Can the criteria for the top 10 leaderboard be improved?

I've seen a few situations where a player gets their rating high enough to get top 10 and then pretty much stops playing. But instead of dropping off the list through inactivity they deliberately play 1 game per week against someone significantly lower rated and stay there forever.

I find it a bit annoying, especially as most of the times I've seen it the players have been significantly over-rated and take away a top 10 place from a more deserving player.

Could there be a different rule in place to avoid this problem? Maybe a minimum rating change (gross change, not net) over a week? Even a change of 1 elo point would be an improvement, as it would stop people playing +0 rated wins to stay on the leaderboard.

To encourage activity, perhaps the leaderboard could only include players who played a pool game in the last 24 hours (or N pool games in the last D days).

Originally I thought that the leaderboard could be supplanted by a ranking system more like 81dojo or StarCraft; however you do state an issue which incentivizes sad behavior.

It helps to think about this problem taking into account different perspectives.

What about people who get a high rating but were dropped from the leaderboard because they have not played in a week? I might argue that I should be in horde top 10. By similar argument one can say that it is annoying to see significantly lower rated players on the leaderboard just because they play a few games. As far as I am aware Carlsen was not dropped from the FIDE leaderboard in January, nonetheless he has not played a game in that month.

What about a player, whose skill is so high that he does not even has a reasonable opponent to play? en.lichess.org/@/nonpareil is not even on the leaderboard in horde, nonetheless he is 300 point higher than the best guy on the leaderboard. More than that, because he plays only against "lower rated players", he is marked as "artificially increases/decreases their rating". In case you wonder who are these low-rated players, I would say that they are really strong (in top 10 in horde), but just significantly lower than him.

@nikolajtesla I've been dropped from the atomic leaderboard due to not playing in a week, which I think is completely correct. If players don't risk their rating/position regularly then they shouldn't be able to maintain a place on the top list. You haven't played a horde game in a long time and therefore shouldn't be on the leaderboard. If you want to get back on there then you know what you have to do. To me it's clear that the timescales involved for inactivity should be much shorter when dealing with online chess, compared to FIDE rated events.

I don't know the details regarding nonpareil, however I'm suspicious of variant players who play the majority of their games as correspondence. It's much easier to abuse the rating system that way (as well as computer cheat), which I've seen done before to inflate ratings.

@tipau you missed my point. I was not arguing that I should be on the leaderboard (I do not care). My point is that you should take into account others people preferences. And some people might think that this is unfair that they are removed from the leaderboard only because they have not played a few games.

>> it's clear that the timescales involved for inactivity should be much shorter when dealing with online chess, compared to FIDE rated events.

ok, taking your point in the consideration you need to decide how much shorter your timescale would be. Is 3 days enough, or 1 day, or 6 hours, may be let's do it 1 hour? P.S. if one will not play FIDE games in one year, he would not be thrown away.

Regarding nonpareil, look at his games. He plays 1 min, 3 min games and win almost all of them. Also he check his games against hardest stockfish (in last 10 or so he has not lost a single game). With what can you cheat if you consistently win the hardest level of engine (yes you can cheat with a better engine or with longer wait time, but based on his performance I doubt this is the case).

One more time, I also think that the leaderboard is broken, so I am not telling that let's not change it. I just tell that you should think about other use cases.

@nikolajtesla Regarding the case of Nonpareil, you should check this if you want to understand why he has been marked :
fr.lichess.org/forum/team-fair-horde-players/the-player-who-is-bothering-us
Also about his wins against the engine, you should be objective and see the insane amount of takebacks he allows himself in order to win in the end. Sure it must take him a lot of time, but this is for me nothing to be impressed of. Plus, I personally have a positive overall score against him.
My last point is that nowadays in my opinion the Horde field has become much stronger than before, so I'm not sure he would beat every one of the top players.

Now to the actual topic of this thread, I see indeed penguingim1 doing this for his Horde rating, playing only one game every one and then just to stay in the leaderboard. Even if he knows he's not strong at Horde since he only plays very fast time controls (and without any increment) as Stubenfisch beat him easily in one of his simuls and penguingim1 refused a rematch. Now, I don't know if there could be a better system to avoid this kind of behavior, but as long as I know where I stand it doesn't really bother me. Just a trophy color :D

@svenos wow, I had no idea about it. The time when I played horde, he looks to me like a genuinely strong player.

I also remember you :-)

@nikolajtesla

>>some people might think that this is unfair that they are removed from the leaderboard only because they have not played a few games

I don't think that and if I understand right neither do you. I haven't heard someone complain about it before so not sure it's worth spending time thinking about such a hypothetical problem. The only people I can think of who might argue this are people who have high ratings and are afraid to risk them. The number of people who would rather inactive players are removed from the list would dwarf those who don't.

>> you need to decide how much shorter your timescale would be

It feels like the thread is being dragged off topic, but personally I'm happy with the 1 week that is currently in place.

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. Something should be in place to avoid that in my opinion, as per my suggestion in #1.

> If players don't risk their rating/position regularly then they shouldn't be able to maintain a place on the top list.

+1 to this "ladder" concept.

+1 to #1, for the mentioned reasons.