Multiple World Plot Development and Contingencies

Of the three major MMORPG’s on the market right now (AC, EQ and UO), only AC makes any attempt at an ongoing storyline.  UO has some "lore" from the previous Ultima series games, which makes no sense and doesn’t matter anyway:  any plot that happens in UO is usually by accident, and created by the players themselves.  EQ has some sort of weak background lore based on a combination of 15th generation Tolkien ripoffs and DikuMUD, but it never ever changes; the world of EQ is even more static than that of UO.  AC has an ongoing storyline, with events and such, despite the fact that the vast majority of players don’t care about it past the concerns of what new loot is available with the event, but at least there’s the attempt.  This section concerns itself with AC-like models incorporating an ongoing plot as part of its design.

In an MMORPG model, one has to resign oneself to the fact that a large portion of your plot is going to static, and therefore linear.  What this means is that if next month you are planning to have the Rovers of the Frost Barrens start raiding border towns, they’re going to be doing it on every iteration (server/shard) of your game world, even if those towns happen to be very well garrisoned in some worlds.  This also means that if in January you drop a bunch of clues and lore and such about an upcoming catastrophe into the world, but only one of the shards manages to find your clues, that catastrophe is still going to happen on all worlds in February.  If the world that managed to figure out the lore ahead of time deals particularly well with the catastrophe as a result, while all the other worlds get bombed into the stone age, the following month’s events will not reflect this divergence between worlds.  If you start diverging like this, eventually you are running as many games as you have servers, with that much more content development overhead.  This is similar to the problems outlined with open-ended storylines above.

Again, the situation relates to development power in the hands of players, but as it impacts the company storyline, it rests more on the technology you develop to deal with player power than the players themselves.  If, for example, you allow players to build houses through a dynamic construction engine that allows for building damage, repair, and the like, then there are definitely variable circumstances that will work themselves out.  Depending on what sort of frontier system you are using, the Rovers of the Frost Barrens might run right into a popular player built town at the edge of civilization, kill the NPC guards, and start burning buildings.  This sort of thing definitely changes the dynamics of one quest from world to world.

Occasionally, though, you still run into problems.  The business with the Herald Crystal in Asheron’s Call is a perfect example.  On nearly every server, eventually somebody killed the uber-godlike Herald Crystal despite other players trying to defend it, which was expected, and necessary for the next monthly event to make sense.  Every server, that is, except Thistledown.  Eventually, Turbine had to send in an invulnerable admin/player to kill the thing to further the plot.  Everyone felt cheated by this, and they had reason to.  However, it had to happen somehow.  Could it have been handled in a better way?  Possibly.  The point is that it had to happen.  If Thistledown had successfully foiled the plot int his manner, implementation would have had to scramble their asses off to give them a different event from every other server.  Not only is this horribly impractical, but you would get even more whining from all servers about being "left out" of the events happening on Thstledown, or on any other server besides Thistledown.

There is an interesting trick one can extrapolate from various hackneyed time travel novelists.  One theory about "changing the past" is that time is self-healing, and that while stepping on an ant 20,000 years ago might have some repercussions, eventually things will smooth over, and Rome will still fall, Hitler will still come to power, and the atomic bomb will still be developed.  If we reduce the scale of this example a little, we can apply the self-healing aspect of a multi-server plot to the overall chain of events, but allow things to diverge slightly on a server to server basis if we have the man-hours to implement it.

For instance, say your overall plot functions on a 2-month cycle.  You are just now implementing an event that involves refugees from a neighboring kingdom swarming into the play area, followed closely by a tribe of marauders that has evicted them.  You’ve already figured out that the next big event will involve an invasion by the players into giant territories to stop their impending threat.  During the two-month refugee/pursuit event, many different things can happen.  The players may, as expected, kill every giant they see, and at the end of the two month period, the local nobles rally and exhort the players to go kick some giant ass.  However, maybe the players are particularly brutal, and kill the refugees, which might lead to a temporary truce with the giants who can assume the players are on their side.  The players may fail to deal with the giants, and some towns and such may suffer as a result.  The players may be trying to fight the giants, but the fluid "frontier border" of giant influence is at a standstill.  Different servers can have very different versions of the event, but somehow they all have to lead into the next development.

The key here is to always have a contingency plan, and to be able to make one up on the fly if something really strange happens.  This is not unlike the normal refereeing of an open-ended plot, except it’s not truly open-ended.  You must have all game worlds coincide at the end of the cycle.  Looking at this as a graph, you start with a point (the event), then branching lines into various possible continuations of the event based on player action, then a reconvergence with rationales in place to bring the players back into line.  The preceding example might look like this, if you happen to use bad web tables in lieu of graphs:
 

EVENT
Refugees enter play area, followed by pursuing giants.

POSSIBILITY
Players kill the giants and save the refugees.
POSSIBILITY
Players kill the refugees.
POSSIBILITY
Players are overwhelmed by the giants.
POSSIBILITY
Players are deadlocked with the giants.
RESULT
Refugees are grateful and swear fealty to the local monarch, etc.
RESULT
Giants figure the players are of a like temperament and cease major hostilities.
RESULT
Giants have pretty much the run of the land in all but the most well-defended areas.
RESULT
The frontier of conflict stays pretty much where it is.  Neither side gains ground.
CONTINGENCY
Monarch declares that the players should go kill the giants before they can mass another assault.
CONTINGENCY
Monarch decides he doesn’t trust the giants and decides to pre-emptively attack them before they can betray the monarch.
CONTINGENCY
Panicked monarch pleads with the players to go strike at the giant homeland while it’s relatively undefended to get the giants the hell out of the play area.
CONTINGENCY
Monarch is worried by the drain on resources and lives, and asks adventurous players to sneak into the giants’ homeland via a newly discovered passage to strike a decisive blow.

EVENT
Players gain entry to the giants’ homeland to attack them.

If successfully implemented, a contingency web like this can give the players a real sense of consequence and accomplishment, rather than the feeling they’re just being railroaded into the next chapter.  You will also be far less likely to hear players complain when you have to pull a deus ex machina to kill the Herald Crystal because you didn’t plan for it.

One Response to “Multiple World Plot Development and Contingencies”
  1. *Applaud*

    I’ve been saying this about as long as I’ve played MMO’s. Problem is there’s too much emphasis on release dates and budget… …or at least that’s the message we typically hear, even as content delivered on time is lackluster or release dates get pushed back, and back…

Leave a Reply

*

© 2009-2018 Howard Collins All Rights Reserved

SEO Powered by Platinum SEO from Techblissonline