Dispute Resolution via Instance Splitting

Psychologically it is very different, and this is important. You are not forceably ejecting someone from your group -- you are leaving. It also gives each member, without thread of coeserion, to choose who they want to follow. It also maintains equal power distribution, no one person (a leader) has more ability than another.

Kicking is an us-vs-them, someone-is-right/someone-is-wrong tone. With an option to split, someone who justs wants to break off and do SoC has an amicable mechanism that lessens heart-ache if the group doesn't want to cooperate with their wishes. In short, kicking produces a winner/loser categories, splitting may produce a loser, but the mechanism itself does not force a loser.

The implementation has many steps; but each person has only a single dialogue box. For the person who wants to split, they are shown a dialog to confirm and select those who they want to allow to follow them. For each other person in the group, they have a single dialog to ask if they want to follow the member who is splitting.

The actual number of actions required by the user is equivalent in user-level complexity to any sort of voting mechanism. You need to call the vote, and then collect responses. It is more complicated than the "dictator" model, where the group leader maintains the ability to kick anyone -- however, that option is quite draconian.

The henchmen-replacement issue is a side point; since both kicking and splitting need something along these lines. So, any additional complexity here is shared.

You are making a host of assumptions:

(a) that the person who wants to leave (say doing an SoC) wants
to even continue the mission;

(b) that there has to be a group that 'continues' and a group that 'fails'
(why can't both groups disagree, and both finish in their own way);

(c) that if he wants to continue the mission, he can't finish it
with henchmen (I've done almost every mission /w henchmen); and

(d) that only one person leaves, and that two people playing together
don't leave togther.

Overall, the split mechanism introduces very small amount (if any) of additional end-user complexity, while providing a great deal of options for a group in disagreement. Where a 'kick' is openly hostile -- a group of people is forceably causing another individual to disconnect -- the 'split' option is based on personal choice, and therefore allows for win-win situations.

it couldn't be exact... through this method you could very easily turn 1 rare item into 8. Simply keep splitting until each person has their own copy of the item. Not slamming the idea, just saying that a bit of refining would be in order. Perhaps the group to leave would NOT get any dropped items brought with them, again putting the responsibility on the leaving group to make sure everything has been taken care of.

Here's the way I see it.

1) If you're doing it to leave a genuinely annoying arsehole, then it isn't different very much. But, you do get a henchman, so you're not a player down.

2) If someone is using it to be a smartass, then not only do you not lose your progress in the mission, you continue with a full party (if you started with a full party)

Overall, this seems to be a stellar idea.

exactly, the item would have to be reproduced or deleted or whatever. Imagine the server having to split one instance into 8, and having to sort out all the items. What would happen if they left a gold item until it went unclaimed (as in anyone can pick it up) and split?

what happens if a rare drops on top of a cliff you cant get to until later, and then a person starts being stupid? if you leave you lose the item but if you stay you dont finish the mission

(a) & (b) Well the way I was think was that presumably, in your example, the point of it not being a "kick" option is so that the other person isn't, you know, kicked from the mission. And if he still wants to do an SoC, that would still be very hard alone (but I see how that doesn't really matter [see (c)]).

(c) I overlooked the henchmen. I can see how that could work.

(d) simply isn't an assumption at all on my part. It was in fact based on your example, hence the reason I said "in this example".

Just have a Kick button and have done. People will abuse no matter what system you have, with or without a kick button. We live in enough of a nanny state as it is. If you get kicked by a bunch of abusing Guildies, chalk it up as xp and don't team with them again. Even the biggest missions only take 2 hours and tha majority only take 30mins so its not exactly a hugh loss of time and effort.

You would need to automatically have the henchmen come in, otherwise the potential for increasing farming would be massive:

A guild could form a group of 8, rush through the level to the target monsters really quickly and then all split. Simple 8x loot .

All the details like "loot splitting" are just that: details. ArenaNet will certainly figure obvious things like that out, and anything they don't figure out they can still just change a week later when We, the Players Who Will Find Exploits, bang on it.

A solution to loot splitting: Each and every item is reserved for a certain person. I assume for this idea, that an exact copy of the map is created, and each group gets one. For example, team A gets original copy, and team B gets the copied version of say Underworld. Team A will only have their member's loot in it. The loot for team B will be ported to team B's map.

So the replacement problem should Not be automatic. What if you have an Ele/mo but is a dedicated healer? Would you really want Orion to step in? As such, I would suggest that the leader choose. The original leader that created the party in the first place or the person that chose to split.

The voting process could be similar to skipping cinematics. It would tell the total number of people that voted. . . however, what of those who refused to decide? Would you implement a timer? If so, what team would be designated default?

I lke this idea. Today we ran into another "loophole" regarding this. We were "rushing" towards Thirsty River and we are like at the last segment. Now some idiot in the fray decided to go afk , all of us rush and do all kinds of "pull" tactics to get the sprinter through [which is fun tactics by itself] . Anyway the sprinter [and all of us] die before he reaches the entrance, and now we just need get resurrected at the shrine...

BUT the damn AFKer is standing in the middle of nowhere [obvsiouly still alive] and there we sit. We can't get rezzed at the shrine, we can't KICK the afker [or split from him] to trigger a party rez and inevtibably we are forced to redo the ENTIRE trip [which is 2 areas to cross].

Now mentioned split idea would be ideal to solve this problem too.


Other reasons why i might consider the split idea as cool [and no, this is not what the idea is for i know, but i think it would be really nifty :P]

1. Ability to have more coordination in terms of having 2 leaders. For example, splitting up in 2 groups and one group going one way and another group going the other way , for example to attack from 2 sides. Obviously can do it now, but ive found unless you know everyone in the group it will take longer to sort out who goes where than simply splitting up and following your mini-group.

2. Some instant PvP action in the middle of an instance :P [yea yea open for abuse, but i tihnk it would be fun to take a party of 8 and split up and see who can get across the map first.....even if it means to kill each other on the way]

um i dont see any connection between the idea and what your saying
its not just splitting into 2 groups
its on group entering into a completely new instance
its the same as if your all in one town, and one group of people exit the town
its impossible to interact with this other group

There is still a way it can be abused. For instance say A died and a gold storm bow drop for him, the rest of the team happen to be guildies, so they say some craptacular thing like "Hey you n00b! You died!" and voted to split him off to a new instance.. and *yoink* if you get what I mean.

I'd say under this scheme items in the immediate vacinity of that target ally should be carried over with him to the new instance, that would deal with 2 scenarios:

1) AFKers (since they are far far away.. tough! Free gold storm bow for those who worked!)
2) Thieves (Like above)

I think items should stay/go with the player. Items assigned to no player disappear (you didn't pick it up, it's gone.) Solves issue.

I see the point about the class based subs - you don't want your Me/W tanking IW/mantra vs physical type replaced with Dunham either. Selecting from the standard list of henchies should be fine.

i like the split option that would also stop ppl trying to scam a run to droknars.if they dont pay half at snakes bend,split and see if they survive.

I was in a group heading to lions when out of the blue the group leader and someone else got into this mad drama fight, about I don't even know what now. I ended telling them to fight or leave, they didn't listen to me. We did end up making there. But no I don't think there should be any voting, spliting, kick options, because it would lead to many abuse issues. You're best bet is to leave if need be, there are plenty of nice people out there that will help. That's what the guilds are for. And someone who mentioned about being able to kick someone out if they didn't pay, I agree.

I have to say I find "conflict resolution by splitting instance" overly complex and open to exactly the same abuse as both the present system (penis-drawing, aggroing mobs, "give us teh gold item or we leave" etc) AND the votekick system (a griefer guild can gang up on single players).

The only real difference between a votekick and a split instance is that it adds complexity, and is designed to have all _well-behaved_ players leave the instance to go to another instance instead of kicking the misbehaved player back to base.

It seems to me the purpose of splitting instances would be mainly to spare the misbehaved player from being punished for misbehaving, and sparing the well-behaved players from having to confront the misbehaved player.

And if the purpose of the excercise is to maximise _everyones_ happiness, then that is a good way of doing it.

However, I think I should point out that by making it easy and painless to split the team, you'll make this a very common event. My guess is that you'll get split instances pretty much every time you take a PUG out on a mission; every little infraction, real or imagined, will lead to a split. After all, all that happens is that the useless monk gets replaced by Alesia.
On the other hand, a successful votekick will be a rare event (there'll be tons of unsuccessful ones, though). Actually I don't even think the kicked player should be replaced by a henchman - there SHOULD be a cost involved in kicking someone. The votekick is for when a player such a nuisance that you're better off without him; it's not there to allow kicking a player just because you feel Little Thom is a better tank.

Technology-wise I also think it would be difficult to do; the "simply copy the chunk of RAM" argument doesn't hold, as the problem is game mechanics (e.g. avoiding duping, griefing, and exploiting) not cloning the instance.

Personally I think the choice is between having a votekick and having the present system. I don't think instance splitting is a realistic option.

I never thought much about these abuses because I've played WoW for months before joining GW and we never had anything like kicking people from instances. Even though I did a majority of my end-game instance runs with guildies, I've done PUGs and there was never a lot of need to kick people.

I think giving the leader the option to kick people who idle purposely/afk during missions and so forth is justifiable. Abuseable? Maybe. But I'll give the doubt that 90% of GWers will not kick people abusively. Why kick people out in the zone after you grouped? But then thinking on that, say you get a full party up to somewhere and you beat up a boss, he drops 300g. You boot everyone in the party. Poof, 300g in your pocket.

I have mixed feelings about a voting system too. :[

This community needs to blacklist people who mess up the gaming experience. :\

The issue is never people who go AFK; they've effectively kicked themselves. Sure, maybe they get through a mission without doing any work, but that pales in comparison to real abuses.

The real problems are when there's a mission quest item, and someone picks it up and refuses to finish the mission. That one person can stop the whole mission from going on.

That's why some forms of vote-kicking would help, at least. Right now one person can grief seven. If it required a otherwise unanimous vote to kick one person, then seven people could grief one player, but that's way less likely.

Thank you all for your fine critique; it is always helpful.

You didn't really outline any of those abuses, nor how they stack-up to the current abuses of the existing system. To back up your claim, it'd be nice if you listed abuses that you feel this system would have rather than making a blanket statement.

I agree that this is a complicated option. Are you saying that this introduces a democratic process where it should be a strict dictatorship (by the person who formed the team)? That's hard to argue with -- dictatorships are certainly much simpler. However, I think you're just moving the potential abuses around, not resolving them.

Penis-drawing is a separate issue -- it happens because drawing is anonymous; if you have each person have a different color crayon on the map I think this behavior will be significantly curtailed.

Extortion usually happens at the very end of the mission, and I feel instance splitting would solve it. If a significant number of players want to do a sit-down-strike, then you can just split -- leaving them to sit by themselves.

There are two sorts of complexities. User interface complexities and implementation complexities. The vote splitting is clearly has a more complex user interface than the current system or a system where the PUG leader has strict dictatorial control. However, the vote-splitting has exactly the _same_ level of complexity as any kick-voting system. Someone must call the vote, and people must vote.

The implementation complexity is really out-of-scope of this proposal; it could be very simple to implement (or impossibly difficult) depending upon how their system is structured. Speculating on this, one way or the other, is probably not helpful. Let's stick with game effects in this discussion, as Arena Net will of course, have to bring implementation complexity into their final decision.

Yea, it's a "I'm OK, your OK" approach. I wouldn't say that it avoids confrontation though, the act of splitting and asking other people to follow is clearly a confrontation. Further, I wouldn't go so far to assume that there is always a "misbehaved" party (and, by contrast, one that isn't).
Many times there are just honest disagreements that are not resolvable.

Ok. Let me try to understand this issue. Assume that I'm irritated with my PUG (the healer sucks, and an elementalist is tanking). I've got a choice, I can map-travel back to town and abandon the party -- or I can stick with it and live with the pain. My "investment" in the group makes it likely that I'll stick with it the further along we are. So, your point is that I might be _more_ likely to abandon the group if I could continue on my own with henchies, rather than being forced to restart from town? Perhaps you are correct. However, if each team is filled-in by henchmen, then I don't see the problem.

Personally, I think you overestimate the effect -- if I wanted to do it with henchmen, I'd do it with henchmen from the start. Further, I'd like to remind you that disconnects when things turn even remotely sour in the average PUG is already around 20%.

That's not a bad idea. I like there being _some_ penalty for splitting. Perhaps henchmen are filled-in to be N-1 slots; ie, you end up one down (but at most one down). Or, perhaps you make it a strict currency charge for the one calling the split?

That said, instance splitting has essentially the same user interface complexity as any vote-kick system, with the primary difference being that instance splitting does not force "winners vs losers" mentality. The implementation cost is just not knowable on this forum, while it obviously will be part of Arena Net's decision making, it shouldn't be a factor in this discussion.

Thank you all for your thoughtful comments and feedback.

Agreed. Maybe split groups would be likely, but it's probably better than what currently happens in the worst situations. And for the purposes of player frustration, the worst situations are what matters. If there end up being a few extra splits where the group might have toughed it out normally... well, that's a small cost for taking care of the worst offenses.

That's almost genius. If you just make it cost some money for everyone who splits off, or for the person calling the vote, it would cut down random splitting quite a bit, while still leaving a fine out. It wouldn't have to be high, and it could even be on a sliding scale, depending on: (1) how many people join you (if you think six of the other seven people will join you in the split, then the offender is probably the eighth guy, and the cost should be low, but if you split off alone, maybe you are the problem), or (2) how many times you've split off "lately" (which is harder to define, but there could be a penalty for the guy who constantly splits the group).

I imagine that instance splitting is still very unlikely for technical reasons, but I'm still a big fan of the general concept of voting-to-kick, which this would be one version of.