« Hamlet who? | Main | Mimicry, Melting Faces, and The Proteus Effect »

Mar 09, 2006



* I'm so sorry. delete my previous comment.

Interesting and nice stuff. Actually, I’ve been considering similar idea for this one. In Korea and China, there are many illegal ‘servers’ run by MMORPG lovers. This is called as “free servers” in Korea. Many of popular MMORPG have its free server clones under the ground. Recently, the evolution of network environment in Korea made operating free server so easy. Unlike most of xDSL, upload speed is increasingly improved with many types of FTTH service.

Anyway, some MMORPG players said that they felt much more comfortable in this illegal and secret playground. The speed of level-up is much higher than original. There is few agonizing factors like PK(players killing) or other kind of virtual harassment. As you pointed out, there is lesser pressure for time spending to elicit fun. Also, many players who enjoyed in this rogue place told that free server made them an easy to get juicy parts of MMORPG. Players in a free server could hunt down legendary monsters in original game easily!

I think this would be remodeled as commercial business model. Actually, some operators of illegal server have already made a considerable fortune. I’m so sorry not to tell you more for their safety, but this underground side of MMO games is one of interesting point to observe in Korean gaming.


I definately think there is an important place for short-term commitment games.

Skotos has experience with three: "Galactic Emperor: Succession" was a week long role-playing game, which starts with the old Galactic Emperor being assassinated, and a new Emperor selected by the end. Unfortunately, it failed for a variety of reasons (as described in a post-mortem at Galactic Empires, Part One: Failing at Succession ), but in my opinion the most important of which is that there were a number of problems associated with it being a week long.

We then created a more strategic and less RP game Galactic Emperor: Hegemony, which is a three-week long game for 12 players. It was initially quite a bit more successful, however, after 3+ years it has lost a some of its charm and is not nearly as active. The problem is that it has been expensive to bring in new customers when old customers stop playing.

Most recently we created what we call a stage, the first of which was Lovecraft Country: Tomb of the Unknown God. It was intended as a two-day (typically weekend) heavy roleplay game, however, more recently has been shortened to an intense 5-6 hours. It is limited to about 30 players, but runs with as few as a dozen. Really this is much more like an online version of a LARP -- you are dropped into the body of a character where you can do things like "recall Achmed" and it will tell you everything special you remember about achmed. To date it has been run a couple of dozen times. Alone I'd not say it was successful, however, it has a tie-in with our Lovecraft Country: Arkham After Dark persistent RPG, and with our horror comic book Return to Arkham, so it has worked out reasonbly well for us.

I personally have not been able to play many online games lately, even my own at Skotos. They require far too much dedicated time. But I don't want casual games either. What I want is shorter term but intense games. I've been getting the fix I need from periodic weekend-long LARPs and game conventions, but I'd love some online alternatives.

-- Christopher Allen
Skotos Tech Inc.


Can't wait to get the data.

I assert that reset is a part of society's psyche. In the arena of fun and entertainment, the common periodicity is approximately nine months with a three month hiatus. From a business perspective, mapping development and play cycle to this harmonics should yield great results.

Reset is a part of society’s psyche because at an early age we learn that if we fail, we can try again. As we become more aware of competition, we learn that if we lose, we can also try again. As gamers, we also quickly learn that we can start the game over or reload our last saved game state.

Reset is seen in many areas of fun and entertainment. In sports, television, and other forms of entertainment the term Season is used. You have football seasons, tournament seasons, and seasons of serial entertainment.

The common periodicity is about nine months. Some are longer. Some are shorter. In general, they are around nine months.

The more important aspect of reset and seasons is that there is usually a set period of hiatus. It is a break period which can be used in many ways:
1. allow for a break away
2. develop suspense and anticipation between seasons
3. develop renewed interest after break
4. as simple as WOW’s rest state

From a business perspective, there are lots of benefits to map online game development and play cycle to this harmonics. In an online sport game, players can associate closely with the familiarity of offline sports, with the built-in expectation that there will be a break and a new season to start anew. In an online narrative game, players can associate closely with the familiarity of television cycle, with the built-in expectation that there will be a cliffhanger, an anticipatory break, and a renewed interest in what comes next, whether a reset or a continuance, but with unexpected twists.

As online games become more popular, this form of entertainment will start completing with other forms of mainstream entertainment. Therefore, a harmonization of cycles and subsequent counter-programming and cycles is highly probable. The research will shed more light on this, and at a lower cost than a business attempting to do the same in the marketplace.


Richard Bartle >Imagine it wasn't an academic textual world, though, but was a commercial graphical world. Would people still play it, or something like it? <

Based on my experience with “A Tale in the Desert”, I would say yes. There, the “purge” cycle for a “Tale” was intended to be six months. The first two Tales have taken about 18 months each, so it missed by a long way. We do get some people saying “its all going to go away, what’s the point?”, but it’s a positive for some people. You do get a bunch of people that drop out at the end of a Tale, it’s a convenient point to do so instead of just fading away. But you also get a bunch of people jump in at the beginning of a Tale because it is somewhat of an even playing field.

Though the end of the world is scripted by the developers, the twists and turns of getting there do create a real history for each Tale. As a story oriented player, that is a big benefit to me. Eighteen months seems to long a cycle though. Frank may be on to something with the nine month idea.


I also have the impression that this style of gameplay could work, especially if each incarnation offers a new story set within the same structure.

This is pretty much the formula for episodic TV shows. The rules are the same, but the story changes -- the comfortable familiarity of understood rules are the framework through which novel entertainment is provided. Millions of people seem to like it....

Something a friend and I have been discussing is the possibility of an online game structured something like "Survivor". For each incarnation, we'd invite people to participate in a story (colonists on a transport to Mars that gets damaged must work together to survive, for example), then tap several applicants to be the active participants for that round. For each chapter in the story, we'd turn on the server for a few hours nightly or weekly, and the participants would play through that scenario.

Meanwhile, everyone else could watch as the participants interact in real-time to solve problems. Observers could vote on challenges they'd like to see integrated into the storyline, either in a discussion forum or perhaps online while the game is running.

This would combine the familiar episodic-storytelling nature of TV with the participatory power offered by Web access. It would also seem to achieve many of the goals Richard lists in the initial post of this thread.

So has something like this been tried already?



The old Kesmai Corporation Games, "MegaWars III" on CompuServe and "Stellar Emperor" on GENIE and later GameStorm, all were reset nominally every month.

On CompuServe and GENIE, only a single Game-instance (called a "War") operated at a time, and they would occasionally throw in a one-week or two-week War -- often at players' request -- either to spice things up a little or to synchronize the start of a new War with a long holiday weekend.

In MW3, which was Before Graphics, the positions of the stars in the Galaxy remained invariant from War to War; only the resources available at each star changed. This was not publicized, and there was advantage to be gained from discovering it: the first stage of every War was a "land rush" as players hunted for valuable planets on which to develop bases, and knowing how the stars were layed out let a team organize its scouting efficiently, minimizing travel time.

When player Allnone of the "Dorsai" team not only made a MW3 map but *published* it on a public Forum, Kesmai said they were not surprised that someone had made a map, but were amazed that that someone had published it, rather than keeping it secret to gain advantage.

On GameStorm, they ran several instances in parallel, and offered a variety of time scales. I don't remember whether the star positions were stable.

More recently, "Time of Defiance" from Nicely Crafted Entertainment has operated with periodic resets and multiple instances operating in parallel.


I've had many players comment that they only tried Everquest 2 because the "new servers" aren't corrupted by late-game economies. There's a growing perception by gamers that, if you don't get in close to the launch of a game, or at least the introduction of a new server, you can't expect to really compete- the economic barrier to entry is too high.

Purges would eliminate that barrier. Even if you started late, you can see the time as a "warming up" and learning the game, ready to start running at the next purge.

Anyone else see the "economy barrier" affecting late-in-game adoption?


A Tale in the Desert is the classic example of this idea, since it restarts a new "tale" periodically. From a commercial standpoint "Tale" hasn't been terribly successful. However, the game had so many innovative features it's impossible to say which ones affected its market success. The same could be said for the many other examples quoted, most of which are quite obscure compared to the "household words" of MMOGs such as UO, EQ, Lineage, DAoC and WoW.

Based on patterns of business success (and failure), I believe that a successful MMOG has certain important characteristics. Among these is persistence of both the world and the characters within the world. While short-term gaming pleasure and satisfaction occurs when you overcome certain obstacles, or achieve certain rewards, the classic "road to riches" in the MMOG business involves long-term satisfaction: your customers keep coming back month after month.

Long term player achievement and satisfaction only occurs if the players have something they can show or demonstrate as an accomplishment. Once it was your level 50 (or 60) character(s) in EQ or DAoC. In Lineage it's owning castle X for Y period of time, or having a certain weapon or armor set (which implies significant character level and accomplishment prerequisites). In WoW it's all these things. Needless to say, if you reset regularly you're at best crippling and more often destroying the mechanics of long-term accomplishment.

Most game designers recognize that games MUST provide periodic rewards to the player. The more frequent and meaningful the reward, the better! The worst possible thing you can do is to take away those rewards, especially if they were visible to other players. Blizzard is past masters at the "frequent and meaningful reward" technique - just look at Diablo or WoW. Even their RTS games slowly ramp up, with a steady stream of new stuff in each new chapter of the game.

It is still possible to create virtual worlds for non-gaming purposes, and those can reset as often as needed. It's also possible to create games that are entirely short-term affairs, like the solitaire player who says "Let me reshuffle and start over, then I'll do better." Nevertheless, if someone gave me $20 million and wanted the most successful MMOG possible for that investment, I'd build something with long-term accomplishment to keep my customers paying for more as long as possible.


Can players save stuff on their own? How hard is it for a player to restore favorite places/items/bios? Or don't they even try to do that?


Ahhh the beauty of Dungeons and Dragons because afterall that is exactly what you are hitting on. Some campaigns could run for a year or more, but most often I've found campaigns to last 6 months at most and then it is all restarted a new. And most D&D groups meet 1-3 times a week. Not always did you wipe everything. Sometimes you keep characters and sometimes you keep just the world. Stories can change and characters can change, but some sort of "reset" is needed to keep the game fresh for everyone.

This was best translated to PC by Neverwinter Nights. It was the truest to form D&D translation to video game ever and Neverwinter Nights 2 is picking up right where it left off. I can't even count the # of downloaded modules my D&D group played through NWN. And as we played more we actually enjoyed the shorter 1-2 night modules the best because we always had something new!

Another viable example is ladders in Diablo and a myriad of FPS games. They reset every so often and there is people flocking to start over again. Its a suggestion that I've had for Battlefield 2 forever because BF2 stats keep adding and adding and adding giving the newer or less hardcore players no chance of ever showing their worth.

BF2 stats also show the weakness of purging. The hardcore are going to "win" every round. So what is important is making the majority (aka the losers of the round) have some impact. In BF2 I don't want to be #1 killer, but whats wrong with having a reseting ladder where best "aim" is rewarded after 1,000 sniper rounds fired? Also the badges and ribbons you can gain in BF2. Even if the ladder reset there is no reason these couldn't be awarded. Then each reset could result in more awards to be handed out.

Problem is that its just easier for developers to create that static growing system. WoW tried to go with a non static honor system and they FAILED. Don't even need to discuss that here because they fell right into the trap of purging... rewarding those that play the most in the given time.

Lets even look at profesional sports. The idea of seasons comes into mind. The NFL plays 16 games a year, the playoffs, the superbowl, and then its on to next season. Works great for them, but I doubt you would want to incoroprate an "off season" in an MMORPG!

There is examples all around, but MMORPG developers don't tend to follow outside examples. They look at the narrow of "what has come before" and then see what they can fix that the last guy did wrong. Thats how you wind up with sequelitis.


I used to play a text based MUD that had a reset every 6 hours. The only thing you could take through a reset was your character, XP and GP. All your gear, if you hadn't sold it yet, was wiped. After the reset there was a boot rush to re-equip. This starts with shop bought gear which you then replace with drops.

The nice thing about this system is knowledge rules. Experienced players know exactly where to go to get the best items for their level. A 20th level character often takes 3 hours to fully equip before tackling a quest. Completing a major quest is always a run against the clock. Sessions are 6 hours which aids personal time planning. You know when it starts and when it finishes.

While this is a free MUD, not a pay for play, it has been running since the late 80s and as far as I know is still going strong.


Arnold Hendrick>Long term player achievement and satisfaction only occurs if the players have something they can show or demonstrate as an accomplishment.

I agree, but that "something" doesn't have to be all that great. An entry in a "hall of fame" can be enough; players who want to cut back on their commitments can demonstrate that they've been there, done that, so although they may be playing a low-powered character right now, they don't have to take any gloating from high-powered characters. They've already demonstrated competence, commitment and understanding; now, they're just playing for fun.

>Needless to say, if you reset regularly you're at best crippling and more often destroying the mechanics of long-term accomplishment.

The current mechanics, yes, but we don't have to use that system; indeed, one benefit of having a virtual world with a fixed lifespan is that we may be able to escape from it.

>The worst possible thing you can do is to take away those rewards, especially if they were visible to other players.

But if you take them away and replace them with something better, or more permanent? And if players know when they start that the in-world rewards won't survive the purge anyway?

Most online and face-to-face games have a limited duration. It may be that there's something special intrinsic to virtual worlds that would be compromised by such a mechanism, but I don't believe there is; the problem is cultural (what players expect) rather than systemic (what is required for them to have fun). As Zyn has pointed out, older textual worlds got and get along just fine using a far harsher reset regime. OK, so they don't have the player numbers of a WoW or even a TSO, so it may be we're talking niche here; on the other hand, it may offer a way out of the current DikuMUD paradigm we're stuck in.

>Nevertheless, if someone gave me $20 million and wanted the most successful MMOG possible for that investment, I'd build something with long-term accomplishment to keep my customers paying for more as long as possible.

So would I, but I wouldn't necessarily make that accomplishment in world; meta-game accomplishment could also work. I also think it's much kinder to set players free (allowing them to reinvent themselves) rather than imprison them (defining their progress only in terms of one character), but I can see why people looking at profit margins may be alarmed by that!



I think there should be a distinct discussion between developer-set purges and player-enabled purges. The former is set by the developer which then set the expectation of players and subsequent playstyle.

I think there should be a distince discussion about different type of purges. The word purge just conjures an absolute action. There could be reset of characters, items, worlds, etc. It should not be a forbidden tool in a designer's toolkit. It shouldn't even be shunned.

For standard MMORPGs there is an expectation of persistence. Well, most worlds and characters have a set "shelf life". The default purge for game world is when the operation can not be sustained (with no reset unless someone resurrects it). For characters, the default purge is when the player is bored of it. The character may keep it on the shelf as an alt, but the primary "utiltiy" has decayed to a point that it might as well be "purged." So going back to my primary suggestion, what's the different between a developer-set purge and a player-enabled purge?


As for the commercial issue of having a break for an operator, it can be utilize as an asset rather than a liability.

One possible implementation is stagger server resets. So instead of playing continuously on one server for 8 years, you could play on two servers for 8 years. People already play on multiple servers, so designing a break in-between in a persistent game world creates a "gated" staircase progression rather than an ungated smooth progression. Gated progression is obviously more controlled and allow content creators time to create new content which will then be shared by all participants more fairly and equally.



Again, my old game FaitH...

The game is played on 2 servers, a battle server and an rpg server. The game is played in 'eras'. The 'battle server' eras run approximately two and a half months and the 'rpg' eras run about 3 months. I wont go into details but will note that there is a persistant community of players that play and have played in some cases for several years, myself included.

What keeps players coming back is the community and the ongoing competition. Player run irc channels abound, forums and a responsive developer who is close to the community has been a key ingredient.


magicback>I think there should be a distince discussion about different type of purges. The word purge just conjures an absolute action.

In the Norseheim example, that's exactly what it is: the entire shard is wiped. I don't know about accounts, but everything from the character level up disappears.

>There could be reset of characters, items, worlds, etc.

Yes indeed. It was clearer in the olde texte world days, but "what survives a reboot" is a good indicator of the "personality" of a virtual world. Some worlds only save characters, some add objects characters carry, others add objects by class or by location, then there are some that store all physical facts (object properties and locations) until finally there are those that save everything, including any additional functionality added on-the-fly while the world was running.

I suppose "time between reboots" would also be a factor. A virtual world that resets every few hours has a different dynamic to one that resets once a week or once a month, or never.

>It should not be a forbidden tool in a designer's toolkit.

I agree, although of course it's the players who have to be convinced.



I would not be surprised if players replaced risk-aversion with risk-loving behavior as the reboot time approached; I guess we'll have to wait for the study to know for sure.

While it makes sense for there to be MMOGs with finite, preset time between reboots, I am not sure these could still be called virtual worlds. One aspect of persistence needed by a virtual world is that players do not know specifically when the world will end. Is it a good gaming idea? Certainly. But is it a virtual world? I tend to doubt it.


Reset is a part of society’s psyche because at an early age we learn that if we fail, we can try again. As we become more aware of competition, we learn that if we lose, we can also try again. As gamers, we also quickly learn that we can start the game over or reload our last saved game state.

I think the challenge for the developer is not developing the reset, but the rhythm and context of the reset. While I appreciate that some players may enjoy a rush against an arbitrary death clock, and some seem to like the so-called Hardcore mode of Diablo II (aka perma-death), I can't see that arbitrary server resets have much of a future in persistent worlds--people just like their stuff too much.

I know this concept would convince me to not start playing a game, probably... unless it was well designed into the entire structure of the game (a la the Tale in the Desert examples given above). But to think of Norath or Azeroth suddenly wiping and having to start again as a level 1 druid in whatever newbie zone just sounds awful. I'd rather game creators use their energy to make these games less grindy--in my opinion WoW is a great leap over Everquest for that reason alone.


I think a regular purge is a really good and interesting idea. As has already been said, it opens up new potential for long, episodic narratives and the potential for players to experiment with a larger and evolving variety of character. For example, I might be interested in honing the best wizard possible. To achieve that in a limited span of time requires experimentation and skill rather than simply time and money. But I can also see that some people would be turned off by being forced to start from scratch after every purge, losing their character and any attachment they had to the playing world.

Here's an idea that addresses both sides: The world itself is persistent, and the character can be persistent or not as the player chooses. I foresee an online world containing many character paths which have a standard beginning but also several possible endings; these endings would be built into the quest system. Once a character completes several important and difficult quests to become the sovereign of a certain area, for example, the player is rewarded, and the character is removed, giving the player a chance to start anew with a fresh character and a new outlook on the game. Games would be continually revitalized by the steady stream of experienced players who come back time and again to explore new gaming avenues in the same world.

How can the player be rewarded without a persistent character? In different ways: each subsequent character can be given a permanent advantage, each subsequent character could be given a cosmetic item, a badge or something, or, my favourite, the persistent world could somehow recognize the accomplishments of the character, through a monument that records their name, through a closet-drama legend (consider a very difficult character path that could be completed only once!), etc.

Players need not follow a character path through to the end, but might take up a more simple lifestyle of hunting or farming, etc., but the option to reach that ultimate accomplishment is always there.


It does not at all surprise me that "refresh" would work in a game like that. Surely I am not the only person who has opted to watch an episode of Futurama I've already seen instead of a movie I haven't because it's less of a time commitment.

Two anecdotes in support of the idea of "refresh":

1) In college, one of the ways I burned copious free time was as one of a dozen or so players of the beta version of a text-based sci-fi colonization and trade game (MegaWars III sounds very similar, and in fact the friend who introduced me to the game described it at the time as "just like a game on CompuServe"). As the developer worked out kinks and tried new rules and features, s/he would reset the game every few weeks. Every time the game was reset, we had a new galaxy and a new starting position and a new map. Everyone enjoyed this in part because (a) we knew about the resets from the get-go, making them more "fair" and (b) because if we bombed out in a given iteration we knew we could learn from that experience and try for a better outcome the next time.

Did people want continuity, though? They did. Among the dozen or so of us playing, a group-written meta-fiction sprang up on the in-game discussion boards (it felt very BBS-y, in terms of interface). Our "races" took on histories, personalities, creation and apocalypse myths which explained why the world would end and why it began again. By the end, we had complex systems of rivalry and friendship and a player-created storyline that explained that each of the competing/cooperating civilizations had ways - some magical, some technological, some psychic - of perceiving the universe's impending death and an ability to open portals to new, parallel universes to colonize. Thus, reset happened on schedule but our identities to one another and our collective fiction could survive.

2) For several years I played on and helped run a MUX using the system and rules of a popular pen-and-paper RPG. After that game's natural end, we pondered what to do next. No one wanted to commit to another five years of constant game stewardship, but we still wanted to play. Eventually someone had the idea to run a text-based game for one year. We would warn people from the start that the game was going to uppercase e End in one year. The players were to make characters knowing that this was not going to be where they played a character from newbie to uber but were going to explore one year of that character's life. We would award no XP because the sorts of stories we were going to run weren't going to lend themselves to lots of "leveling up," it was going to be character- and plot-driven rather than combat- or loot-driven. No one should come into it thinking they would go from being decked out in rags and rusty daggers and in a year's time be driving a Mercedes or wielding an Epic Anything. We were afraid that the player base from the old game, who were our first targeted potential participants and who had years of playing characters that grew and changed dramatically in terms of power and abilities, would be so uninterested that it would never start. Instead, their response was overwhelmingly positive. The ability to know that if their real lives were different a year from now and they could no longer devote time to the game was very liberating. Knowing that there would be no room for jockeying, or pressure to keep up with the virtual Joneses, was very liberating. I was shocked at how into it they were, and how quickly, and how they instantly *got* the pros and the cons of that approach. That the game never happened had nothing to do with the game's design or lack of players and everything to do with our real lives interfering before it could even start. In fact, we had so many interested players we were afraid we would be overwhelmed.


Sam> But I can also see that some people would be turned off by being forced to start from scratch after every purge, losing their character and any attachment they had to the playing world. <

Starting the world over, doesn’t imply losing your character history or attachment to the world. At least, not in an “A Tale in the Desert” type of world. In that world, you restart (if you wish) as a descendent of your previous character, with the same name. So all your history in the previous age comes with the character. The Player also retains all the world knowledge and social connections from the previous Age. A Tale in the Desert has a minimal amount of character stats though, power is in player knowledge, social connections and access to production resources. Purging would be tougher in a EQ world where much of the power is in accumulated character stats.


Norman Maynard>I would not be surprised if players replaced risk-aversion with risk-loving behavior as the reboot time approached

This happens when virtual worlds are closing for commercial or practical reasons, but those aren't scheduled and the players are generally depressed by the idea. It's a "go out with a bang" kind of thing.

For a world that was always going to close down on a given date, I'd expect the process to begin sooner and be less of a wake and more of an excuse for mayhem. This is what happens when we have a "no consequences" event in MUD2, where we snapshot the database and then give people a day or so to do whatever they like in the knowledge that the saved version will be reinstituted at the end. Risk-taking is much more frequent when you don't care if the risk doesn't pay off.



There are two time-cycles we are talking about:
1) a reset cycle set by the designer
2) a decay cycle set by the player community

A purge is an absolute form of the former and the "death" of a virtual world is a typical example of the latter, resulting in purge but no recent.

There may be issues when the expectation of the two cycles are not in sinc. For example, players like to keep the world going, but the operator can't keep it going. Another example, is operator want to keep it going, but the players do not want to.

In a time-based achievement game, players may not want to lose the time investment and rather monetizes it at auction houses. Nevertheless, for this type of game you can still make the process fun. Example, modern casio gambling odds are ~51% in favor of the house. This means that it will take a while for the house to win. In the meantime, players are having lots of fun in the process of losing.

That is why this experiment is so interesting to me. It's like studing a virtual economic game (with greater dynamics than most current economic games) or studying people going to vegas to lose $10,000 (this should be a reality show!!!).



Oh let me clarify. The objective of the idea is to study the process not the end results. It would be good to see hard data to confirm what I found as common/logical sense.


I'm not sure why all servers necessarily need the same policy. I could imagine a single resetting WoW server being popular and considered a "contest" server of sorts where people would try to achieve as much as they could before the reset. Such a server could even go on ultra-short time -- a reset every day, for instance.


I'm a member of a MUD that actually took this idea to extremes.

It's called Genocide and its a PK (player killing) MUD. As with most MUDs, the virtual environment has been created by successful players who have received "developer" characters for their longevity and dedication to the game.

In that respect, its no different from a thousand other MUDs.

What makes Genocide unique, however, is every war resets every 10-30 minutes. Each war, players start naked and must gather ("bot") as many items as possible, sell them for coins to buy heals and weapons. Then we kill each other. It's pretty frenetic.

I have to say, though... I can't play other MUDs anymore. They bore the hell out of me.

The comments to this entry are closed.