Question

Sometimes, Project-GC miscalculates the Favorite percentage, e.g., at https://coord.info/GC2M93Y. Groundspeak shows 98%, which is correct. Project-GC shows 87%.

Answer 1

The most common reason for this is what pinkunicorn said. The second most common reason for this is that people that have left a FP have stopped paying for premium membership and geocaching.com uses an incorrect formula for the calculation. The third is that people left a FP and then had their log deleted and geocaching.com does not handle that exception but Proejct-GC does. For the cache you've linked the issue is the second one.

The issue is not at Project-GC, it's at geocaching.com. They use the formula (total FPs)/(current premium members founds), Project-GC use (Total FPs that are connected to a log)/(current premium members founds + basic members that left a FP). Since geocaching.com includes the FP from the people that are not longer premium members but excludes the account of them, their numbers will be higher for many caches until they fix this bug.

For the cache you've linked these accounts left an FP but is now no longer premium members:

Burstelhexe: https://coord.info/GL9NZJ29
CCCNH: https://coord.info/GLDF0M6N
dieroes: https://coord.info/GLA14W05
Fullhousemaddim: https://coord.info/GLG5W07E
jamesone1: https://coord.info/GL9WEWAR
judy.roughneck: https://coord.info/GLEAGVQ2
misahast: https://coord.info/GL9M16DJ
Schokokekse: https://coord.info/GLB0XTYP
sunnythedog1: https://coord.info/GLA5Q4X9
Team Uncharted: https://coord.info/GL9V98Q6
viniausparis: https://coord.info/GLA5M5Q4
XdreamTEAM: https://coord.info/GLB19C44

The cache has 134 found-logs, out of which 28 are from basic members. This means Geocaching.com is calculating 103/106 which is ~97,2% but I'm pretty sure they just round everything up and get 98%, Project-GC is calculating 103/118 which is ~87,3% and with correct rounding this shows as 87%. The correct FP% of this cache is ~87% as Project-GC shows it, not 98%. If these 12 people paid for premium membership again then the percentages would match or at least only differ by rounding.

Comments

Thanks! This clears things up.

---

To be honest, my interest in FPs is that I want to find caches that are good. Thus, if someone at any point felt they wanted to award a cache an FP then I think that's relevant, regardless of whether their log has been deleted or they have stopped paying for premium.

---

@pinkunicorn I agree with the reason to look at FP but I don't find incorrect data helpful so I still think geocaching.com should correct this. :) The issue here is that you can have a cache that have 100%FP on geocaching.com while in reality only half the people that logged it gave it a FP and therefor the 50% at Project-GC would be a more correct picture of how many people felt the cache was worthy of a FP.

It could be argued that PGC should do (Total FPs)//(current premium members founds + basic members that left a FP + deleted logs that left an FP) I guess, but the effect of deleted logs shouldn't be huge. My main point is that doing (total FPs)/(currently active PMs) is bad because it gives incorrect and "inflated" numbers.

Answer 2

Not really. The correct phrasing is more likely "sometimes Project-GC has not had time to detect favorite points added after the posting of the initial log".