@Arcticpenguin This still allows for some typical Arena abuse. For example, if a strong player is waiting for a pairing, you can pause and wait, to get a more easy player to fill in your game cap. Not that you cannot do this now (but you cannot do that in Swiss), but such a change will exacerbate the issue, as pausing doesn't have much negative side effects anymore.
@Wolfram_EP #81 That is certainly true, the only way to get the most fair pairings is to wait for each round to finish. I also think there is sufficient demand for a pure Swiss system despite the drawback in pairing time. In particular, I think spectating Swiss tournaments with titled players would be way more fun than Titled Arena because it would guarantee exciting, decisive matchups between the leaders towards the final rounds.
However, given Lichess's past experience with the format, there seems to be little developer interest in introducing anything with fixed length rounds. While an accelerated pairings "Swiss-lite" system could be abused the way you suggest, the game cap also allows a longer delay between pairings to wait for a suitably performing player on roughly the same win percent. Of course, this also takes a greater change to the pairing code and some testing, but I still think the option to limit the number of games that can count towards a tournament would be a simple way to improve the fairness of the Arena format for those wanting it.
However, given Lichess's past experience with the format, there seems to be little developer interest in introducing anything with fixed length rounds. While an accelerated pairings "Swiss-lite" system could be abused the way you suggest, the game cap also allows a longer delay between pairings to wait for a suitably performing player on roughly the same win percent. Of course, this also takes a greater change to the pairing code and some testing, but I still think the option to limit the number of games that can count towards a tournament would be a simple way to improve the fairness of the Arena format for those wanting it.
I have proposed a feature which would satisfy quite a lot of the needs expressed in this thread, and I'm guessing the implementation effort is limited.
I have so far seen limited interest in my light-weight suggestion of organizing tournaments outside lichess, with the support of a lichess framework without automatic pairing. A director (much like in OTB chess) can organize the tournament using lichess and this framework, and some limited scoring assistance from the lichess tournament.
lichess.org/forum/lichess-feedback/feature-suggestion-tournament-type-where-lichess-is-not-doing-pairing
Please let me know if there's any additional input I can provide. I'm thinking this should be possible to have up and running in very little time, while we keep discussing how to make a fully fledged automatic Swiss tournament!?
I have so far seen limited interest in my light-weight suggestion of organizing tournaments outside lichess, with the support of a lichess framework without automatic pairing. A director (much like in OTB chess) can organize the tournament using lichess and this framework, and some limited scoring assistance from the lichess tournament.
lichess.org/forum/lichess-feedback/feature-suggestion-tournament-type-where-lichess-is-not-doing-pairing
Please let me know if there's any additional input I can provide. I'm thinking this should be possible to have up and running in very little time, while we keep discussing how to make a fully fledged automatic Swiss tournament!?
I think there are basically two desirable conditions:
a) We want the tournament to be convenient for the players, so that they can pause and come back any time, if possible.
b) We want the tournament to be fair, so that stronger chess players consistently perform better.
These two conditions, although both desirable, are contradictory in nature. The more "convenient" in the a) sense you make the tournament, the more "non-chess skill" you need to play it well (which does not work well with b), and the more freedom there is in choosing pairings. One can try to mitigate the issue, but it is still here, independently of how you intend to "fix" Arenas. I do not see much sense in trying to create some middle variant, because it will turn out to be not well-suited for both purposes. Arena is already good enough to fully satisfy a), and Swiss is already good enough to fully satisfy b) - there is nothing to invent here.
On the developer interest. I can say my honest opinion, though it may be unpleasant to read, but just based on many similar threads... I see it as:
"We are the progressives! We created the best format ever possible! It is clearly superior to anything ever existed before for all online chess tournaments and we will dismiss all the evidence of contrary."
The "past experience" with the Swiss format is just an excuse for dismissing this evidence, not some real issue. As well as the difficulty of coding is not a real issue, as there are many people who would like and are able to implement Swiss, but the lichess team made it pretty clear that it will not be merged and deployed, because they don't bother about this feature and thus don't want to maintain it. But come on, are team battle arenas more easy to maintain or more needed feature without which lichess feels somehow incomplete? I don't understand that.
a) We want the tournament to be convenient for the players, so that they can pause and come back any time, if possible.
b) We want the tournament to be fair, so that stronger chess players consistently perform better.
These two conditions, although both desirable, are contradictory in nature. The more "convenient" in the a) sense you make the tournament, the more "non-chess skill" you need to play it well (which does not work well with b), and the more freedom there is in choosing pairings. One can try to mitigate the issue, but it is still here, independently of how you intend to "fix" Arenas. I do not see much sense in trying to create some middle variant, because it will turn out to be not well-suited for both purposes. Arena is already good enough to fully satisfy a), and Swiss is already good enough to fully satisfy b) - there is nothing to invent here.
On the developer interest. I can say my honest opinion, though it may be unpleasant to read, but just based on many similar threads... I see it as:
"We are the progressives! We created the best format ever possible! It is clearly superior to anything ever existed before for all online chess tournaments and we will dismiss all the evidence of contrary."
The "past experience" with the Swiss format is just an excuse for dismissing this evidence, not some real issue. As well as the difficulty of coding is not a real issue, as there are many people who would like and are able to implement Swiss, but the lichess team made it pretty clear that it will not be merged and deployed, because they don't bother about this feature and thus don't want to maintain it. But come on, are team battle arenas more easy to maintain or more needed feature without which lichess feels somehow incomplete? I don't understand that.
I remember times when we were announcing record number of online players. In october 2015 record number of players online was 7000. And the same year earlier the peak was 3000. I'm here from december 2015 and I'm not aware of existence of RR or swiss tournaments. So they were there probably before that, when the number of players was much smaller. So arguing with "we tried that and it didn't work" is just an excuse in my eyes. Time's change, players pool's change, give it a shot. If it works on chess.commerce, i don't know why it could not work here.
And if it was once implemented, the source code is probably reusable (with same changes, beucase the coding of lichess probably changed a lot since then).
And if it was once implemented, the source code is probably reusable (with same changes, beucase the coding of lichess probably changed a lot since then).
As a Lichess user, of course I always want more features, especially the ones like Swiss tournaments that I would probably use! But as someone with practically zero knowledge of how much work it is to maintain a giant codebase, it *seemed* like the option in Advanced settings to cap the number of games (like the existing "Disable Berserk" one) would be easier to maintain than a separate Swiss system.
Maybe the developers will decide to add a Swiss format someday. Maybe it will even be this thread that convinces them. But I've seen enough Swiss threads over the last 5 years to think it's at least worth discussing lightweight alternatives to achieve something passably similar. Anyway, I don't want to dismiss the passion of all those requesting true Swiss Tournaments, so don't mind me and keep at it!
Maybe the developers will decide to add a Swiss format someday. Maybe it will even be this thread that convinces them. But I've seen enough Swiss threads over the last 5 years to think it's at least worth discussing lightweight alternatives to achieve something passably similar. Anyway, I don't want to dismiss the passion of all those requesting true Swiss Tournaments, so don't mind me and keep at it!
I do not understand why the moderators and developers of Lichess waste so much time and energy defending "the supposed benefits" of Arena.
It's not about eliminating Arena tournaments, it's about adding new tournament systems: Round Robin, Swiss.
I suppose that means a lot of work for them, and in that sense I understand them: they have their hands full with so many projects and fixing day-to-day problems.
Unless new developers join Lichess and they take responsibility for scheduling and maintaining the protocols for running Swiss tournaments and Round Robin, I highly doubt we will see such events around here.
It is a great loss for chess. I like Lichess because it is free, it is libre, it is open and very respectful of privacy. I do not want to leave from here.
It's not about eliminating Arena tournaments, it's about adding new tournament systems: Round Robin, Swiss.
I suppose that means a lot of work for them, and in that sense I understand them: they have their hands full with so many projects and fixing day-to-day problems.
Unless new developers join Lichess and they take responsibility for scheduling and maintaining the protocols for running Swiss tournaments and Round Robin, I highly doubt we will see such events around here.
It is a great loss for chess. I like Lichess because it is free, it is libre, it is open and very respectful of privacy. I do not want to leave from here.
The Swiss and Round Robin systems would be muchly appreciated. Arena is a beautiful platform and I find nothing unpleasant or underdeveloped in its systems (the number of players issue is not precisely a developer issue).
This thread is built in such an interesting perspective of whether or not to implement the requested tournament systems. My take on that is that *a)* these should be endorsed and implemented, and *b)* should the developers have a problem with coding such features, I suggest there be a programme such as the Lichess Classes, whereby designated team leaders can have the authority to create and coordinate the tournaments in discussion, and hold responsibility to ensure the participation and punctuality of the players through extensive communication. These team leaders may then also be able to communicate with other team leaders to organise team battles using the Swiss and Round Robin Systems.
Since Swiss and Round Robin tournaments require a fixed time for the start of each round (upon conclusion of all games in each), it would seem wise to have chess puzzles as a means to wait for pairings, or allow players to spectate the remaining games of each round.
• Make use of the BYE system in the case of an unpaired player and perhaps a game with a bot to pass time.
There are many ways to checkmate. Let a trial phase be our check (it is good to check because it could possibly be a checkmate).
Thank you for your consideration.
@thibault @tpr @Ugalde @Toscani @Kwatschewsky @bufferunderrun I acknowledge your strong view points.
This thread is built in such an interesting perspective of whether or not to implement the requested tournament systems. My take on that is that *a)* these should be endorsed and implemented, and *b)* should the developers have a problem with coding such features, I suggest there be a programme such as the Lichess Classes, whereby designated team leaders can have the authority to create and coordinate the tournaments in discussion, and hold responsibility to ensure the participation and punctuality of the players through extensive communication. These team leaders may then also be able to communicate with other team leaders to organise team battles using the Swiss and Round Robin Systems.
Since Swiss and Round Robin tournaments require a fixed time for the start of each round (upon conclusion of all games in each), it would seem wise to have chess puzzles as a means to wait for pairings, or allow players to spectate the remaining games of each round.
• Make use of the BYE system in the case of an unpaired player and perhaps a game with a bot to pass time.
There are many ways to checkmate. Let a trial phase be our check (it is good to check because it could possibly be a checkmate).
Thank you for your consideration.
@thibault @tpr @Ugalde @Toscani @Kwatschewsky @bufferunderrun I acknowledge your strong view points.
@thibault @bufferunderrun I seem to remember reading that Swiss and/or Round Robin had already been implemented in lichess back in the days?! They have been discontinued due to lack of interest or had there been technical issues as well?
Would it be feasible to restore/reuse much of the relevant code base? Or would a complete re-write be required?
Glad to read (#21) that repeated pairings are fixed, but I, too, still hope for Swiss / Round Robin for our small club tournaments here at lichess. :-)
Keep up the great work, we all very much appreciate it!!
Would it be feasible to restore/reuse much of the relevant code base? Or would a complete re-write be required?
Glad to read (#21) that repeated pairings are fixed, but I, too, still hope for Swiss / Round Robin for our small club tournaments here at lichess. :-)
Keep up the great work, we all very much appreciate it!!
I would propose.
1. Organizers can set points for wins (example, win = 1 point or 2 like Sofia rule)
2. Disable repeat pairings (which you mentioned is already fixed)
3. Option for disable streak point system. Please :(
(With 1 and 3, it becomes 1 point per win, just like swiss. Please give priority to 3, because some players get undue advantage from walkover or easy wins)
4. Optional manual control option for organizers:
a. Manual Next Round start button.
b. Kick players out: In paid entry events mischievous players pay once and share password with friends. This is why we had to switch to other sites for high prize money paid events. No control for organizers.
c. More options for organizers in chat box. Like, Message players or send some alert to all (optional). Mute option to restrict unnecessary noise during important tournament announcements.
With these advanced options, I think Arena becomes Swiss :)
1. Organizers can set points for wins (example, win = 1 point or 2 like Sofia rule)
2. Disable repeat pairings (which you mentioned is already fixed)
3. Option for disable streak point system. Please :(
(With 1 and 3, it becomes 1 point per win, just like swiss. Please give priority to 3, because some players get undue advantage from walkover or easy wins)
4. Optional manual control option for organizers:
a. Manual Next Round start button.
b. Kick players out: In paid entry events mischievous players pay once and share password with friends. This is why we had to switch to other sites for high prize money paid events. No control for organizers.
c. More options for organizers in chat box. Like, Message players or send some alert to all (optional). Mute option to restrict unnecessary noise during important tournament announcements.
With these advanced options, I think Arena becomes Swiss :)
This topic has been archived and can no longer be replied to.