LOL. I missed this thread and just proposed two options at game formation in the main Rengo thread:
Based on my experience on DGS, I guarantee you that the VAST MAJORITY of Rengo games with a large number of players (for pretty much any definition of large ) will time-out.
I’d propose that at game creation, the organizer has the option to allow a time-out to either
act as a pass to the other team
play moves to the next player in the rotation on that same team, possibly with a shorter definable time-out.
The remaining question is what happens to the player who times out. There are several good ideas for that… perhaps another game organizer option?
As for cheating, this is a game for fun online. If someone wants to cheat, they’ll find ways to cheat regardless of this feature.
It would be really amazing to have a hundreds vs hundreds slow-correspondence rengo. It could take years to finish and each player might only get to play one move over the entire game, if at all. This would be more of a collective experience, rather than a typical game, but it would be really interesting to see how it played out. However, given such a huge participation list and the likelihood of someone missing their turn, this type of thing would need to be handled gracefully.
I think that if a player times out, they should be immediately removed from the rotation in their team and the next player in the rotation gets to make that move for them. Also, I like the idea of requiring some threshold of players voting for the resign (perhaps 50% or more, so that in a typically two-player team, one player is enough).
My issue with removing players who timeout (instead of treating it as a pass) is that Rengo is of course easier with a lower number of player, so that makes it easier for the remaining persons.
In other words, a team is rewarded for having one of its members timing out, which is not intuitive.
Maybe the behavior of handling certain things like this should be an option in the game settings? Some game creators might want stricter “timeout equals pass” rules, or even “one player can resign for the entire team” or “timeout equals loss” rules for a more, serious, competitive rengo match.
On the other hand, some players might want to use the rengo feature to create a more casual game with flexible and forgiving handling of contingencies where individual players quit or time out.
Some players might not want to play traditional rengo at all, but rather just use the format as a convenient way to manage (a prearranged) game of team go, where private discussion among the team is permitted (in a separate channel, like a group private message thread on the forums).
Actually, we’ve played a few games of team go via OGS and these forums before, and using this rengo feature would have made things much more convenient. The way we did it in the past was to just have the team captains play against each other in a standard, unranked game, and then consult privately with their team before each move. However, using the rengo feature would be a bit nicer since it would rotate who gets the duty of actually entering the move (motivating more engagement from everyone), and everyone would conveniently have the game on their homepage, rather than having to bookmark a link and potentially forgetting about it. For such a use case, I think the more flexible handling of timeouts would be more convenient.
I very much would not want this change; I like it how it is now. Sometimes you’re going to get a troll, or someone who, through no fault of their own, simply makes a mistake, as we all do, and times out, but that is already a danger on OGS, the only difference here is that it could be your teammate.
I think part of rengo is managing time and leaving enough for your partner. You’re not playing alone.
Another advantage of the normal way of doing it is that if one player is playing more quickly than they should for the time control, their partner can still make use of that time.
I think we should be emphasizing the aspects of rengo which set it apart from normal go, not shoving them under a rug where they cease to be relevant.
That does sound like a very useful use case for some of these suggestions, but I would not want it unless it was an option so that we could still play normal rengo with a shared clock, since that’s the core feature.
Settings overload is certainly a valid concern. To avoid having too many options, I would strongly advocate for the most permissive and flexible handling of timeouts and resigns:
Each player has their own clock.
Timing out kicks that player out of the game, but gives the next teammate in the rotation the opportunity to make the move (i.e., timeout does NOT equal pass).
Resigns requires 50% or more of players clicking on resign. With more than two players, maybe the number of resign votes tallied so far could be indicated in the interface some how.
@qnpnpmqppnp raises a point about the above behavior being unfair to players that want a serious rengo game that punishes timeouts.
However, I think that a stricter rule of “timeout equals loss” or “timeout equals pass” might make rengo games highly unlikely to finish for larger numbers of participants. It makes the system less flexible to support casual games where these problems are bound to happen.
The more flexible rules that I describe above could even still support the possibility of “loss on timeout” or “pass on timeout”, if the players agree to play that way, since the next player could simply pass or the remaining players could agree to resign.
I disagree: if it be practical, let players choose for themselves amongst the available options, and only as a last resort should a feature be removed to “clean up” the settings of something. Obviously this difference in philosophy is going to impact our views on this.
As stated above, I very much do not want this unless it can be disabled. It would be like saying that either teammate can play any move. That’s team go, and there’s nothing wrong with it, but it’s not rengo. Separate clocks could be a useful option, especially for correspondence and many-playered games, but not at the expense of vanilla rengo, where time management becomes, like your play, something about which you have to think of not only its impact on you, but on your teammate as well.
I’m more ambivalent on this, though I would greatly prefer still having the option for any timeout to be a loss, since in the basic use case of 2v2, the game is fundamentally changed by allowing a player to time out without at least voiding the game.
Sounds fine to me. Seems like some solid logic to pick >=1/2 as the fraction to require, too.
That is a very roundabout way to get expected behavior.
“Pass on timeout” could be a useful time setting for all games, not just Rengo. E.g. “10s/move, pass on timeout” might be a more fun way to play a fast game.
But the simplest solution to making large Rengo games playable seems to be something we need anyway: give organizers the power to kick out players at any time. As long as they are paying attention, they can get rid of absent players before they cause a problem.
I had never thought of that before, but that sounds awesome. I would always check that on fast games if it could be ranked, though it probably wouldn’t be. Would still come in handy for games which have to be unranked anyway for reverse komi, though.
EDIT: interesting observation, the reason I am okay with this despite it arguably being just as big a change to normal go as it would be to rengo, is that it clearly and unambiguously is a disadvantage (almost always severe) to the player who timed out.
That’s a fair point. There would probably be cases of players getting kicked for playing poor moves, even though that’s a part of rengo.
Another option is a vote involving all players on both teams, as in many FPS games, but that may suffer from the same issue, as well as making it more difficult for a well-meaning director to boot a genuine troll, which is the main use case here.
Good point. I would like this kicking functionality in some form.
Yeah. I don’t think most games would have this issue, but it would happen.
In this instance I can see the strong benefit of that being an option, though based on watching current games, I’d make clock-per-player the default … maybe shared clock would be default for invitation-games…
) will time-out.
