Quote:
Originally Posted by DBACE
I am inclined to think it is even more complicated in this case. You yourself have a rating that is more than 1 standard deviation, which according to the Bimmerpost formula is in the top 15% of the population. Since Reputation points are measured in a positive direction only (no negative Rep points), the software ignores the possibility of a negative score, yet tells you where you lie in a standard deviation on a bell curve that includes the possibility of a negative score. 
I am sure one of the moderators can better explain the whole thing and in simpler terms. |
Thanks for explaining it. It is a little more complicated only because we have a bimodal distribution of the population (alot of post whores and alot of people with 0-10 reputation) so we have a very large SD.
The other thing is you assume that a bell curve or any population distribution needs negative numbers and so a half bell is used? You don't need negative values to create a population curve. We simply do not display the negative SDs to people since it would not be as informative due to our non standard population distribution.
In the end its almost to the T how you explained it in the first post except we have some tweaks to account for our population distribution and we also take into account relative activity (surfing, etc..)
