Loading ...

A different way of conducting video analysis | Embracing Technology | ConnectedCoaches

ConnectedCoaches uses cookies. By continuing to browse the site you are agreeing to the use of the cookies. For more details about cookies how we manage them and how you can delete them see the 'Use of cookies' part of our privacy policy. Click the cross to close this cookie notice. X

Home » Groups » Embracing Technology » blogs » Rob Maaye » A different way of conducting video analysis
Embracing Technology

Leave group:

Are you sure you want to leave this space?

Join this group:

Join this space?

Add a new tab

Add a hyperlink to the space navigation. You can link to internal or external web pages. Enter the Tab name and Tab URL. Upload or choose an icon. Then click Save.

The name that will appear in the space navigation.
The url can point to an internal or external web page.
Login to follow, share, and participate in this group.
Not a member?Join now

A different way of conducting video analysis

Avg: 0 / 5 (0votes)

Blog by sports coach UK’s Research Manager John McIlroy

When I first heard that Disney Research existed I imagined they would be testing some of the most essential questions like whether elephants can actually fly. However my facetious attitude changed when I looked into what they actually do and discovered an interesting sport video analysis project they were working on.

In sports were players constantly change role quickly throughout the game they found that focussing on, and analysing, the position rather than the individual was best for detecting tactics, strategy and style of play for a team. As they state in their research paper

As players constantly change roles during a match, we show that employing a “role-based” representation instead of one based on player “identity” can best exploit the playing structure.

At the heart of this approach is an attempt to make sense of data. You can collect lots of information on an individual but does it help understand team behaviour or have you collected so much information that you are stuck with that eternal problem of too much noise associated with having too much data.  

This method reduces complex movements of individuals throughout a game to a more manageable ‘compact representation’ thus making it easier to understand what is going on.

The research uses all sorts of things I don’t understand like spatiotemporal modeling but what is perhaps most interesting is the idea behind it, that sometimes video analysis by position is better than by player.

If you want to read the research paper you can click HERE.

Comments (1)

Avg: 0 / 5 (0votes)