Scoring Bug? (Could Affect "Cult of the Dragon" Contest)

docsc00ter

I've noticed a potential bug in the rating system for Foundries that might have an impact on the Cult of the Dragon Foundry Contest.


Basically, the "Adjusted Rating" that appears for a quest in the Catalog in-game (i.e., when you hover your mouse over the title of a quest) may not be accurate.


For example, as of this posting, the in-game rating for my contest entry, "The Dragon's Prize" (NW-DLVIDYP25), shows as "3.97".


However, if I open the quest in the Foundry editor and look at the detailed ratings, it's listed as "4.5".


Since the Foundry editor gives us more information about the ratings -- how many people gave it 5-stars, how many gave it 4-stars, etc. -- it's easy to calculate the average rating. Doing so reveals that quest with an average rating of "4.46857" right now. So the Foundry editor is correct in rounding it up to "4.5".


So there's an inconsistency between the actual ratings received and the ratings that are showing up in the Catalog in-game. I'm not sure which figure will be used in determining the contest results, but I'm guessing Authors would prefer the more accurate (and, at least in my case, higher) rating displayed in the Foundry editor.


Am I missing something here? Or is there a bug?


Doc


P.S. If you want the actual math, the quest example I'm citing currently has 43 5-stars, 14 4-stars, 3 3-stars, 2 2-stars, and 2 1-stars. So out of the 91 plays currently, 64 have left a unique review. The total number of stars earned is 286. So 286/64=4.46875 average rating.

eldarth

The tooltip on live server is actually some weird weighted average that seems to always be much lower than actual average until you get it into the tens of thousands of plays.

imaginaerum1

Honestly, since it's the star system, and thus has all of the issues that come with the star system, I've given up hope on any kind of decent resolution to the contest.
I'm still writing my entry, but I'm writing it just to write it, rather than to try to win a contest that will inevitably have an unsatisfying resolution, as people get their guildies to flood the ratings, or otherwise try to manipulate the results.
But yes, the weighted average has always been there, and is yet another way in which the quests that already have tons of plays continue to get tons of plays, while good quests with very high rankings among a smaller number of people flounder far down in the lists.

grimah

thats just how adjusted ratings has always worked.

It starts off at a big penalty (i think 60%), and works closer to 100% (which it will never do because it curves). foundry editor displays it at 100%)

mrgiggles651

What grimah and others said, its an adjusted average. It's been this way since foundry beta. Among the archived threads we calculated how it works to within a few percent, which matched how devs said it worked. There were plenty of threads about whether the rating method was good or not as well.