Prank and Rankem

[MarkRBright]

Two further thoughts on Prank and Rankem.
I just realised that being all about ranks, Prank and Rankem would probably have to be excluded from Squarepants. In these cases the lower the value the better, so for example, new Boincers would tend to get a huge value up Rankem and Prank. Unless someone can think of a sensible way of inverting the score?
My other thought relates to those potentially huge values of Prank. Perhaps if the maths isn't too problematic, then the highest Project/SubProject rank (or maybe the highest n% of ranks) could be ignored so as to effectively not 'punish' someone for taking up new projects/subprojects.

[MarkRBright]

LOL Well I still think this is a good idea!

[Bok]

I did think about this a little. Ranking across both projects and subprojects is a little more convoluted as they are in distinct tables. But I did construct some sql for just projects or just subprojects, even discarding highest and lowest so it is doable

MariaDB [static]> select avg(projrank0) from dcfree.boinc_user where cpid = 'fa1a611d7c2bb42d13617eb6b8903f8c';                                
+----------------+
| avg(projrank0) |
+----------------+
|      3298.3542 |
+----------------+
1 row in set (0.02 sec)


MariaDB [static]> select (sum(projrank0) - max(projrank0) - min(projrank0)) / (count(projrank0) - 2) as average from dcfree.boinc_user where cpid = 'fa1a611d7c2bb42d13617eb6b8903f8c';
+-----------+
| average   |
+-----------+
| 2479.9255 |
+-----------+
1 row in set (0.00 sec)

[MarkRBright]

I can't help but think it is almost pointless to discard the lowest, as in my case the highest is 83484 and the lowest is 47. I know which one has a bigger impact on my end results ;-)

BUT ... I have no idea if I am a better person for knowing those numbers or not. I almost need to see where I sit on a league table of them in order to see where you stand in the grand scheme of things - which of course I don't because the number says it all.

I kind of think it's probably not worth the effort you would need to go to to display the results. Especially as a smaller number is better compared to a bigger one, which is a little unusual if it were to be shown on the MM table.
Possibly they would sit quite well in the Summary section of the main User by CPID Stats screen, but I am thinking "Meh!" Maybe it wasn't such a good idea.