First, we capture every activity within a match.
Our data acquisition specialists (all wrestlers and former wrestlers) view archived matches on video and, using a mobile app, track all of the action in a match. We analyzed over 7,000 matches from the 2019/20 D-I college season, we will continue to accumulate data and our algorithms will continue to learn in real-time.
All match activities are time-stamped and recorded as a sequential stream of data.
The match activities that we track include the various types of offensive attacks and the corresponding result (scored, defended, stalemated or defensive score). We also track neutral time, riding time, bottom time, defensive activities, cautions, penalties scoring and much more.
Then, our cloud analytics platform calculates over 550 wrestling statistics.
We calculate both the offensive and defensive effectiveness of each wrestler using statistics that recap their productivity and effectiveness. Because there’s a time dimension to our data, we can calculate a productivity statistic like “Shots per minute of neutral time” and then test how that statistic correlates to results (scoring).
We have created algorithms to provide accurate wrestler performance insights.
We talked to DI wrestling coaches to get their ideas about what really matters in wrestling. We tested each hypothesis to see if the statistics support the thesis. We (and the coaches) were surprised by the results. We also use certain analytical techniques to search for and discover patterns in the data that may not be part of conventional wisdom. Armed with this analytical framework, our algorithms can predict outcomes with impressive statistical reliability.
Now you can use fresh insights to coach athletes and scout competitors.
The statistics used as inputs to our predictive models represent the most important factors in wrestling with the highest correlations to the desired outcomes. This provides insight into future match-ups and points out areas for athlete development as well as opportunities to scout opponents.
