I wonder if it might be a good idea to pre-place a stone for each player in some regular fashion (depending on number of players and board size). Otherwise, if there are a lot of players, you may get really unlucky with collisions in the first round and not place a stone, which may disadvantage you for the rest of the game. Or am I being too paranoid? 
2 Likes
I strongly prefer not to introduce such a mechanic, since it creates a public, binding promise about behavior in future rounds. A player can always publicly announce their intention to submit standing pass and withdraw from active participation, except that it would not be binding and would not require the referee to step in by making such a statement.
Your proposal would benefit the other players by providing more certainty. However, uncertainty about otherās actions is fundamentally part of the basic philosophy and premise of the game of diplomacy. Further, decreasing uncertainty would not necessarily provide some other benefit, since a player leaving at an inopportune time fundamentally disrupts the game that is not necessarily ameliorated by certainty that they have left for good.
A marginalized player that feels that they are doomed to lose, might want throw the game in favor of a particular player (maybe to spite another). A publicly binding resign (confirmed by the referee) provides more certainty to the remaining players and rules out the possibility that only a public statement (from only the leaving player) might only be a ploy, where the player might actually return to active participation later by withdrawing their standing order.
I also oppose this proposal since it deviates more from standard Go (which is starting with an empty board and flexible strategic decisions from there). Starting with some fixed stones removes very important strategic choices, such as shape, positioning, and also who you keep as neighbors.
Since communication is allowed before each move, avoiding collisions is not really just a matter of luck, but rather a matter of how one can understand and influence the intentions of others, and adapt to the information gained through diplomacy. Of course, there is always some degree of uncertainty, but it is far from pure chance, and a skilled negotiator can do much to shape their own future.
2 Likes
I remember from earlier that this is your stance on resignations I donāt have strong opinions either way, so Iāll leave it up to what other people prefer.
While I agree with @yebellz regarding pre-placed stones, I would propose that the elimination rule only be effective after a fixed number of moves, for example three moves. Otherwise players might be eliminated in the first couple of rounds. I feel that after the first three moves, the possibility of a very early elimination drops a lot.
Edit: This is in the context of the elimination rule where players must have x/2 stones, where x is the number of rounds. This would mean a player that doesnāt place a stone in Round 1 immediately gets eliminated, right?
1 Like
Yes, we could change the requirement to 0.5x - 3 or something to get a more forgiving start. Maybe the coefficient of x should be slightly higher if weāre making the expression more complicated anyways, but itās hard to say before trying it in practice.
2 Likes
I also like the idea of this rule. I have a question though: Is the goal to be the last player to be eliminated, or is the goal to maximize oneās rank in a ranking that is determined by when players are eliminated? In other words, should players only aim for the top spot or should they try to āoutliveā as many players as possible, even if they canāt live the longest?
I felt like this question should be discussed still. Personally after considering all arguments that have been mentioned so far, I prefer if the rules only state that reaching the top spot is the goal, and it is up to the players which outcomes to prefer if this goal cannot be achieved.
1 Like
Right, in general we have a cut-off function, where having strictly less than f(x) stones at beginning of round x means elimination.
For the first game, it was f(x) = 0 (for the first three rounds) and f(x) = 1 thereafter.
@le_4TCās amended proposal is f(x) = x/2 - 3
The function does not have to be linear either (e.g., āa * log(b * x) - cā or āa * sqrt(b * x) - cā would result in a slower cut-off rate), but I think in general it should be monotonic and increase by no more than 1 between rounds (to avoid score quantization).
Thatās a very good point. I guess another way to consider it is, should the goal be being the leader, or last as many rounds as possible? Does the last person(s) standing in the game get all of the money? Or is money portioned out proportional to which round you leave?
1 Like
I cannot resist the urge to make this pedantic comment: The goal to last as many rounds as possible is not the same as the goal to outlast as many players as possible 
2 Likes
Yes, I agree and am aware of that. I was intentionally bringing it back toward the framing of winner takes all vs score maximization, which is more aligned to the earlier discussion of objectives above.
1 Like
I would say that the primary goal is to be the last player standing (i. e. get first place). If thatās not possible, aim for second place, etc.
Within a given placement, staying longer is better.
Itās sort of analagous to how we treat score in normal go. The real goal is to win, but in practice we think about maximizing score (well technically score difference in that case), and we would consider a move which loses points unnecessarily a āmistakeā in the abstract sense even if youāre still winning.
We could decide on a way to distribute a prize pool, but it felt like most people found that too complicated, so I would focus on the more basic āstay in the game longer than as many other players as possibleā.
3 Likes
Ok, that sounds good to me. Oh my god, did the three of us just reach consensus on this issue? I hope @Vsotvep likes it as well. What do the other playerās think?
For simplicity, I like the specific cut-off function of f(x) = x/2 - 3
2 Likes
Who would have thought this was possible? 
But I seriously believe this would play well. Iām assuming players have the liberty to go against these incentives though, if they see fit.
One quick thought: The number of stones on board seems a good indicator for how good a position a person has, and how much area they control on the board. One can also consider a slightly modified version of this rule, where the number of stones + number of all liberties is counted and compared to f(x) (arguably the more liberties you have, the better). This seems like a minor change to me, and possibly the drawback of complicating the rule is not worth it. But I wanted to ask about your opinion anyway.
1 Like
Yes, I think keeping it simpler is better.
An interesting consequence to this sort of rule with a cutoff function (regardless of the details):
Consider a scenario similar to the end of the first game, where red and blue canāt be captured, but they could capture white together. Red has more stones than white who has more stones than blue (this rule does result in a sort of stone scoring of course).
We might expect white to be captured, but actually red will just wait for blue to be eliminated (the blue stones stay on the board still afterwards). The result is that white gets second place.
Now if white had more points than both blue and red, then they would be forced to work together to capture white, and decide the game between them afterwards.
2 Likes
Yes, I think Iām agreeing as well, although I havenāt found the time to read the whole discussion I gave my approval of this type of rule here already.
The specific cut-off function is a tiny detail
3 Likes
Another possibility is that after blue lets white get a higher score than red, and blue then uses that as leverage to get both red and white to agree to the three-way draw. Of course, whether this would actually work out that way would depend on the specifics of how the players value the ranking vs drawing and their risk tolerance. It doesnāt work if they (particularly red) feels that second place is not such a bad threat in comparison to accepting a 3-way draw.
Since simplicity seems valuable, I think a cut-off function of
f(x) = 0 for x = 1, x = 2, x = 3 and f(x) = x/2 otherwise
is preferable over f(x) = x/2 - 3
(Or in other words, in the first case the elimination rule is effective after Round 3. I think many people would find this simpler)
2 Likes
Hm, Iām really not too happy about agreed drawsā¦ But I think this formulation from earlier actually takes care of this by itself:
(I meant by this āmaximize the number of players that are strictly below you in the rankingā)
This means that a 3-way draw is the same as each player getting third place. In particular, second place is better than a 3-way draw.
1 Like
Yes, I guess since we are not taking the winner takes all perspective, the agreed draw has diminished importance (unless foreseen to be inevitable) since the threat of throwing the game is less important.
However, to save time, perhaps the players can vote to end the game by unanimously agreement (as expressed through their voting) on the final ranking (including ties if needed and foreseeable) of all remaining players.
For example, in a game situation with three players left, and the eventual outcome being certain, they might all submit votes in the form āAlice > Bob > Charlieā (if thatās the appropriate ranking), or something like āAlice > Bob = Charlieā (if Bob and Charlie cannot avoid being tied).
1 Like
Yes, this is what I meant by this comment:
I think this is the natural generalization of a normal go resignation to arbitrary games. So for the earlier proposals with prize pools, it was natural that a distribution of the prize pool could be suggested and accepted at any time if everyone agrees, and for the current proposal I guess the player ranking is what should be decided. (I said earlier that staying longer is better, so technically the agreement should then include the precise number of rounds for each player, but that seems unnecessarily picky)
2 Likes
I mean at the moment, playing with eliminating players who have less than x/2 stones at round x also deviates lot from go, and is kind of like a battle royal type mode in a sense. Not that Iām against this variant at all, it does sound very interesting.
On the other hand though, it probably encourages playing throwaway moves every now and then to stay in the game. Imagine round n-1 and a player has n/2-1 stones. If they risk playing important moves which are likely collisions they might end up passing and be eliminated so they play a safe bad move (maybe in someone elseās territory eg eyespace of an alive group.) itās kind of like a timesuji for a round, a roundsuji!
Re initially placed stones, it could be something to consider. That or say one randomly placed initial starting stone per player, from a handful of reasonable starting locations. Iām thinking of games like risk where if you can choose your starting location and some players arenāt picking locations sensibly one player can get a large advantage. Maybe for future games though rather than this one.