PROJECT
I joined the SBG team as a freelancer, with goal of identifying opportunities to translate the data collated from telemetry devices, cameras and OB units situated at major sporting events.
The application was made up of a video playback panel, pitch tracking component which presented the position activities of individual players, and a host of QUANT data presented in individual panes built - compiled in real time from the raw data captured across the ecosystem.
The goal make the data USABLE, reduce the effort required to understand WHAT was happening. Help users (trainers, players, team managers) make better decisions, predict future performance, and identify training gaps within the team.
- The data set contained vital insights into events, performance, and ISSUES
- But no ideas considered to date helped end users understand the data the system captured
The SBG team wanted recommendations on
- But no ideas considered to date helped end users understand the data the system captured
The SBG team wanted recommendations on
- how that data could be broken into a digestible form.
- moving beyond discrete performance statistics
- the content had to be prepped to enable football trainers make use of the rich context captured.
- patterns within the data had to be "identifiable" visually (and should be presentable in an easy to understand language by the system)
- moving beyond discrete performance statistics
- the content had to be prepped to enable football trainers make use of the rich context captured.
- patterns within the data had to be "identifiable" visually (and should be presentable in an easy to understand language by the system)
FUNCTIONAL REQUIREMENTS
Users need means to :
- filter the types of events showing
- play, stop, jump back 20 seconds, jump to next/previous marker
- set manual “in” and “out” markers to loop through a sequence of footage
- compress / expand timeline, to enable users to zoom in / out
Users should have means to print the timeline so that team / match performance can be reviewed both on and offline.
Events to be shown
Should ultimately be configurable, but would consist of at least:
- Kick Off, Full Time, Half Time
- Incidents (Yellow Cards, Red Cards)
- Events (Goals, Substitutes)
- Performance metrics (shots on target, shots off target).
There may be other metrics which trainers would like to display but these events were outside of the initial requirement.
e.g. a video clip capturing the events which led to a goal).
Counter proposals from separate teams would continue to propose
- the SBG Sports approach (separating data across an array of discrete UI Components)
- ultimately the timeline was built out, and enriched inline with proposals made in the proof of concepts
- Kick Off, Full Time, Half Time
- Incidents (Yellow Cards, Red Cards)
- Events (Goals, Substitutes)
- Performance metrics (shots on target, shots off target).
There may be other metrics which trainers would like to display but these events were outside of the initial requirement.
e.g. a video clip capturing the events which led to a goal).
Counter proposals from separate teams would continue to propose
- the SBG Sports approach (separating data across an array of discrete UI Components)
- ultimately the timeline was built out, and enriched inline with proposals made in the proof of concepts
WORK
With the objectives defined by the team and limited access to users based on the sensitivity of the solution I had opportunity to reflect on statistical packages available on the market with a view to identifying the paradigms presently in existence.
Having a background in digital media, and post production I had a strong sense where the solution existed and set about working up concepts for how the information might be best integrated into the existing components in the screen.
INVOLVEMENT
- Freelance UX
- Enrich basic application
- Addition of more context
- Support Users unfamiliar with stats packages - negotiate context
- Freelance UX
- Enrich basic application
- Addition of more context
- Support Users unfamiliar with stats packages - negotiate context
FIGURE ONE
MATCH REVIEW SCREEN
- Data split across the breadth of screen
- Data can only present in XOR configuration
- Balance skewed towards numeric statistics
- Little space for video context
- Data split across the breadth of screen
- Data can only present in XOR configuration
- Balance skewed towards numeric statistics
- Little space for video context
FIGURE TWO
PROOF OF CONCEPT DELIVERED
PROOF OF CONCEPT DELIVERED
Having combined calls with the team discussing ideas which the team had previously explored, I delivered concepts for an enhanced timeline component.
FIGURE 2, 3, 4 show an overview of functionality, data presentation and the component itself.
FIGURE 2, 3, 4 show an overview of functionality, data presentation and the component itself.
FIGURE THREE
Context integrated in hierarchical configuration - to put user in control of level of detail presented alongside screen - dedicated to presenting sensible view of video footage (rather than "tiny panel"
Context integrated in hierarchical configuration - to put user in control of level of detail presented alongside screen - dedicated to presenting sensible view of video footage (rather than "tiny panel"
FIGURE FOUR
Player centric data entry, stats supporting user to:
- navigate huge array of context captured within matches
- without sacrificing huge swathes of onscreen real estate
- context presented in individual panels / panel combinations (XOR)
Player centric data entry, stats supporting user to:
- navigate huge array of context captured within matches
- without sacrificing huge swathes of onscreen real estate
- context presented in individual panels / panel combinations (XOR)