Seperate beserk ratings - idea from Atrophied case

3 &

What was the Atrophied case? Didn't guy get kicked off for sandbagging? What sandbagging has to do with berserking?
The current system is perfect, if someone can't see the logic behind berserking it's their limitation... many tournaments happened to be won or lost just because someone berserked a lot and someone didn't ;) 3 instead 2 pts (50%), and later 5 instead 4 (25%) is huge difference, and of course if You berserk You risk more >>> put rating points on the line and streak in the tournament, but instead You get reward and some chess players (works only against patzers ;) ) get really nervous when they are berserked, which helps berserking guy to win... This Lichess berserking idea is really brilliant as it is, sandbagging is totally different thing and can't blame berserk option for that. If someone intentionally throws points = get's kicked off the Lichess, end of story, fair story ;)
... or yeah let's make ratings for 3+2, 3+2 berserked, 10+0, 10+0 berserked, etc... seriously, I guess it will make useless additional workload for webmasters... :)

@lovlas you said that:
1) You play the bongcloud a number of games where you lose the majority but you are trying to win = fine
2) You intentionally throw games to get a lower rating = not cool
I wonder in what category fits the following hyper-bullet game. What's your opinion?

Wouldn't it just make more sense to reduce the K factor of the berserker so that they lose less rating points (coherently with the fact that his berserk rating is "lower" than expected), but still get punished rating wise?

e3 ? BEST
######

I think this solution is largely unnecessary. Berserking ought to be bad for your rating and somewhat useful for tournament results. Just don't ban people for playing less well when berserking.
Really why ban people for any form of artificial rating alteration except for creating new accounts and losing/winning to them or resigning in the first few moves of a game.

@nebronmace
> Berserking ought to be bad for your rating.

This signifies a departure from what rating was meant to be. Rating was meant to be the best possible measurement of a player's real strength. A measurement that could allow a fairly accurate comparison of two players, even before they seat in front of the chessboard against each other for the first time.

Imagine Morphy playing against Maurian with knight odds at New Orleans, 1869, winning and losing equal number of games. Would you give these players equal ratings, because «Giving knight odds ought to be bad for your rating»?

http://www.chessgames.com/perl/chess.pl?page=19&pid=16002

There is one problem here. Making berserk another rating category will allow to avoid losing rating against some players by berserking them. So it could be used not merely as a tool to reach more tournament points, but a tool for avoiding some rating losses too. For example, if I know that the player's current rating underestimates their playing ability, I will berserk them to decrease chances of higher relative rating loss. If I know that a player's rating overestimates their playing strength, I won't berserk, to skim the cream off their inflated rating number.
Probably, these consequences are not more dangerous to the rating system than the current fluctuations caused by berserking. And it definitely should be considered. I would like berserk to be a separate rating.

@lovlas Everything is unclear about "As long as you are trying to win". Who is there to prove whether you are trying to win or not? How do you define trying?

Some people might even say that you have not tried to win a few games by yourself: lichess.org/PutRtbA8, lichess.org/YH4RDKbS, lichess.org/j3To7Dxd (you can find huge amount of such games for every player)

Of course you do not need to worry about such ambiguities as long as you are the guy who casts the decision, but you might agree that things like "trying to win" are might sound subjective at best. I think that people are ok with almost any rules when they are properly defined and do not have something like "trying to win".

I am sure that people are ok with Atrophied being banned. What they do not ok with (at least this is true in my case) that no explanation was given up till now and people just suppose to believed that lichess has done everything correctly. As far as you can see people do not believe this and lichess takes the approach "we do not care what peasants think, we are the best".

This might work for now and maybe for many more times, but remember that this is the road to a disaster.

@nikolajtesla I did some favor for You boy... I pressed analysis button (perhaps You found it hard to find) and spent some seconds analyzing those games. First game is really beautiful with queen sack and advantage gain, and then he apparently missed that Queen takes with a check, otherwise would be a win. Third game advantage, but missed a tactic, opponent made really nice avcpl and as well beautiful game, but yeah he was "sandbagging" playing 58 moves 3+0 game... Second game seems to be disaster blunder, if You would find more of those games and let's say 200-300 pts loss over certain short period of time thanks to those blunders, then yeah suspicion would have some base. But to make a decision to ban I guess You need to analyze more games, actually to make statistical pattern of the person and to look for deviations, which thoroughly would have to be analyzed with external conditions and rating deviation within time frame.
So please don't make it look like one mod just woke up one day and said "Ah what to do now... let's ban Atrophied, it will be fun :)"
As for "We do not care, what peasants think...", man stop that bulls..t, to be clear, Lichess owe You nothing. You play for free here, learn for free (it is arguable since You don't even bother to analyze IM's games You posted) and yet You claim here some stuff, come on man... They banned the user, user appealed and appeal was declined, so there might be something, which nobody wants to disclose could be for different reasons: maybe to hide detection method, or may be to not harm Atrophied more or some other reason, which could really be nor obvious nor logic to us...
There is some truth in what You wrote it is that - "we are the best" - yes Lichess is the best ;)