I agree on this and think that following through on diplomatic threats, or retaliation against betrayal, are sound reasons to overrule the defined objectives. I attempted to analyse the given position in a “vacuum”, without the information of what had happened in the game up to this point, which is perhaps naive.
Nevertheless, if we are talking about which direction the defined objectives lead the players, I believe the above argument is still valid.
This depends on the player, I presume. I see Dave’s reasoning as severely flawed, and don’t believe retaliation, especially if not announced, is a good strategy. But we’ve had this discussion a few times in this thread already.
In Diplomacy, the game is typically not decided by clever board plays or very deep reading, but rather hinges on who the players choose to help and who they hinder.
Direction of play becomes much more nuanced, beyond urgent and big, as we must also consider who benefits from the foundations we lay, and who we are casting stones against. The understanding of such is even more perplexing, since all the subtlety and duplicity of language can be employed to shape the perceptions of others.
Ultimately, the winners and losers are chosen by the collective actions of the players, but often what predominantly guides these actions occurs in the communication off of the board.
This is not a matter of strategy, but rather a question of preference. If you were reduced to such a choice, would you care more about taking 4th place instead of 5th, while letting the person who put you there win? Or would you find more consolation in ruining their game as well?
Frankly, even if we think about these as repeated games in a series, I won’t look back at an old game and care that much about finishing 4th vs 5th, and I doubt many would remember much about the rankings beyond just who won in each game. However, playing for satisfaction makes for a more memorable experience. And, of course, the expression of one’s intent to do is vital as well.
I would rather have 4th place. I see no merit in taking revenge. Also, especially with this game, I would blame myself for not seeing the betrayal, and not the opponent who betrayed me. At least, assuming it was a rational betrayal, they must’ve gotten to a better position by betraying me than by not betraying me; moreover this has to be significantly better, since my opponent does not know if I would retaliate.
For example, in round 17 I was betrayed by le_4TC, who had agreed to capture Haze. But It was really my fault for being naive enough to believe le_4TC would grant me such a nice position with barely getting anything in return.
That even extends to daily life: only harm comes from revenge. Of course it’s bad strategy to announce that you don’t retaliate or never take revenge, since that makes you exploitable. Therefore this might all be a very elaborate ploy to trick people into mistakes in my future games. Or at least, that’s what I want you to believe ( or is it? )
The form of “revenge” here shouldn’t be viewed as a form of maliciousness that extends beyond just the competition of the game. Making a threat of retaliation, and even following through on it, is a potential tool to influence other’s play, not to harm them on a personal level. However, there is definitely a psychological aspect to it as well, in whether such a threat seems credible to actually work.
It’s important to also draw a psychological sandbox around these games. Even though a game of standard Go might be described as “fighting”, “violent”, “battle of life and death”, etc., we quite easily view that abstract competition, which we ultimately view as a joint exercise in recreation rather than actual conflict, as something separated from how we interact with that person outside of the game. With Diplomacy, however, it can be bit tricky to naturally view things that way, since the game revolves around interaction via language, yet, we still must understand that the concept of retaliation and revenge should not be an interpersonal attack, but rather part of only the artificial conflict of recreational competition.
Related to such concerns is how much psychological factors may influence the interaction between the players across multiple games. Ideally, each game should be a clean slate, and generally the rules for a series of games include prohibition of any explicit alliances, deal-making, grudges, etc. that carry across games (i.e., no one should offer a quid pro quo or make a threat that spans two separate games). However, how the psychological biases and reputations between players might carry over across multiple games is a much grayer area, and anyways very hard to cleanly define or draw boundaries around.
One could certainly argue that some of the discussion in this thread, such as expression of one’s intentions, preferences, and general strategic thoughts, falls within the realm of building reputation and revealing intentions/preferences such as to potentially influence future games.
(At the risk of starting a philosophical dispute, which is not my intention)
While I sympathise with this sentiment, I would argue that revenge =/= retaliation (although english is not my native language, so I might use the words in a wrong way). To describe what I want to convey: I think it is a good thing that certain actions generally result in bad consequences in modern-day society (i.e. there is something like “good retaliation”).
For example think of a criminal getting arrested by the police. This is retaliation to the committed crime, and hopefully demotivates some other people from committing the same crime.
Going back to the game, there is no law or neutral third party that could “punish” a player for betraying someone. In this context I believe that in-game revenge is justified as a means to retaliate.
After writing this I had a fun thought. Imagine a diplomatic go game with hundrets of players on a huge board. The players decide to establish laws to create order. If player(s) disobey these laws, a big group of players would intervene and punish the culprits. That would be like a miniature simulation of society. 
Just as an interesting thought experiment, would it be possible in theory to enforce this? Here’s the best way I can think of:
For each new game, completely anonymize the players. For a game played on the forum, this could mean creating a brand new account for each player, or perhaps reusing a small set of accounts (but shuffling them around between the players).
This still leaves the possibility of setting up an alliance in one game, sharing some secret information, and then finding each other in the next game using that information. This automatically becomes less feasible if the player pool is larger, and you can’t know which players you will meet in the next game. But to make “re-finding” a player in the next game even harder, we should make all private threads public after the end of each game (this makes it useless to agree on some secret for the next game). Apart from ensuring fairer play, this also just makes sense for a “spectator sport”, in the ideal environment all the spectators would even get to see all private communactions while the game is being played (like what yebellz allowed by sharing information from the players to the spectators via the referee, but implemented such that everything is automatically shared).
Even with these precautions, it’s impossible to remove the possibility of recognizing other players by the way they play and communicate. Still, I think such an environment is more fair than what we will be actually doing, where it’s very possible to carry grudges and alliances between games, even unintentionally.
I’m not advocating for doing what I described above, I don’t think the benefits are worth the hassle. But I will personally try my best to mentally isolate each game. Whether other players want to do the same is up to them, but I would encourage everyone to make a conscious decision about it 
It’s very difficult to enforce and even a bit tricky to precisely define where the line is. Players that are committed to this sort of meta-game cheating could always claim to have made certain decisions based on some factors, when they are actually motivated by others. It’s impossible to prevent or monitor communication over external channels as well. Ultimately, it comes down to the players following an honor code.
Note that even though the game allows deception and dishonestly in the interactions between the players, the players should remain honest when dealing with the game master/implementation. At the end of the day, it is just a game, with nothing really at stake, so hopefully people commit to following a general honor code of fair play (which is often summed up as “don’t lie to the game master”), even though the game may inherently involve duplicity and betrayal between the players.
A lot of diplomacy servers offer the option for anonymous games for this very reason. On WebDiplomacy, a lot of the serious tournament games are anonymous. On Bounced, all games are anonymous, and players have the option of not revealing their identities even after the game.
This is probably the easiest work around to achieve some level of anonymity, although, as you mention, there are still risks that some might recognize the particularly pedantic speaking styles of others.
However, we also have to be careful about players that have multiple accounts on the forums. The game master should always clearly establish who is actually involved in each game and make sure that one person is not controlling two or more positions in a game via dummy accounts.
If players are determined to cheat, it’s impossible to reliably uncover all of their side channels, which may even exist off of the forums.
I like the idea of making all chat available for the spectators. However, the biggest issue is probably keeping things organized, and checking that everyone that signs up for a game is okay with revealing their chat logs. If the solution involves making all PMs threads public (after the game), then we might even need new forums categories just to help organize them all. Revealing everything to the spectators during the game does carry some potential risk of a spectator (perhaps even inadvertently) spoiling something to the players.
Yes, by “enforce” in this context I perhaps meant something closer to “provided that everyone follows the rules of the game, make sure that each game is isolated as much as possible”. Even though we can’t stop determined cheaters, we can make it clearer what is and isn’t allowed. With the game as we’re playing it now, we can’t have any rule forbidding keeping grudges between games, because that’s impossible to control, even for a player who wants to follow the rules.
To be clear, having some alliances and grudges bleed over from game to game is not a terrible thing. It just puts the game in a slightly larger context, and makes it a bit less “pure”, but hopefully it won’t make it less fun to play.
A sort of practical way of doing this would be to have the referee be a part of all private threads. They could mute the thread so that they’re not overwhelmed by notifications, but by checking in every now and then they could choose some interesting parts to share with the spectators in the kibitz thread.
Yes, this is a risk, but this risk always exists if we want to allow discussion of the ongoing game. Personally I think it’s worth it for the improved spectating experience (which also includes the ability for the players to understand what happened after games).
I think the most important opinion in this case is from @Vsotvep, since this will lead to potentially some extra work for the referee. Do you (@Vsotvep) think it sounds like a good idea to allow players to invite you to their private threads, to signal that all that information can be shared in the kibitz thread?
We can discuss whether inviting the referee should be mandatory or not. The advantage of having it be mandatory is that the referee can then answer questions like “What has been said between A and B before this move?” from the spectators with certainty. The advantage of having it be optional is that players can choose to use a private thread if they don’t trust the spectators to not interfere.
I’m fine with being present in private threads, but what would probably be easier for me, and perhaps more interesting as well, is to have players submit a Malkovich log where they explain their moves themselves. That way also hidden thoughts, like “we’re discussing collaboration, but I’m planning to betray”, get forwarded to the kibitz thread.
Diplomatic Go: the only way you can call binging Among Us studying.
I bet someone can write a psychology dissertation about Among Us.
Game Theory would be more interesting, I think. If we model the map as a graph with each point representing a room or smaller unit, and each connection between two points being associated with a positive real number denoting travel time, is there a Nash Equilibrium for movement, task completion, murder, &c.?
How about eliminated players have a choice to exit the game discussion (for example if they want to join the kibitz thread) or stay in the game discussion? I don’t mind if eliminated players want to keep influencing the game by means of discussion, for example to remind players about promises made in the past, or to show their support of a player. Maybe they don’t intend to influence the game but just enjoy discussing with the remaining players.
Well, that sort of blurs the lines on what elimination really means. In a game where the discussion off of the board is vitally important, one might wish to eliminate a player for fear of their diplomatic influence (and their ability to provide strategic help to one’s rivals).
Of course, with a player pseudo-eliminated (no longer able to play more moves, but still able to talk), some of the other players might just choose to stop discussing with them. However, others might seek to use them for advice and influence.
I think it makes a lot of sense that you can’t influence the game anymore once you’re eliminated. To make a slightly dramatic analogy, think of being eliminated as being killed… while the ghosts are able to spectate, they are not allowed to interact with the living 
(it also opens up the possibility of betraying a player, without other players necessarily knowing that you betrayed them, since they’re not alive to tell on you)
We start tomorrow, right? Better get things ironed out…
Most importantly, if dead players can stay in discussion, we’ve got the problem with incentives again!
I think things have been ironed out already.
So what’s the current discussion over? Looks from where I’m standing like basic rules are still being debated.
The major issue was which goal players should have and how players should be eliminated. That has been solved after pretty much everybody agreed on le_4TC’s suggestion above.
There was then a discussion about a minor point in this rule about if the strongest two players being drawed should count as a shared 1st place or as a shared 2nd place (which is a really pedantic kind of thing to discuss, but you know us).
We then had a discussion about whether it makes sense to retaliate or not, which had nothing to do with the rules and more about general game theory / personal preference.