An interesting problem that crops up occassionally in Second Life is the question of Whether or not a Viewer bug that affects rendering should be fixed.

In most MMO’s the idea of fixing a bug is a very simple thing. Bugs are bad and there to be squashed. However as soon as you have an economy that pushes roughly $450m a year around the world you start to get into all sorts of problems.

The most recent issue is a fix to the order of Alpha rendering on Avatar Skins. Ever since I can remember The Eyebrows and Lips were rendered below the texture and this has been used by Skin makers to allow users to modify their skins without destroying them by making the lips a little darker or adjusting the overall skintone.

As you can se ein the picture above though changing the render order to draw these elements above the texture can have some very nasty side-effects that could threaten to ruin the Skin industry.

Fortunately this fix is being reverted by Linden Labs as they aren’t so stupid as to incur the wrath of their entire customer base and destroy a lucrative industry with a few lines of code.

The original bug report (and hysterical comments) can be found here.

You can find my musings on the impact of Second Lifes economy on it’s technical progression beneath the cut.

Over the last ten years we have witnessed significant changes in Videogames rendering technology that have made games look more realistic than ever before. However these changes have always come at the cost of increasing the amount of art and amount of skill required to make that art.

Take the mode scene for Unreal Tournament or Half Life. These two games have a HUGE amount of fan-made content of a good quality. Fast-forward to today and there are less mods and maps being made for Half Life 2 and UT3 than previously despite the communities actually getting larger (well at least in the case of Half Life).

The problem lies in the difficulty of creating content for these games compared to their older more simplistic versions. Mods require bigger teams with more talent and skill. It’s harder to make a fun Deathmatch map for UT3 than it used to be not because of any gameplay reasons but because nobody will download your map if it doesn’t look any good. You may be a great Level Designer but if you aren’t also a good artist nobody will care.

This problem is even worse for a Virtual World like Second Life. For years now it’s residents have been creatively working around the Viewers limitations and even using bugs to their advantage (the entire Jewelry industry is based on twisting Primitives into sizes that they technically shouldn’t be capable of).

Second Life is stuck in a position where embracing technological advances can destroy existing industries and as a result the livlihood of countless entrepeneurs.

That’s alot of pressure to be under.

Take Shaders for example. Second Life can’t afford to increase it’s minimum requirements to support more advanced shaders for several reasons.

1. A huge number of Residents are using very old hardware that won’t support it.
2. A Two-Tier economy is created between those who can’t take advantage of this new technology and those who can.

The first point is fairly self-explanatory. Linden Labs don’t want Residents to leave the world and take their money with them.

The second point is slightly tricky though.

Taking the Skins industry as an example I would love to see a viewer that supported advanced shaders so I could see even mor erealistic Avatars with better lighting and shadows etc etc. However anybody that decides to create content that takes advantage of this new technology will find themselves in a situation of creating content twice. The High-definition versions for those with the Hardware to render it and a Fallback for those that cannot. After all even if you can render and see your spiffy Normal Mapped Avatar somebody stood next to you in-world may not. However they still need to see something and you don’t want to show up as the default avatar after spending all that money on your new look.

Could the economy in Second Life support this route of premium content for players with modern hardware? I don’t know. That’s really for an economist to answer. I do however know that it will put alot of pressure on Skin makers to support this market in addition to the existing one and only reduce competition driving out those who aren’t able to adapt to the demands required by the new technologies.

So for the time being the rendering in the viewer remains largely untouched in an effort to retain the status-quo and a stable economy.

What does this mean for Second Life in the long-term with lots of new Virtual Worlds coming along claiming super-realistic graphics as a feature? I believe these newer worlds will struggle to generate enough worthwhile content to draw people away from Second Life until it is as easy to make content as it is in Second Life.

Second Lifes biggest advantage in the battle of the Virtual Worlds is in it’s low barrier to entry by using older more stable technology.

Now if they would just let me rate Places with a star rating so their in-game search engine no longer sucked that would be abug worth fixing!

Firstly I think it’s amazing that this research is being done at all. The sheer amount of data that has to be processed to reach any conclusions is huge: 60TB for 4 years worth of gaming. Secondly I think it’s great that Sony let them have access to this data; many wouldn’t have.

However this post isn’t about the technical challenges they faced. While reading I picked up on a few points in the piece that I think are much more significant than they first appear and decided to elaborate on them.

“The average age of players turned out to be 31. “These aren’t just pasty white teenage boys in a basement—to be sure, they’re there, but they’re not typical,” he said. The older players tended to play more than the kids and, although the total hours played seem large, he said that the time mostly displaced either TV watching or movie going. And the surveys showed that those who viewed TV news in the first place continued to do so, suggesting that gamers really slotted EQ2 into their entertainment time.”

Emphasis mine.

While developing your game it’s easy to lose sight of the fact that you aren’t just competing with World of Warcraft for your players subscriptions. You’re actually competing against everything else they could be doing instead of playing your game.

This isn’t something you have to worry about in the singleplayer games market normally as you aren’t reliant on players returning to your game night after night for your revenue. If you buy Gears of War then Epics bottom line isn’t going to be affected by you deciding to watch the football instead.

However this is something that has really been evident to those playing the games. Real world events impact how many people log in to your game. During the World Cup there was noticeably less people online than normal and even the release of a Single Player game in a genre entirely unrelated to your game can affect your subscriber numbers.

In the wake of a period like last November where there was a big title being released each week for 4 weeks how many people will have felt compelled to return to their MMO after spending that long away from it playing something else?

During the recent (and still ongoing) Planetside Blog-o-War between Rock Paper Shotgun, Boing Boing and The Escapist the Dawn of War 2 beta was released. That very night I logged into the game to discover that over half of the outfit that would normally be on at that time were playing the DoW2 demo rather than Planetside. Probably only half of those people actually returned to Planetside after the demos appeal had worn off.

This brings me to the topic of churn…

“…Srivastava described how he could explore the phenomenon of customer churn, something that’s significant for any sort of subscription-based service, like cell phones or cable TV. With the full dataset, the team can now track how individual customers dropping out of the game influenced others who they typically played or interacted with. Using this data, the spreading rate and influence factor could then be calculated, providing hard measures to work with.”

Anybody who has been a member of a social group of any kind (real or virtual) understands that sometimes when people leave the group that group will start to fall apart.

This is a completely natural phenomena and one that once begun will tend to run it’s course. This is well understood by the large guilds and clans who know that if your group stays too small or doesn’t have a frequent turnover of new members then the group will become too familiar and if a popular member of the group leaves it can cause a spiral that brings the entire group down.

This is often why the largest guilds sometimes feel the least personal. If the relationships between the members gets too strong then if one of them leaves the entire structure is weakened (a little similar to a tantrum spiral in Dwarf Fortress).

What developers should be taking away from this though is that keeping individual players happy is incredibly important. Their community managers should be identifying who the “leaders” among the community are and trying to ensure that these people keep playing the game.

A well known member of the community bad-mouthing your game because of game imbalance (for example) can affect the morale of the entire community causing players to stop playing (maybe just temporarily) and dropping out of the social network they formed in your game. This can lead to other players who never visit the forums but know somebody who does stopping their play because the people they play with aren’t playing anymore. Repeat this enough times and you start to lose subscriptions.

WoW and Eve (which in my book are the most successful MMOs of recent years albeit on different scales) have clearly understood this. Churn is inevitable as many reasons why players stop playing are completely beyond the control of Developers however it’s important to identify the causes that are under your control and address them quickly (rarely quick enough).

Additionally like the large guilds maintaining a constant stream of new players is important in preventing the community stagnating and refreshing the bonds between established members. WoW and Eve have never stopped marketing themselves, always pushing to get people who have never played (or used to play but have lapsed) into the game. The number of MMO launches each year is always increasing but very rarely do you actually hear anything about them after their launch.

However I think Twitter is destined for much greater things than it is currently being used for.

Right now Twitter is largely a device for spreading memes around the Internet. It’s larger scale than IM and more immediate than email. It is the perfect platform for viral marketing.

However I was thinking about how Twitter is going to fit into other forms of media. This mostly relates to how Twitter is going to completely revolutionise the ARG (if it isn’t already) and potentially change the face of Television.

ARGs have been fairly slow moving things in the past. Where updates are carefully staged (often according to coincide with specific marketing pushes) and planned out months in advance. I imagine a future (and it may already be here) where instead of elaborate puzzles to be solved over several days or weeks we’re dealing with rapidly unfolding story-lines with many virtual actors facilitated by services like Twitter.

Updates to the story-line can happen in real-time. The sensation of being involved in something as it happens will be far greater and be capable of generating a sense of urgency in what has typically been a fairly sedate form of media.

Members of the public could be dragged into the story as well. If run well enough genuine actors could be hired to follow up on Tweets like “Am going to Starbucks, I need a coffee while I get my head around what just happened.” If the general location of the character is known to those following the story people nearby might actually be able to spot the character at a Starbucks and engage with them, take photos of them and Tweet about it themselves perpetuating the story even more.

Granted this assumes a lot on the part of the participants, that they are capable of Role-playing sufficiently that they don’t break the universe of the game (I remember the ARG surrounding the Transformers movie where you could read the inbox of one of the characters and everybody on the Internet had spammed it into oblivion with “Zomg Optimus Primelol” ).

Television can also use Twitter to grow beyond it’s episodic boundaries. Characters from TV shows are already getting Facebook pages but these are often fan pages and not genuine attempts at creating a virtual persona for the characters in a TV show that actually appears to be living in that world.

My current favourite show is The Big Bang Theory (wonder why…) and it is incredibly easy to imagine the characters of that show adopting a service like Twitter as part of their everyday lives (references to Facebook are common in the show already). Extending these characters lives beyond the weekly episodes on to the Internet through social networking services should really be a no-brainer. There are thousands of people out there who would want to be privy to the Tweets of characters like Sheldon Cooper so long as those Tweets were entirely in character (so no references to the existence of the show).

However I actually believe that if this happens it will be more likely to occur for kids shows like Hannah Montana as the goal for these shows is to create characters that kids will want to relate to and be friends with (and lets not get into the reasons for this here) and that the audience is far less cynical and more likely to believe that the person writing messages on Twitter is actually the character in the show (and not some intern on minimum wage).

Other things that might happen in Twitter:

Role-playing Games played over the Internet in public (the short message length will impact the rules used in such games in a way that I think could be very interesting and help keep momentum going).

Tweet Aggregators becoming more common place. We’re already seeing these in Cursebird which is a fairly base use of the service but it will make commentary on events (sports matches, current affairs etc) interesting. You will be able to follow the opinion of the Internet by filtering tweets based on a few keywords in real-time.

So as I said at the start: I think we’re only just scratching the surface of Twitter. This service is going to head in directions that will surprise us all (both for good and bad I suspect). I’m reminded of the War of the Worlds radio play which convinced listeners that it was a news report. I look forwards to the day when mistrust of the media is our base state due to our perceptions of reality being completely eroded.

Well it was inevitable really. If only so I could find out what they’d done to a game I technically spent 18 months of my life laying the foundations for (and by that I mean developing the failed Climax version which actually has nothing in common with the Mythic version other than the name and the creative director).

I like it. It’s more fun than World of Warcraft to me and I put that down largely to it being Warhammer. I’ve never cared about Blizzard games overmuch but I (like many UK games developers) spent much of my teens painting lead figures when I should have been chasing girls.

That plus the fact you can dive into the fun stuff straight away and not have to grind 20-40 levels before it’s worthwhile. PvP is where it’s at for me.

However…

Warhammer is proving a theory of mine: Game balance has far less to do with stats and numbers and far more to do with art.

Allow me to run with this for a while:

Take a look at the current server queues for WAR. The majority of servers have queues on the side of Destruction but only a handful have Queues on the side of Order. Now unless Destruction have less slots assigned to them this means that there are far more people who want to play as the bad guys than good.

This is interesting because it goes against what happened in WoW where Alliance outnumbered Horde.

Back when I was designing an MMO for Climax (Not Warhammer Online but another one that never got signed) we chatted with some guys from NCSoft and we got around to the topic of the “Good guy bias”. The idea that on average players want to play as the “Good guys” more than the bad guys. At the time the anecdotal evidence was in favour of this theory with just about every MMO having a bias towards the obviously Good side.

Now I think this theory is flawed. My new theory is that players want to play on the side that looks “Coolest”. 

Currently as an Order player myself (and I chose Order on the basis that Witch Hunters looked cool) we are noticing some trends in the make-up of the Destruction vs Order playerbases. Destruction fields far more tanks (Chosen and Black Orcs) than the Order side who field far more Damage Dealing classes like Witch Hunters, Bright Wizards and Shadow Warriors. We are very lacking in frontline fighters meaning Destrcution tends to run straight into the softer ranged classes right past the few front-liners we have.

This may be leading Mythic to investigate why everybody is favouring Destruction and the Chosen and Black Orc classes in particular (I don’t hav enumbers to back any of this up but I see more of them than any other class in PvP). I can save some of their effort: It has nothing at all to do with numbers or combat balance.

There is no flavour of the month effect going on here. Chosen and Black Orcs aren’t any tougher or more dangerous individually than SowrdMasters or Ironbreakers. However what they are is “Cooler”. Faced with character creation and the choice between being an Elf in what appears to be a nightdress or a Huge guy with a beard in spikey plate mail armour you instantly think that the Spikey armour guy is “Better” than the Nightdress guy.

This is what I think of as “MetaBalance”. These two classes may be perfectly balanced numerically but because one of them has art that matches the stereotype closer it affects how players choose their character and even how they play them. I’ve seen Swordmasters complain that Elves have no tank class when Swordmasters ARE their Tank class. The problem is they don’t feel or look like a Tank class and this changes the psychology of the person playing making them believe themselves to be weaker than they actually are.

This theory is backed up by my experiences as a Destruction player fighting Order on another Server. I’ve seen Witch Hunters (Who carry a pistol in their off-hand) playing as if they are a ranged class and not a melee class. One of my friends who has done more Destruction PvP genuinely didn’t relise that Witch Hunters are supposed to be a Melee class until I told him.

This gets further backed up by Warrior Priests who try to stay out of combat and heal people when they are at their most effective upfront with the Sword Masters and Witch Hunters.

On the other hand we have Destruction. Where all of their classes play alot closer to how they look. Big guy wearing Spikey Platemail? yep he’s a Tank. Goblin with robes and a staff? yep ranged caster and healer. Elf with a dagger in each hand and no clothes? Damn right that’s a melee DPS character.

Classes being played alot and well on the side of Order however are: Bright Wizards (Ranged DPS and they look it), Arch mage (healer/Caster and again they look the part), Iron Breaker (Dwarf in lots of armour and a big axe yes he’s a tank), Shadow Warrior (Blatently Legolas).

So how much does Art affect population balance in an MMO? My answer would be “Alot more than people think”. Power Gamers are the only ones that will consider statistics when choosing their class while the rest of us go with the character we like the look of.

Warhammer is currently suffering a population imbalance because the Destruction side just look cooler and appeal to more people than the Order side and importantly more closely match the stereotypes of their class.

The net result means I’m sick of seeing bloody Bright Wizards everywhere and no bloody Iron Breakers.