
Sign up to save your podcasts
Or


# From the Pitch to the Algorithm — with Ben Teune In this episode, Ian Renshaw and Keith Davids are joined by Ben Teune, Sports Scientist and Analytics Lead at the WTA (Women's Tennis Association), whose PhD work — embedded with the Western Bulldogs AFL club at Victoria University — tackles a problem every constraint-led coach runs into: you're told to "manipulate the task," but rarely given a principled way to decide what to manipulate. The conversation covers how to give representativeness a number using the specificity index and association rule mining, why looking at one constraint at a time is misleading when machine learning can surface which constraint *combinations* actually produce breakthroughs or breakdowns, and how clustering algorithms can classify drills by variability so coaches can build sessions with intentional exploration versus exploitation phases. Ben also walks through his "outnumber" research — the two distinct solutions players find when given an extra player, and how a coach can use that insight to shift team behaviour with a simple time constraint. Underpinning it all is Ian's standout line of the episode: start with the coaching problem, then find the data that answers it — not the other way around. --- If you're getting value from our podcast check out our membership options and coaching community at [www.theconstraintscollective.com](https://www.theconstraintscollective.com) where you can access podcast summaries, narrated presentations, early release podcasts and monthly online meet ups with experts from the Constraints Collective. Support the running of the podcast at [patreon.com/TheConstraintsCollective](https://patreon.com/TheConstraintsCollective).
By The Constraints Collective5
55 ratings
# From the Pitch to the Algorithm — with Ben Teune In this episode, Ian Renshaw and Keith Davids are joined by Ben Teune, Sports Scientist and Analytics Lead at the WTA (Women's Tennis Association), whose PhD work — embedded with the Western Bulldogs AFL club at Victoria University — tackles a problem every constraint-led coach runs into: you're told to "manipulate the task," but rarely given a principled way to decide what to manipulate. The conversation covers how to give representativeness a number using the specificity index and association rule mining, why looking at one constraint at a time is misleading when machine learning can surface which constraint *combinations* actually produce breakthroughs or breakdowns, and how clustering algorithms can classify drills by variability so coaches can build sessions with intentional exploration versus exploitation phases. Ben also walks through his "outnumber" research — the two distinct solutions players find when given an extra player, and how a coach can use that insight to shift team behaviour with a simple time constraint. Underpinning it all is Ian's standout line of the episode: start with the coaching problem, then find the data that answers it — not the other way around. --- If you're getting value from our podcast check out our membership options and coaching community at [www.theconstraintscollective.com](https://www.theconstraintscollective.com) where you can access podcast summaries, narrated presentations, early release podcasts and monthly online meet ups with experts from the Constraints Collective. Support the running of the podcast at [patreon.com/TheConstraintsCollective](https://patreon.com/TheConstraintsCollective).

108 Listeners

371 Listeners

58 Listeners

279 Listeners

159 Listeners

271 Listeners

20 Listeners

190 Listeners

234 Listeners

559 Listeners

22 Listeners

21 Listeners

41 Listeners

192 Listeners