You are right.
I have added the paper about glicko original rating.

I think only volatility was added in glicko 2.

Cheers, filer.

Hm, tough stuff, Glicko certainly doesn't have much in terms of wiggle room.

After reading through the paper, I think the best option here would be to declare each poker round a Glicko "period". Then we could view the weighting as players having played N times against each other - so in my approximate "linear" mode, the heads-up would be equivalent to 9 games of 1st vs 2nd, whereas finishing last would be equivalent to only 1 game against every player at the table.

Thank you for your input.
I will test your suggestion and try to clean up the code some more asap.
The beauty of glicko(2) is that inactivity is handled automatically.
When i am satisfied with the results, i will publish the java code on github.

Cheers. filer.

Well, the inactivity bit is the whole point of rating systems with deviations. Not that Poker-Heroes really needs it - the seasons thing is probably the better approach and fits Poker quite well.

But I am looking forward to playing with your code

It is nice to nice the fellow mathematician (or computer scientist) on the forum.
I am implementing this for our WEC games.

You are welcome to join, quality of the game is much higher than in regular ranking games, despite my silliness.
See:
www.pokerth.net/component/kunena/13-gene...041-weekend-cup.html

Cheers, filer.

Hi, been playing and reading this forum for a while, but this is my first post.
I found this on sourceforge if it helps the developers at all to implement an Elo based ranking system, which I think is a must: sourceforge.net/projects/multirank/?source=directory