Advanced

Re: Checker exception confidence score

Checker exception confidence score
July 17, 2023 10:31PM
Just a query
At what point does a cache that is on the exception list (and has been voted on), does it either get removed or added to the 'This is a challenge' list and therefore appear within the found list on a cachers challenge cache tab
An example
GC3DD4E - Challenge of the Century: 100 words
This is showing with a negative vote and a score of 34% and has done for some considerable time
Though this does appear on lists on the challenge cache tab which suggests it is showing that this is a challenge and yet it falls below the threshold of 35% and with an overall negative vote.

Note: this isnt about whether this particular cache is a challenge or not - this is about the threshold and the process around the exception list and this is an example
Re: Checker exception confidence score
July 21, 2023 06:37AM
The voting page currently says 3 votes are needed. Assuming that means that there are 3 caches below the threshold, it would put the threshold somewhere between 41% and 53% (the current 3rd and 4th lowest confidence levels).
Re: Checker exception confidence score
July 22, 2023 01:25PM
It actually says 3 votes of the same kind (i.e 3 ups or 3 downs) but once there is conflict in that then 3 wouldn't be enough - so I don't think there being a difference of 3 isn't what it means.
I didn't think a specific number of votes specified a threshold or that three was the required gap
I am not sure how many moderators there are though the bottom of that list suggests there are at least 34
Another query with that is that I would suspect that with GC7AJN2 - Mo Farah 10,000 Metres challenge cache which shows 34 up votes that there at least 34 voting moderators - I doubt there is a down vote (I would be surprised there is) but the %age is only 90 - does that suggest there is a downvote (or 3)
Is this linked to how many moderators there are and some haven't voted ??

However my main query was it is evident with the original example that the threshhold hasnt been met - does that mean it isnt clear either way and so no action is taken with regard to declaring it IS or ISNT a challenge cache and so remains as one
Re: Checker exception confidence score
July 22, 2023 07:17PM
If you hoover on the number the tooltip shows how many votes there are of each kind.

The 90%/+34 cache has 34 up-votes and 0 down-votes so it suggest 34 moderatos have voted on that. I would assume the 90% is because it was probably added as a "not a challenge" and I would guess it therefor starts of at a negative score.

The "Challenge of the century" have 9 up-votes and 10 down-votes so the "difference of 3" is not it. I agree that it looks odd that it's showing up in the tab though. Could be a bug?
Re: Checker exception confidence score
July 22, 2023 07:32PM
If you hoover on the number the tooltip shows how many votes there are of each kind.

learn something new each day - didn't know that

The 90%/+34 cache has 34 up-votes and 0 down-votes - irrespective of how it was added the numbers would suggest this should be 100% (as with many of the very obvious ones)

Also - hovering over a few of those it does show how some have had 30+ votes and some far less than that
Re: Checker exception confidence score
July 22, 2023 07:58PM
Since the info text says "Only geocaches with a confidence score of at least 35% will be be considered as an exception. Three homogeneous votes should be enough for that, while for example 3 up-votes and 1 down-vote won't." I don't understand how you come to the conclusion that 34 up-votes should equal 100%? If 3 up-votes and 0 down-votes equals 35% my conclusion is that the % is not based on how many have voted on the cache but it's a calculated score using an algorithm that happens to use the % as it's unit. To me it sounds like it might be impossible to actually get 100%.
Re: Checker exception confidence score
July 22, 2023 08:13PM
I acknowledge that - so it would be interesting to know the math to get that %age figure
A confidence score - to me simply implies that a proportion of voters have voted one way and that is higher than the proportion who have voted the other (which way is irrelevant as we know some that go on the list are challenges and some are not)
The less than 35% suggests the vote is close and a decision can not be made either way

The example (now shows even lower at 32%) started off as a challenge and so presumably will have to remain that way - much the same that a cache that ended up on the list was not a challenge

That said I would have thought the simpler way would have been majority wins (just a thought :-)
Sorry, you do not have permission to post/reply in this forum.