« Avatar Rights | Main | Maslow Short-Circuited »

Oct 13, 2003



In the post, Edward Castronova wrote:
"A real phase shift would be, say, a PvP world where one could identify long, stable periods of time when newbies were not ganked, interrupted by equally long, stable periods when newbies *were* ganked, with there being no obvious design changes at the switching points. Rather, the player base, as a group and for no design-related reason, simply changed gears in response to some internal historical dynamic."

I suspect that this is due to the places you are looking. Meridian 59 has had cycles like this happen continuously for as long as I've been working on the game. Groups dedicated to newbie survival or eradication rise and fall as the server progresses, and people remember old allies and enemies fondly. There seems to be a cycle in M59 where periods of relative peace are broken as random murderers (PKs) have grown in power and challenge the status quo. The result depends on how the "forces of good" fight back against this challenge. If the PKs win, most people are chased off for the short-term and the PKs get bored. Eventually the server builds back up if enough new people (and old people checking back) sign up again. If the PKs lose, then they usually leave (or go into hiding by rerolling a new character). The forces of good then either go back to the status quo a bit more battle trained, turn upon each other for continued sport, or get bored and leave. Each cycle is affected by the previous cycles, whether it's through warnings of veterans, rivalries between old enemies, or simply strategies learned from previous encounters.

Design changes tend to affect the strategies used (which can affect the cycle), but rarely affect the cycle directly. One of my goals as a designer of the world is to make it so that the strategies used result in the cycles becoming more extended; this gives more time for each side to get the upper hand over the other side before decending into defeat. I've also tried to design it so that an end to the cycle can arrive before the characters of one side or the other are rendered mostly useless due to death penalties; that is, I've tried to introduce the concept of war by attrition and victory by outlasting your enemies instead of destroying them. This allows characters a reason to stop fighting before the bitter end and to attempt reconciliation.

I suspect that this PK cycle is mostly a function of the world's size. Meridian 59 has much smaller populations per shard averaging about 100 or so currently with a realistic maximum of about 250 simultaneous logons (not yet reached since the game's relaunch). This smaller community size leads to a different feeling than the larger games provide. I imagine that the larger worlds have a lot more inertia, where a small group of a dozen or so people simply cannot disrupt the status quo as easily in a server of several thousand compared to a server of a several hundred. In the offline world, it's easier to motivate people to care about the world around them (even if you get them to "care" in destructive ways such as genocide), but it's harder to movivate people about a commercial world they have to pay to enjoy.

There's my wild-ass theories. Enjoy.


Actually Brian, that makes a lot of sense. I think there is a micro/macro issue in these worlds. LambdaMOO, Meridian 59, and ATITD all have a different feel from the behemoths. They also are most commonly referenced when one speaks of 'exciting new things happening in virtual worlds' - democratic movements, player-made currencies, historical cycles. Small groups of intelligent, net-savvy, energetic people. The communities of large worlds don't seem to exhibit those features.


I think part of the "different feel from the behemoths" may be due to inherent design differences. The behemoths are specifically designed to appeal to a (not the) mass market while the other games seem more targetted.

As such, the behemoths are inherently more game-like, more arcade-like, than the niche players, which seem more world-like, sim-like.

I can't speak for LambdaMOO or M59, but EQ, DAoC, The Realm(originally at least), AC, and nearly every other MMOG I've tried/beta'd have, at their heart, been vast (arcade)games, while the niche MMOGs I've tried like ATITD, There, and a couple of others, have been almost like grad-school research projects on emergence, society, and simulation. They've been truly virtual worlds.

I think the distinction is an important one, though I'm not sure how to lable the two sides. It does seem, however, that the arcade-style games with a pseudo/meta static world overlay are much more financially successful than the virtual worlds, the MMO Sims(not to be confused with "The Sims" or with "simulation games" like BF1942.

Does the same dichotomy exist in M(UD|USH|OO)s? I tend to think it does, but am not experienced enough to actually make a definitive statement about "success" as measured by "subscribers" or regular, serial players.

Getting back to the "history in games" bit at the very beginning, I'm a NASCAR fan, but don't hold that against me. Having recently heard the PS2 might be available for $99 soon, I was reading a review of NASCAR Thunder 2004, which apparently has a history mechanism for both the single season and career(20 year) modes of play. Your actions on track, like blocking another driver, wrecking another driver, and oppositely, playing nice with another driver, earn you prestige points as well as alliance/rivalry points, which affect how the other drivers, in turn, treat you on the track throughout your season/career. If you know anything about NASCAR, this (seemingly) is a huge feature for real fans and hardcore players.

You actually have a history in the game that affects the game, and it's not prescripted.

I know that a (primarily single-player) racing game doesn't quite fit the mold of a Virtual World, but features like this could start to blur the line if adopted across more genres.

Sorry to have rambled, but you guys always get my neural juices flowing.


Perhaps what's missing is journalism embedded within the game world.

History is alive to the participants, especially those highly networked players who are at the center of the gossip chains. Outside boards provide more history, but it's disjointed and generally guild-focused. (At least in my most-familiar EQ.)

Support the telling of stories and consideration of events in their context, and maybe the history will start to evolve, based on a common understanding of events.


The telling of stories is, IMO, supported in any "game" with a chat engine.

It's the "permanence of change" that's lacking. I can understand the gods of the planes and dragons respawning, but just how many lives DOES D'Vinn have, really?

Telling the story of how you barely scraped by in a harrowing battle with all the evil denizens of Crushbone has very little value when every other character above level 20 has done it a gazillion times too.

Or the interception and slaying of the orc messenger in the Butcherblock mountains - it has zero significance because every other player has done it and it doesn't change anything.

It's an arcade game, albeit a persistent one where your character grows more powerful(level-up = power-up?), but still an arcade game that replays nearly the same for everyone, every single time they "start over".


The game where I found the most history is/was Nexon's Dark Ages. I think the point about journalism in game, or in the extended out-of-game but related media, is on target. Dark Ages promotes the literary efforts of players as part of their character development. As a result, eent are records, built upon, and referred to in other, newer, works.

But size, and cohension, also play a part. The modern, sharded nature, of MMOGs does not promote a shared history. Every "galaxy" or "server" should have its own history, but the parallel nature of them tends to glob it all together and thus dilute it, or normalize it, to where there is no history, no sequence of events tied together by causality. There is little common causality beyond that of the manipulated-by-the-developers environment.

Earlier, smaller games had more history. Thinking back to Island of Kesmai, especially, I definitely have a historyical sense of that game. I have far less of UO and EQ, though there is a little for DAoC, once I started paticipating in PvP in the frontier lands. In DAoC it is diluted by the shard effect though, most definitely. Still, the taking and holding of real estate following big battles, sneaky maneuvers, long grueling repairs of keep gates, these are the sorts of events that COULD become history, if they weren't diminished in importance in that game's community by the fact on 10 other servers something else happened.

ATITD suffers from a bit too much scripting in this regard. Yes, the efforts of players to some degree control at least the pace of the script. But what unfolds if largely pre-determined. Not all of it. Players and groups do come up with their own ideas. I took part in the development of the first widely-accepted currency there, and that was a history-naking sort of event. But the lack of combat in ATITD also decreases the sense of history. If we look back over the study of history, a lot of it concerns battle and struggles for power. One half of that combination is missing in ATITD.

I think history inside games does a lot to enrich them... for certain audiences. I'm amazed and amused to find a lot of real Star Wars fans playing SWG. They actually know that universe. That's a new phenomenon to me. In games developed whole cloth with thinnish backstories, no one knows much about their world. In those that encourage historical-like movement, more learn all that is available.

I like SWG's idea of having the monthly 'acts' push and pull events somewhat, but that game suffers from the sharded personality. It's hard to accept the rebels have gained power if in your galaxy the Imps outnumber them 5:1 and kicked them all over the known worlds... even if in every other galaxy the opposite was true.

I can live with it, but it doesn't do much to promote a real sense of history.


"Perhaps what's missing is journalism embedded within the game world."

That's the first thing i thought of, Ian. For an example of embedded journalism within a virtual world, please check out Wagner James Au's "New World Notes" over at http://secondlife.com/notes/

There's some real history in there -- specifically the multi-part series on the the "War of the Jessie Wall" which starts on the bottom of this page: http://secondlife.com/notes/2003_07_07_archive.php


When I was observing class balance metrics in Camelot, I saw many "phase shifts" and suprising emergent results. For example, at one point I notice an odd jump in the RvR performance of a class, when I looked deeper this jump was actually confined to a single server, where suddenly the performance of this class had jumped by a factor of 4. This had been preceded by a slow *decrease* in the number of this class on that server.

I noticed another server where the number of this class was declining slowly, and was approaching the same proportion as that of the one that just jumped. Two weeks later, the RP performance of that class on this server also jumped. Meanwhile, the proportion of the class on the original server had started to climb, 2 weeks later (4 after the initial jump in RP performance) the RP performance dropped as quickly as it had risen. In 3 weeks, the second server dropped. Then the population of the class began to decline.

This cycle by then was showing on many different servers, all with pretty much the same characteristics. It continued for as long as I was observing, a 4-6 week repeating cycle. The cause appeared to be the damage type of the class, which was unique to that class. As they decreased in population, people stopped defending against that damage type, discarding items that offered resistance to it and forgoing spells for that resistance. At some point, some critical value was passed and the few remaining active members of that class earned huge amounts of RP. This high reward attracted and retained people in that class, causing a population rise until some critical threshold was passed and players again defended against the damage type.

Another example: On most servers, Emain Macha was the "deathmatch zone", the place where most non-objective oriented RvR activity took place. There were two exceptions, for both the zone that was used instead was Odin's Gate. In both cases, Midgard was the smallest realm on the server (on all other servers, Hibernia was the small realm). One of these transitioned to Hibernia being the smallest while I was observing, a few weeks later the "deathmatch zone" changed to Emain Macha.

So much for emergence, now for history: I think history has been rare and limited because the impact of the players on the world has been rare and limited. If the players can't make anything lasting in effect, there's no point in historical records of what they have done.



"This cycle by then was showing on many different servers, all with pretty much the same characteristics. It continued for as long as I was observing, a 4-6 week repeating cycle. The cause appeared to be the damage type of the class..."

In agricultural economics, this pattern is called the 'cobweb dynamics' (referring to the weblike pattern that price/quantity points trace out over time). I guess you would call it a stable irregularity: high prices elicit lots of planting; lots of planting yields high quantities and low prices; low prices elicit minimal planting; less planting yields low quantities and high prices, and the cycle starts again. The pattern happens because people use current prices (instead of current planting) to forecast future prices.

Anyway, Dave's fascinating observation is one of the reasons I wish academics could get their hooks into the datastream somewhere, preferably in our own world. As an ordinary player, it is really hard to see these things. But if we were the wizards behind the world, we could not only see the patterns, we could conduct little experiments on them.


I am curious what sort of data you might make use of, and what sort of experiments you would be interested in running.

In my current position at There, inc. I have access to quite a bit of data and could always make a proposal to grant limited access if something seemed useful. I am especially interested in enabling those in higher academia the ability to study aggregate economic data.

I would also be open to running economic experiments or tests if there would be something we could find out that would help There make better economic policy decisions in the future.

Of course I can't make any promises by myself, but I would certainly take any serious proposal to the right people internally.

aka, There_Economy


Thanks Scott, that "New World Notes" reference was very much like what I was thinking of. Stories about how people interacted. Some will have significance, many won't. But over time (and with the possible nudging of a historian) the stories turn into a history.

Duck I was thinking less of PVE stories (though there might be some place for them, espcially firsts), but more about people interactions. There are of course lots of examples of poor (or boring) stories, but presumably a journalist would tease out interesting ones so he might sustain his marketability, if for no other reason...

I have to say I'm unconvinced of the utility of world-changing events, if we're talking about changing the environment. By any definition I can think of, only a very few players would experience "making" the change, but most/all would experience the "changed" world. This essentially means that for all but that few there is no difference between a developer triggered and player triggered change.


First point: I don't think you have been looking hard enough if you are lacking for history. Browse through the arhcives of UO Stratics or UO Vault if you want to find some. :)

Second point: I can think of many phase-shift events in UO which were not induced by the designers. It may be that because of the original, anarchic nature of UO, it lent itself to this sort of thing. Some of the major historical events are things that I bet resonate with some of the list members here:

- the player tavern movement and the war over Kazola's tavern, which split the playerbase

- the rise of the Shadowclan (this one arcs across multiple games, even) and other non-game sanctioned "races"

- the player city movement beginning in Yew with Enshu Ponfar's leadership, and moving on to the founding of Oasis and the player-run scheduled event (fight nights, etc)

- the last ditch defense against PKing with the Carebear Manifesto

All of these were player reactions to a status quo, not induced by designer actions, and they did lead to "phase shifts." The player tavern and city movements have arguably led to entire new games (by other developers...)


"There's some real history in there -- specifically the multi-part series on the the "War of the Jessie Wall" which starts on the bottom of this page: http://secondlife.com/notes/2003_07_07_archive.php"

I've just been a lurker here since finding this site, but after reading the linked article above I thought I'd make my first comment post... :-)

I had no idea that WWIIOL'ers had crossed over in such numbers... I know a couple of people that had mentioned Second Life, but I hadn't stuck my nose in to see. Interesting!

I have a good bit of experience with MMORPG's (UO Beta, UO, EQ Beta, EQ...) and tend to agree that the sense of 'history' in-game with those systems is diluted to the point of nullity. The games themselves attract entirely the wrong kind of player to make it otherwise. Games that are populated by the 'older crowd' tend to develop that sense of history, IMO; it sounds like SL has that, I'll have to try it out.

WWIIOL has that demographic, but the game world resets too often for any real sense of history to the environment, other than a running count of how many times we've run the Allies out of France. Maybe that will change if the theatre increases significantly in the future.

Hopefully that wasn't rambling too much! I love reading the articles here, great stuff.

Phil Comeau
WWIIOL Veteran, 2 yrs and counting


Speaking as a player, one thing I miss was being able to track the course of events within the game. There are a lot of historical reports and fan fiction on separate sites, but an in-game library system, or news archive, might have an interesting effect on the sense of shared history.

As far as letting the players change the world, you could produce the appearance of change to at least some some degree without actually changing anything. Ever read Roger Zelazny? In one of his "Amber" books he described a Tavern where the name of the tavern reflects the previous owner. Bloody Bill's is actually run by his borther Sam, and the name reflects Bill's ultimate fate. Things like that.

In other words, if I kill a mob it isn't necessary to remove that mob. Just rename it. Replace a dead shopkeeper with a slightly different looking NPC that carries a different name. If asked, the NPC could have a small script describing how his brother/cousin/etc. was brutally murdered and robbed by soandso and he inherited the place.

Little touches like that. Orc messengers could get killed a lot, but do they all have to carry identical messages? How hard would it be to plug in a brief, say one or two paragraph report that desribes what actually happened in Crushbone yesterday. If the players stage a major raid, have the Orc messenger's pouch contain something like: "Dreaded Lord, we have suffered major depradation of late by the cursed Wood Elves and require immediate reinforcement." Stuff like that.

I don't know if any of this is clear, but I don't think you would necessarily need to permanently kill off a God and change the climate in order to give a sense of connection to past events.

The comments to this entry are closed.