ELO score calculation inconsistency

Discussions about BGA (all languages)
Forum rules
Warning: challenging a moderation in Forum = 10 days ban
More info & details about how to challenge a moderation: viewtopic.php?p=119756
Locked
User avatar
Nglob
Posts: 2
Joined: 16 April 2018, 18:40

ELO score calculation inconsistency

Post by Nglob »

I won a Race for the Galaxy game recently (table 169949466), and was surprised how little my ELO rose. It looks like I got the average of the scores against the individual players instead of the sum of the scores.
OK, I figured they changed how ELO was calculated. But then I won a Splendor game (table 177072765) and the total ELO was the sum, as I had expected previously.
Does anyone know if this is a bug, or if the ELO calculation differs depending on the game, or what is going on?
User avatar
ufm
Posts: 2039
Joined: 06 January 2017, 08:38

Re: ELO score calculation inconsistency

Post by ufm »

Actually it needs some extra calculations.
Check this out and please upvote if current system is too confusing: https://boardgamearena.com/bug?id=29568
User avatar
postmans
Posts: 57
Joined: 25 September 2016, 17:27

Re: ELO score calculation inconsistency

Post by postmans »

I also observed differences between Railroad Ink (where it seems average gain against all players) vs Welcome To (where it seems to be the sum of all ELO gains/losses).

Your example of race for the Galaxy vs splendor shows the same difference.

What ufm says about k-factor is also there, but is not accounting for all of the difference. in splendor you had a higher k-factor because it was still one of the first (20?) times you played that game.
Locked

Return to “Discussions”