Can the criteria for the top 10 leaderboard be improved?

>> My point was that whatever the length of time for inactivity a player can still abuse the system by playing only people hundreds of points lower rated, getting +0 and staying on the list.

This is only true until some point. By playing opponents way weaker you end up with an always growing rating deviation, until you reach 110, then your rating is marked with a '?' and can't appear in the leaderboard anymore. Then you have to play against stronger people again.

So, all in all, I think the system is already fair enough. Just a matter of time, there is nothing permanent.

>> By playing opponents way weaker you end up with an always growing rating deviation, until you reach 110, then your rating is marked with a '?' and can't appear in the leaderboard anymore

This would take a huge amount of time. I don't know how the rating deviation calculation works here but it seems to increase much more slowly than where I've played online before. For example I played some antichess last year. I only played 85 rated games and the last one was 11th July 2016 but my RD is 62 now. Maybe it will reach 110 one day, although maybe not if I played a 1500 player once a week.

Just a bump to this thread and to note that in 4 months of inactivity since my last post my antichess RD has increased from 62 to 63. This indicates that it would stay below 110 for many years, especially if I occasionally played a very weak player to keep it active.

@tipau Thanks for the detailed report & bump and I believe you are correct. I will submit a RD calculation fix.

Progress is being made!
http://github.com/ornicar/lila/pull/3210

Great, thanks Toadofsky