Timeline for How to design a database that tracks sports player's positions
Current License: CC BY-SA 3.0
12 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Aug 1, 2014 at 0:14 | vote | accept | SSumner | ||
| Aug 1, 2014 at 0:14 | comment | added | SSumner | thanks. I think that helps. I was going to have the application side track stoppages, and only insert rows during play. | |
| Jul 31, 2014 at 23:43 | comment | added | blobbles | Otherwise I think the design I have suggested would be best, it will allow comparison between games, allow comparison between different players in the same position in different games, allow for the calculations you have specified so far. | |
| Jul 31, 2014 at 23:41 | comment | added | blobbles | Oh OK, sounds interesting. So you may need to make an addition to the ExactPlayPosition to include "Action" to include some of the actions like Catcht/Throw. Distance travelled will be able to be calculated from relative X/Y positions as already stated. You may need an accompany table as well to indicate game on/game off time which will indicate when play stoppages are occurring, in order to correlate this with the ExactPlayPosition's game time. Likely your tracking system will track players even when the game is stopped. | |
| Jul 31, 2014 at 22:51 | comment | added | SSumner | As far as defining the position, it will depend on the resolution of the LPS, but something along the lines of x/y coordinates of 1-m or 1-ft squares is how I was leaning. | |
| Jul 31, 2014 at 22:49 | comment | added | SSumner | It will use some form of Local Positioning System (so no, not manual). A number of things will come out of it - distance traveled, catches, throws, recreation of the game's positions for analysis. | |
| Jul 31, 2014 at 12:24 | comment | added | blobbles | Oh and where will the data come from to record this? Tracking players by using a camera? Will be a horrendous task, and innaccurate, if doing it manually... | |
| Jul 31, 2014 at 12:21 | history | edited | blobbles | CC BY-SA 3.0 |
added 999 characters in body
|
| Jul 31, 2014 at 12:10 | comment | added | blobbles | Oh OK. How you define the position? Do you split the field up into a number of grids (preferred method I would suggest, 1mx1m squares) or use exact geographic references? Plus tell me what your END goal is, what information do you want to get out of the system? How long each player spends in each position? A density graph of the field where the coloring depth indicates more/less time spent in that position? | |
| Jul 31, 2014 at 11:53 | comment | added | SSumner | Would it be better to insert multiple entries for a single timestamp, and just have one player column? (E.g. 20 entries each time, one for each player)? | |
| Jul 31, 2014 at 11:51 | comment | added | SSumner | I think I phrased the question bad. The "position" is not the player position (quarterback, left fielder, running back, shortstop), it is the physical location on the field, so a table will have hundreds or thousands of timestamps with recorded locations. | |
| Jul 31, 2014 at 9:00 | history | answered | blobbles | CC BY-SA 3.0 |