The Build-Up Scouting Tool (BUST) pairs a back-end R script that identifies and aggregates passing patterns with a user-friendly UI that allows scouts/coaches/recruitment analysts to easily find a team's most common passing patterns. We identified most frequent sequences of play, in 2-pass chains, and then plotted them using the x,y coordinates from the Opta event data.

The BUST was inspired by passing network research previously done by David Perdomo Meza and tools currently available for use in Opta ProVision, specifically passing chains, heat maps, and density maps. While the passing chains are a great resource there is not a way to quickly up a teams most frequent passing patterns.

The BUST has the potential to add value to multiple parts of a Club or International team's set-up. It can assist scouting departments, specifically identifying build-up tendencies (location or sequence of passes) and adjusting defensive plans accordingly. The Tool could also be used in recruitment to identify players that are similar in involvement to players a club is trying to replace, or wants to add depth behind without affecting the passing identity of the team.

The BUST in its current form can identify and aggregate passing chains, and plot them in the UI for a stakeholder to view for a single team, using their unique Opta_Id, because the BUST relies on the unique Opta Id's it is easily scale-able.

The most obvious potential improvement would be linking video of the individual passages of play to the chains provided on the UI. Another potential improvement, would be tying the chains to their outcomes. Additionally, determining similarity scores for team build-ups would be useful both for scouting and recruitment. The last potential improvement we considered was looking at the all of the chains involving the same unique players instead of only chains that were identical in sequence.

Ultimately we feel that this tool will help scouting and recruitment staff to make more educated sporting decisions.

Share this project: