Within the MMOG industry there are two primary server structures
employed to house the player base: Separated shards of some sort
(typically referred to as 'servers' though this is misleading from a
hardware standpoint) and a more unified, single-world structure.
Primary examples of each are World
of Warcraft, which employs a
shard-based structure that breaks its considerable population into
separate worlds, and EVE
Online which houses its entire player base in
a single universe.
Since the first gamer began to take notice of this distinction, the
debate has raged on forums and wikis across the internet about which is
a better type of server architecture, and which type better serves the
MMOG audience. The arguments continue as each new game gets announced
and players make their arguments about the merits of one system or
another.
Today I'd like to try and discuss this from an impartial standpoint,
and attempt to outline the primary pros and cons of each style of world
building. I've also come across the thoughts and concerns of some
industry insiders on this topic which I'd be happy to share with you as
we go along.
Social Fragmentation
src="http://www.tentonhammer.com/image/view/92822">
It's
not possible to have this discussion without talking about the
separation of players that occurs in a sharded infrastructure. By
segregating your players into entirely separate worlds with no
connection to one another you put hard-coded limitations on their
ability to form social bonds that extend beyond those boundaries. Out
of sight, out of mind. If you place two groups of people in two
separate rooms with no doors or windows attaching one to the other, the
odds of one group interacting with the other are about the same as
taking on Deathwing solo wearing nothing but a slightly-damp loincloth.
Meanwhile, the primary focus of a unified server is to allow every
single player the opportunity to interact with every single other
player within this game's environment. This structure allows social
bonds and communities to spread themselves across the entirety of the
game's player base, leaving nobody left out. It also allows for every
type of player-made content (that includes guilds!) to potentially be
experienced by each and every player.
That's the theory, anyway. In fact, most of the time when you read this
particular argument, it swings so heavily in favor of a unified server
that you'd wonder why it's being discussed at all.
Well I'll tell you why: People
aren't nice.
border="0">
src="http://www.tentonhammer.com/image/view/92819">
I know that makes me sound like a misanthropic shut-in, but it's the
truth! No where is this more evident than in online games that offer a
unique combination of anonymity and total lack of social
accountability, coupled with the ability to adversely affect another
person's enjoyment of their game time. And if an evil-minded griefer
gets it in their head to make the lives of their fellow players a
living heck, the environment offered by unified server offers them a
bountiful variety of targets without restriction. In fact, I'd argue
that a unified server is an even more appealing stage for these
hooligans to raise a ruckus, as their infamy would be limited only by
the heinousness of their acts and could potentially spread to the
entire player base instead of just a single server. And for better or
worse, infamy is what many players online seek celebrity status among
their peers, even at the cost of their respect.
One might argue that the lack of a fragmented player base would mean
that these misfits are caught and chastised more easily, but I'd argue
that it is easier to hide among such a crowd than it would be in a
shard-based environment. After all, a face in a crowd in Podunk, IL is
a lot more likely to get picked out than if that face is hidden among
the teeming throngs of Manhattan or London.
And what about those crowd sizes? On October 30th, 2010 more
than 3000 players clashed in EVE, in the largest space battle the game
has seen to this day.
After just a few moments of the battle being joined, players began
suffering lag and input delays that ranged from 30 seconds to more
than 5 minutes. In a game where the single press of a button can mean
the potential loss of hundreds of hours of work, this excruciating lack
of server performance would end up causing the battle to be
indefinitely delayed. Both sides basically ended up stopping the fight,
due to the inability to perform at a reasonable capacity, and went home
without resolution.
Keep in mind that was with just a few thousand concurrent players. Now
let's imagine for a microsecond that the entirety of
style="font-style: italic;">World of
Warcraft's concurrent online population (which is
estimated around
85,000 at any given moment across all active servers) were shoved
together in one copy of Dalaran. To put that in perspective, this town,
even on busy servers, typically houses no more than a couple hundred
players at most, including the ones dancing AFK on the mailboxes. Now
that you're done hyperventilating from claustrophobia, I think you can
see why some games are simply not cut out for large-scale crowds. Aside
from the physical limitations of avatar space and breathing room,
there's also the technical aspects to be considered, such as the
aforementioned lag and performance drops that are frequently associated
with large gatherings of players in a single location.
border="0">
src="http://www.tentonhammer.com/image/view/92826">
Another benefit that social fragmentation offers to those participating
in sharded universes, is the potential to avoid excessive clique
formation, and content monopolies. If, to re-use an earlier example,
WoW were to somehow magically accomplish the impossible feat of
becoming a single unified world, it would not be long before a single
uberguild would form that would only accept the absolute most elite
players into their absolute most elite club. This activity already
exists in WoW on each individual server (and many other MMOGs
as well) but the fact that each world is separate from the others
allows for a vastly larger amount of the overall game population to
participate in these best of the best guilds. On the imaginary
unified server, all players that considered themselves good enough to
apply to that single top-ranking guild would find the competition
unfavorable to say the least, as they butted elbows with tens of
thousands of other players for the same honor.
World Building
href="http://www.gamesetwatch.com/2009/05/opinion_designing_a_single_ser.php"
target="_top">
According to Divide By Zero Games founder James Portnow, the
world
itself would need to be crafted in a very specific way to handle a
unified server in a reasonable manner. Primary features this world
would have to possess would be an incredibly large infinite, if at
all possible procedurally-generated terrain environment, and content
that could also be generated on the fly to fill such a massive space.
The idea being that regardless of your population size or density,
every player would receive approximately the same gameplay experience.
Further features such a world would need in order to feel immersive and
cohesive, would be player-created content of an incredible degree:
housing, cities, governments, kingdoms, etc. The mere fact that you
intend to have such a limitless population cap implies that your
content must also be limitless. Yet there is no way to manually craft
limitless content, and so that onus must be placed back in the hands of
the players in the form of content they create themselves. That content
can be social, material or economical, but the ideal implementation
would include all three.
src="http://www.tentonhammer.com/image/view/92825">
Star Wars Galaxies
was among the first well-known MMOG to offer the
widespread availability of player-generated content to the subscribers
when they rolled out their Storyteller system in 2007.
Many will claim City of
Heroes did them one better when the Architect system
launched in 2009, but I feel this is simply a matter of personal taste
and both sit on equal footing in many regards. Through
the use of both of these games' user-generated content systems, players
are able to set up their own stories with unique dialogue, plots, items
and more. Both systems were lauded as welcome additions to these aging
MMOG titles, but a few murmurs persist along the lines of if
I'm paying to play a game, why should I do the developers' jobs for
them? This sentiment continues to be common among critics of
style="font-style: italic;"> Star
Trek Online's impending launch of their own user-generated
content toolset, The
Foundry. Both SWG's and STO's toolsets required countless
hours of developer manhours and resources to create and publish, which
some players still feel could be better spent creating actual playable
content. It's a valid complaint to make when you consider that these
titles are just a few among hundreds that players can potentially spend
their hard-earned money to subscribe to on a monthly basis. If a major
title were to try and launch today that required all of its players to
create all of the content that they experience, I'd bet you ten
quatloos that it'd get quickly buried by the swarm of other titles that
freely offer content by the handful to players without asking for their
creativity or effort just their money.
As for the world itself, there's a fundamental downside to
procedurally-generated terrain and that is that they are not, by
definition, hand-crafted. I doubt there is a single player among you
dear readers that hasn't occasionally stopped to take in the scenery
around you in one MMOG or another. The occurrence of such breath-taking
vistas would be a nigh-impossible statistical improbability when
generating an endless terrain using numbers and variables.
The same would be true of the content created to fill such a vast and
unappealing landscape. No matter how robust you make your madlib-like
spawning system, the chances are that even the most inexperienced RPG
player will eventually spot the patterns of coming across another
[some race] being [attacked/threatened/captured/etc] by [other race]
at [random setting] and will quickly tire of the repetition. There's
just nothing as inviting and interesting as watching content unfold
that's been created with a purpose by a dedicated content team. But if
you have a gigantic world to fill, you're going to run out of money
before you run out of space to fill no matter how good that team is.
Variation of Mechanics
I have a secret. It's a private little thing that I try and keep to
myself when among crowds, but in order to illustrate one of my points,
I will admit this shame among you all.
I
like to roleplay.
src="http://www.tentonhammer.com/image/view/92824">
I enjoy immersing myself into the fantasy of being the character I
portray in the worlds they inhabit online, and acting like a total
buffoon in front of complete strangers that are all acting just as
ridiculous as myself.
But there's a problem with MMOGs, in that they do not reward such a
choice in playstyle. The closest any game comes is to write deep and
well-thought-out plotlines upon which we devout thespians of the
internet can build our own stories and adventures. But even those
aren't all that common.
In order to enjoy my time in any online realm, it eventually becomes
necessary to seek out others with similar proclivities. And you know
what makes that easier? A great big sign next to the server shard's
name that tells me that it is designated for roleplayers like me. I
shout a hardy Huzzah! when I see this tag.
But sadly, in a world of unified and border-free entertainment, no such
tags or signs exist. Instead, I'm forced to wander the same socially
desolate landscape that is populated by every elite min-maxer, every
griefer, every PvP pwn machine, and every socialite teenie-boppers that
pays to subscribe to the same title. I have no privacy, no retreat, and
no safe haven from their harsh stares and rolling eyes.
To put that in a more out-of-character context, the option of
separating your players into well-defined subtypes allows for
like-minded adventurers to seek out their own kind. It's an activity
that we will seek to perform on our own no matter what environment we
find ourselves in, but can be frustrating in a unified environment
populated by the entire player base. Frustrating enough, in fact, to
quit playing entirely.
Furthermore, developers of games that employ a shard-based environment
are free to experiment with different types of gameplay experiences.
SOE has been known to employ this freedom of adaptability to launch new
servers for both EverQuest
and EverQuest II
that possess unique sets of
rules not present on other servers. These new servers have also been
known to allow for content that is otherwise no longer available on
other servers due to updates and expansions, or because the need for
specific content no longer fits the story that has evolved there.
I'm reminded of another example from WoW the opening of the Gates of
Ahn'Quiraj. This massive server-wide event lasted several months,
during which time the entire player base of each server was united in
striving to unlock the new content available in the instance of
Ahn'Quiraj. Some did it for the sake of the new content, others for the
glory and fame of being the first server to achieve victory, but all
participated to some degree. But now, sadly, no server exists that
offers this content. A large portion of Blizzard's work that went into
developing the quests, plots and events that unfolded during these
months, has completely vanished. With the Cataclysm expansion, whole
sections of WoW's classic content is being thrown out the window in
the same manner, and no player will ever have the option to experience
this content again.
src="http://www.tentonhammer.com/image/view/92821">
Or will they? It is entirely within Blizzard's ability to
launch a new series of Classic WoW servers that contain only content
that existed prior to Cataclysm. Or another that has the Ahn'Quiraj
event still underway. I doubt they would do either, but because their
servers are segregated into separate worlds, the ability to pursue
either still exist as an option. Similar content on a unified server
would simply be gone forever.
Instancing Exception to the Rule
src="http://www.tentonhammer.com/image/view/92820">
If you want to look for a model that breaks the rules of this argument,
look no further than Guild
Wars which puts all of their players on the same server,
separated by instances with an ability to hotswap between them at will.
Star Trek Online
and Champions Online,
both developed by Cryptic
Studios, have employed this same instance-based server architecture
that combines a few of the best features of a shard-based world with
the
boundless unified model. Through the use of creative and widespread
instancing, all of these games have managed to segregate their player
base into
bite-sized chunks that are easily rendered by both server and client
software, while still allowing unrestricted access to these instances
by every single subscriber.
In execution, this compromise hasn't managed to address the fundamental
design concern of a fractured community. In fact, the heavily-instanced
environment has in some ways worsened the situation that would have
already been present had they gone with a shard-based server like most
MMOGs employ.
However, it's also been a boon to Cryptic in particular. Many of you
are probably aware that both STO and CO experienced a lot of hype and
build-up prior to their launch only to lose a significant percentage of
their overall population a few months after they reached the public.
This has actually become a common trend in many MMOGs, sadly.
style="font-style: italic;">Age of
Conan is another good example of this sudden drop-off, and
I could
probably rattle off a dozen more but I'll spare you.
In light of this sudden drop-off, a shard-based server architecture
would generally require the developer to shut down and/or consolidate
separate worlds into more populous conglomerate shards. While this move
just makes sense to employ, there's a significant downside to this
action in the form of public relations. The announcement that any game
is consolidating servers is generally seen as a mark of failure by both
players and fellow developers. In STO's case, they were able to
completely avoid any potential PR backlash by having no need to
announce the status of their hardware since their players were
unaffected by it.
Another modern title that makes extensive use of instancing is
Turbine's Dungeons
& Dragons Online. However, this title,
unlike the instanced STO and CO, employs separate server
shards as well. From a player perspective, I can see very little reason
to have both in place except for the physical aspects of attempting to
jam the entire population of DDO into the non-instanced social areas
that they've crafted. But even this could be overcome by instancing
them as well, which is a practice that's already utilized when a
particular social area comes under a heavy load. In this game's case,
the shards serve no function other than to fragment the player base
unnecessarily. As a part-time DDO'er (hey, it's free!) I hope to one
day see Turbine do away with this game's use of shards, but the time
for such a fundamental system design choice may have already passed.
Conclusion and Summary
Shards
Unified
- Fewer excessive cliques or content monopolies
- Unique mechanic options
- Customized, hand-built content
- Segregated player base
- Player-created content available for a portion of
players - Server selection can rift friendships
- Low population server closures cause PR backlash
- Eliminates the hassle of "which server are you on?"
- No artificial social boundaries
- No need to announce server closures or consolidations
- Griefing affects everyone
- Excessive crowds / server performance
- Higher need for procedurally generated content
This debate is largely a matter of what features or comforts a
particular player is willing to live with or without, in favor of
others. It's a personal choice that must be made by each player for
themselves, and one for which there is no clear winner.
---
Got
your own Pro or Con to add to either side of this argument? We'd love
to hear your thoughts on the debate that continues to rage in every
gamer community across the web. Leave a comment and join the
conversation!
