Open awanczowski opened 8 years ago
I agree in principle. I had a similar suggestion long time ago. Anyway. Do we need to specifiy what type of units it is? Like meters or yards or just relative numbers. And is there a need to specify the boundaries? With field we assume a playing field for fotball, basketball or such. But can this be used for positioning a golf ball on a golf course or a sailing boat on the water?
I think a field-location-unit is a good idea. No need for a field-location-x-unit as I believe all units for the field-location will be of the same type. Can't remember if action has a general-unit, but as the field-location units may be of a different unit-type, I am agreeing on your suggestion for specifying what type of units it is. In order to make it less related to fields (soccer-, basketball-, etc) maybe the name should be area-location. I believe x, y and z can appy also for golf courses and sailing courses.
Yes, I agree. A units field would be beneficial. I am not too familiar with sailing or golf. Would those be GeoLocations instead of relative positions within a field?
Today we have start-location and end-location which are strings and according to the documentation can hold a combination of values for each location. So the sample above could be entered something like start-location="63;-27;1". The sample does not distinguish between start and end. And I assume the position is the puck. (icehockey) But is it the location the puck was shot from or where it entered the goal? If we are to use this for alpine, sailing etc it will be the skiier, boat etc that we give location for.
@awanczowski Could you provide some more samples/use cases. We talked about it at the meeting Feb 21th, but it seem a bit limited when it is unclear what you are positioning and start or end position.
Hi @jolla56,
The idea is the ability to have the 3 dimensions of a point expressed in separate attributes instead of having everything in a flat string. The reasoning for this is the ability to do Geometric Shape Queries when the content is indexed in a database. Many databases support spatial indexes now that will be more performant in the long run.
Example 1: As a user, I would like to see a heat map of actions on a field. The broken out geometric points will allow the application to layer these points and show the overall density of actions.
Example 2: As a user, I draw a geometric shape on a field and all actions that occur within those bounds are shown.
I was mostly wondering 3D z-coordinate. What is the use case for that?
The value for the z-coordinate is to future proof for better ball tracking (Football, Soccer, etc.) once that information becomes available.
http://www.smithsonianmag.com/innovation/ball-tracking-tech-american-football-180951841/
OK, I'm fine with this request.
I'm still not clear on the usage when only one point is given. For an action, is it the start location or the end location?
As a SportsML User, I would like to break out location information in grid type objects. New grid types should support 3 dimensional locations as well.
For example, we have a 3 dimensional coordinate "0, 0, 1". The new location object should be able to track x = 0, y = 0, z = 1 in separate attributes.