Why am I not in the rankings in the bullet?

pemo, at least Nakamura was playing the cream - he was playing players who were 2500, 2600 rated and beating them.

The player in question here is playing 1900 rated opponents when he's been challenged multiple times by NMs and IMs who play here. I know Kingcrusher has challenged him before.

He is certainly a coward to not play people near his rating when he has the option to do so, and that is certainly why he cannot be rated: the difference in ability is so vast that it cannot be accurately measured. It is like if Magnus Carlsen chose to only play me to achieve dominance. The gap of uncertainty is so large, that it cannot be considered accurate.

Of course Nakamura is an exceptional example, but illustrates the problem very well.

I admit, that it´s not easy (maybe impossible) to measure such huge differences accurately - but in my opinion it cannot be the solution to punish, blame or ban players having chosen that "way".
They are doing nothing wrong, the are simply playing chess according to the rules - nothing more, nothing less.

#23 "Punish?" Are you kidding? Being listed in the leaderboards is a privilege for those whose ratings have a high degree of confidence (low rating deviation) and are great in magnitude. Being not listed on the leaderboard because your RD is enormous (and therefore mathematically speaking by Glicko-2 definition, it's probable your rating is inaccurate with regard to your actual strength) is not a punishment.

I can see where pemo is coming from, but the people who designed Elo and Glicko(-2) are very, very smart and have already considered these questions.

They were aware that these practices can make ratings inaccurate, and put in these measures to ensure we know what constitutes an accurate rating. This isn't a lichess thing, it's a rating system thing. And I have full faith in the ideas behind the sophisticated Glicko-2 system, which has been extensively peer reviewed by seasoned mathematicians.

Once again, hellball is right.

pemo, he doesn't demonstrate the example. Nakamura was obliterating all competition, winning with a 99 - 1 record against 2500s like it was easy as pie. In our current example, we've got someone who has a 90 - 10 record against 1900s, and a 50 - 50 record against anyone around 2400. He's a strong player for sure, but he doesn't play anyone 2100 or up.

I'd actually suggest that having such a huge rating deviation is a form of artificially boosting your rating. Perhaps lichess should block rated games between people unless they are within +-500 of each other, or something. Keep casual as the full range, but block a certain amount to prevent boosting by selective play.

I know that this discussion is not really a lichess thing, but a rating system thing. But maybe how lichess leaderboards are designed...

Being kicked off the leaderboard after having played hundreds or thousands of games (and having done nothing wrong) because of a rating system, that the guy might have absolutely no knowledge about, is not a punishment? Hmmm...
OK, he can learn how the rating system works and adjust his behaviour (playing stronger opponents to get a lower RD) to fit the ratings systems expectations to get a stable or accurate rating (from the rating systems mathematical perspective). But results this change finally in a weaker or in a stronger player? No! His skills stay the same. Not playing opponents beyond a special limit/rating/strength does not violate any rule. I mentioned already that I do not prefer this way, too. But it´s a fact.

I do not dare to start theoretical discussions about rating systems because of my limited knowledge, but having a huge rating deviation because of playing much weaker opponents is not a way to artificially boost the rating - it´s a result of the rating system.
I don´t put into question, that glicko-2 is a very well designed rating system, maybe the best to measure the strength of a (chess)player base. And I have no idea how to do that better.
For sure the designers of glicko-2 are aware of this problem and put a lot of energy in that. Maybe there is no (final) solution to solve this to the last extent?!?
I just want to point out, that there might be the one or other exceptional case, where the calculation of ratings can lead to... let´s say a feeling of injustice. And that I have a basic understanding or "sympathy" for the initial motivation to open this thread. ;-)

#28 The flaw is that no rating system is good at ranking players. There isn't a single game or sport where an accurate rating system correctly ranks top players.

That's a reason some sites use ladders, pools, etc. for ranking top players.

#29 And in this case a top player (in my opinion) is excluded from the leaderboard because the rating system is not good at ranking top players... that sounds at least a little bit strange. ;-)