourself the edge
TennisInsight now allows you to analyse match-ups based on ‘common opponents‘. Academic research in the UK has demonstrated that this kind of analysis may help provide an edge in predicting the outcome of tennis matches. ‘Common opponents’ refers to the set of opponents that both players have faced in the past. The theory is that comparing common opponents is a little bit more like comparing ‘apples with apples’.
On TI match previews, you can analyse all match statistics by common opponents, including the ability to filter down by surface, time and so forth. We all know Rafa has a huge edge over Federer in the H2H department, but who does better against their common set of foes? The answer may surprise you.
How to interpret common opponents on TI
The following points will ensure you fully understand what you are looking at when analysing common opponents on TI:
- By default the set of common opponents is limited to opponents common to each player within the past 12 months
- The set of common opponents remains fixed when you filter for other time periods (i.e. a given common opponent may not have played both players within the filtered time frame);
- The set of common opponents remains fixed when you filter for other surfaces (i.e. a given common opponent may not have played both players on the filtered surface);
- Statistics are rolled up by match, not by common opponent;
- Match stats only include matches where match stats data is available;
- When you jump to common opponent activity, this includes all matches from our database (i.e. with or without match stats);
The above implementation has been chosen as the most practical for TI and it differs in some ways to the academic paper which inspired the development. Over time (and based on user feedback) we may tweak how we apply the common opponent model.
The academic paper is titled A Common-Opponent Stochastic Model for Predicting the Outcome of Professional Tennis Matches and may be found here.