Apparently this is what they do:
" a 10 x 10 square drawn around the point, then a random point found within this and another 10x10 grid drawn with this random point as the centre. "
And then they record this to 6 decimal places in the database (from the bug when the localities were visible)..
This does not explain why 6 observations are all on one grid, unless:
* the random point is not "random"
* this is a new method and does not apply to the old data brought in from the old site (in which case what is happening there?)
I find it fascinating that the old ZA site had such an elegant solution.
take the locality and display it to 1 decimal point.
map that point as the NW corner of a 10X10 decimal degree square.
I bet you a case of beers people will extract the square and calculate the centroid and visit that spot! And SANBI will be accused of leading poachers and collectors to those spots.
(on the old iSpot the locality was displayed to 1 dd and the square was always the same).
And what happens on projects? Do the species show up in the NEW location if the coordinates overlap? Or do sensitive species never show up?