Recommend
 
 Thumb up
 Hide
10 Posts

BoardGameGeek» Forums » BoardGameGeek Related » BGG Help and How-To

Subject: BGG ratings algotithm questions rss

Your Tags: Add tags
Popular Tags: [View All]
Dan Moore
United States
San Francisco
California
flag msg tools
mbmb
I understand, I think some of the input factors. Watching the course of A Few Acres of snow showed me that - - as other people have pointed out - - the first 10s are sight unseen by those wish the game to do well.

But I'm wondering how, or if, the Geek accounts for Duration. Every game peaks, sooner or later. Posts trail off, ratings trail off; but the game in question keeps it's high water mark. Or do they get continual revisions, and 'deserve' their standing?

I hope this makes sense as a question . . . Of the handful of subscriptions I have, Twilight Struggle is followed, way back in the field, by Caylus as most posted game. Which denoting continual new interest, ought TS have an even higher posting?

Can anyone tell me if there's another forum I can post this too, for more response? Or a description somewhere of the algorithm?

And does algorithm mean what I think it means?

 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Greg Lott
United States
Little Elm
Texas
flag msg tools
Avatar
mbmbmbmbmb
Not sure what your question is, but algorithm is just another word for formula. And no one knows what the formula is exactly.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Dan Moore
United States
San Francisco
California
flag msg tools
mbmb
good, that's what I wanted it to mean.

My question is: the older a game, the less people rate it. Is it likely that ratings can change upon those later ratings being weighted for continuous interest? ie, if TS gets more attention over time than Caylus, do ratings change according?

If it's all a secret then I suppose I'd have to monitor several games and see if I can figure it out.

Thanks for the 411 and the response.

PS now suspicion algorithm secret to keep people from flaming and frothing about it . . .
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Josh
United States
Santa Rosa
California
flag msg tools
Avatar
mbmbmbmbmb
From what I understand, there are basically two things that modify it from just being a normal average:

1. There are a certain number of 5.5 ratings added to make it so that games with a low number of ratings dont get unduly ranked due to a vocal minority. The number of them added is based on the number of users and so it changes over time.

2. Some "schill" ratings are thrown out as invalid.

There's an old post, I think from 2003 or 2006, from Aldie explaining it, but I dont have that link handy.

Edit: Found it.. New Game Ranking System
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
J
United States
Lexington
Kentucky
flag msg tools
admin
Avatar
mbmbmbmbmb
Moved to HOWTO BGG
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Brook Gentlestream
United States
Long Beach
California
flag msg tools
Avatar
mbmbmbmbmb
I don't believe your supposition that the older a game is, the more likely it is to be ranked holds true. I think over time, people may go back and alter their ratings with a more realistic view once the "newness" of a game has worn off. I also think that over time the rate of new players buying (and rating) the game may also decline. However, these seem like they should accurately be reflected in the final scores just fine and from what I understand these scores are changed in more-or-less real-time as people adjust their individual rankings.

So in answer to your questions, yes, I think if we all decide that we hate Carcossone and rate it a 1, then Carcossone's rank will plummet. But let's not test that.

 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Dan Moore
United States
San Francisco
California
flag msg tools
mbmb
thanks J & J
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Dan Moore
United States
San Francisco
California
flag msg tools
mbmb
lordrahvin wrote:

. . . However, these seem like they should accurately be reflected in the final scores just fine and from what I understand these scores are changed in more-or-less real-time as people adjust their individual rankings.



I guess it's not possible for new ratings to un-affect a game. Only wondered if they were weighted more against older ratings. I wasn't supposing more ratings with age, but less, and if the less, would a game decline faster than a game with more?

But, dash it all, A Thing of Beauty is a Joy Forever. So it would be moot to deny that.

Thanks one and all.
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Maarten D. de Jong
Netherlands
Zaandam
flag msg tools
badge
Avatar
catmando wrote:
Only wondered if they were weighted more against older ratings.

No. A rate given the moment the game appears in the database carries exactly the same weight as the rate given when Aldie pulls the plug. (Of course assuming that the system isn't altered in the meantime.)
 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Greg Lott
United States
Little Elm
Texas
flag msg tools
Avatar
mbmbmbmbmb
catmando wrote:
good, that's what I wanted it to mean.

My question is: the older a game, the less people rate it. Is it likely that ratings can change upon those later ratings being weighted for continuous interest? ie, if TS gets more attention over time than Caylus, do ratings change according?

If it's all a secret then I suppose I'd have to monitor several games and see if I can figure it out.

Thanks for the 411 and the response.

PS now suspicion algorithm secret to keep people from flaming and frothing about it . . .


All of this depends on the algorithm, which no one knows. You will notice, however, that normally the older games BGG ratings are closer to their Geek ratings than the newer ones. Make of that what you will.

Eclipse has definitely turned things on their ears a bit.
1 
 Thumb up
 tip
 Hide
  • [+] Dice rolls
Front Page | Welcome | Contact | Privacy Policy | Terms of Service | Advertise | Support BGG | Feeds RSS
Geekdo, BoardGameGeek, the Geekdo logo, and the BoardGameGeek logo are trademarks of BoardGameGeek, LLC.