Return to Project-GC

Welcome to Project-GC Q&A. Ask questions and get answers from other Project-GC users.

If you get a good answer, click the checkbox on the left to select it as the best answer.

Upvote answers or questions that have helped you.

If you don't get clear answers, edit your question to make it clearer.

Categories

Wilson score

0 votes
872 views
in Bug reports by Amberel (120 points)
What is the question ?
by Deepdiggingmole (13.8k points)
I'm sorry about that, I didn't even know I had asked a question - I was searching for existing topics that included that term rather than intending to post my own question! It was a question about why the order of the Wilson scores didn't make any sense, but the question has moved on, I've found that the reason it doesn't make sense is that they are calculating the percentage of FP per Premium User found logs incorrectly. They say the calculation should be the same as on the cache page, but on one of my caches the cache page says 99 FPs at 100%,
whereas PGC has it at only 91%, so of course if that percentage is wrong the Wilson Score will be wrong.
by Amberel (120 points)
I have filed a support question on this, but in case anyone else has an explanation, I have put a composite screen shot that illustrates the problem on http://www.amberel.com/africa.jpg
by Amberel (120 points)
The reason that the percentages are different is because a different formula is used. Project-GC does (number of FP)/(number of premium members + number of basic members that left a FP), Geocaching HQ does (number of FP)/(number of current premium members). Both had the same formula previously but Geocaching HQ has changed theirs, since this version leads to some very odd things (like caches having 100% even tho multiple PMs did not leave a FP and so on) it's being investigated if this was actually an intentional change at Geocaching HQ or not.

For the question about Wilson score I don't see what would be unclear about the order. A cache with many logs _and_ many FPs would have a higher Wilson than a cache with fewer logs and 100% FP, that's the whole point of the Wilson score?
by Pleu (50.2k points)
Thanks, I think I understand that - there are some premium logs that don't have an associated FP, but there are also some FPs from cachers who are no longer premium where the FP is still counted but the no longer premium cacher isn't? As it is a known issue and is being investigated, that's good, thanks :-)
The Wilson score comment you can ignore - at first I believed my understanding of the Wilson score must be wrong because the order didn't make sense, but then I realised the problem was caused by the FP percentage calculation and not the Wilson score calculation.
by Amberel (120 points)
@Amberel Yes, this sounds like you've understood what's going on. :)
by Pleu (50.2k points)

Please log in or register to answer this question.

 | PGC Theme, based on Snow Theme by Q2A Market
Powered by Question2Answer
...