Philosophical Ranting of an Engineer

I’ve been spending quite a bit of time thinking, and playing, Demigod over the last couple weeks. One of the things I’ve been thinking about and meaning to write on is character design and balance, specifically balancing out all of their spells. I went to David Sirlin’s talk at GDC on balancing competitive multiplayer gameplay and it is very relevant to this game. One thing he specifically warned about was dominant strategies. Dominated strategies are bad, but people just don’t choose those and the game continues. Dominant strategies are bad because people pick them and it ruins everything else.

The general trend right now is that Lord Erebus is overpowered and ought to be nerfed, and I’m inclined to agree with them, but I want to mention a few things about Erebus and dominant strategies. Another way to look at a dominant strategy is a dominant talent. All of the heroes have many different spells and you aren’t required to take any of them. You get to pick and choose and for the most part all of the heroes can be played many different ways. But some heroes have dominant spells. They have spells that you, while maybe aren’t forced to take, should always take. There is little reason to not take them. They are as good or better than all of your other spells, so why not get it?

Take for example Erabus’ Bite spell. This spell is available at level 1 and offers several effects: damage, heals yourself, slows the enemy, reduces the enemy’s armor. That’s a lot of bang for your buck. It has a short cooldown and isn’t too overly priced. Erebus is special in my opinion in that ALL of his talents are very useful. Other characters I’m happy to take some spells, leave others, but with Erebus, I would take all of his spells all the time. They give him so much variety. But Bite is special. I always take Bite. I can’t think of an Erebus build that doesn’t take Bite. I could make a build that doesn’t use minions, or one that doesn’t use Bat Swarm, or Mist, but Bite? Bite is always useful. In the beginning of the game I use it to heal and harass the enemy. Later on I use it to reduce their armor and slow them. In my opinion, Bite is a dominant spell and should be nerfed, not because it is too powerful against other players, but because the decision to take it is trivial.

There are other spells like this and this is really just in my experience: Rook’s shoulder towers, Sedna’s Heal, Unclean Beast’s Spit. I’m sure you all have your own. And having your own “must have” spell is fine, unless it is everyone’s must have. Then it should be nerfed or the other spells buffed so we have to make some real decisions picking a build.


April 25th, 2009


I picked up Demigod recently. It’s been one of only a few games I’ve been looking forward to over the last year or so and I was really excited when it came out and started getting decent reviews. In case you are unfamiliar with it, Demigod is a game very similar to Defence of the Ancients (DotA), a custom map for Warcraft III that I spent countless hours playing back in the day.

(Note: I mostly playe DotA 3.7 by Eul and not DotA: All Stars, so don’t come complaining to me that my statements about DotA don’t apply to All Stars.)

The gameplay mixes some RTS elements, some RPG elements, into an action based arena game. You control one hero amidst a battle between a friendly army and an enemy army. Your objective is to destroy the enemy’s citadel (Demigod has a couple different modes, but I’ve only played this one) and protect your citadel. Your hero levels up by killing the enemy NPCs that spawn (creeps) and killing enemy heroes. It sounds simple, but there are so many other things to consider: enemy towers to destroy, flags to capture, as well as items to buy for your hero and upgrades for your citadel, there’s a lot to keep track of. And yet, compared to most RTSs it’s less frantic as you only have your one hero to control.

I’m really enjoying the game so far. Using my mechanics vs. content model, Demigod is very clearly on the mechanics side of things. There’s very little in terms of back story and no single player campaign worth speaking of. The game is shamelessly designed to be playing with others. This is exactly the kind of game I enjoy playing and designing, so I’m having fun.

Enough dilly dallying, I want to talk about how this game differs from DotA, what seems to work and what doesn’t, and then some of my thoughts on the different heroes. The gameplay really is just like DotA, no striking differences although there are a few. Half the heroes are assassins and half are generals. Generals have the ability to summon more units to fight for them. There was a little bit of this in DotA, but not too much. I was a little concerned before I played demigod that the generals would be a huge change in gameplay and might break it. But after playing all of the heroes, they definitely fit and are just as, if not more fun than the assassins. It doesn’t feel like playing an RTS, just a different kind of RPG hero. Demigod introduces flags, control points around the map that give your team various bonuses: more mana, more hit points, stronger stats, or more concrete things like controlling a gold mine giving your team income or a portal that will spawn units for you. This was another aspect of the game I wasn’t sure I would like, and while it gives the gameplay a different feel from DotA, it works. On some of the larger maps it seems like you spend more time capturing flags than fighting, especially if the flags are off the beaten path. Last but not least in my list of differences is the citadel upgrades. I’ve played enough other DotA style Warcraft III maps that have citadel upgrades that this didn’t really shock me and it does give you something to think about other than just buying weapons for your hero. It works, adds variety and options, and keeps heroes from being completely overpowered against creeps and towers late game, an issue DotA did have.

One more expected change from DotA, is different maps. DotA, as a custom Warcraft III map, had only one layout and I know it very well; three paths, one right up the middle and two more on either side of it with very few ways to get from one path to another. With five heroes and three paths, you had to carefully pick which paths were going to have two players and which player would be alone. Demigod has an interesting variety of maps ranging anywhere from almost completely linear with Crucible to huge with many criss crossing paths on Mandala. None of them feel as tight as the original DotA map, but that just be because I haven’t played them enough. My favorite so far is probably Exile, a windy sort of map that takes place on the backs of two snakes being wrestled by what reminds me of a Greek god sculpted out of stone and floating in space. You can see constellations in the background and if you look down you can see a green planet. Which brings up an important point, the game looks amazing. The maps are especially beautifully with their detailed backgrounds, although the terrain itself is completely flat. The heroes are also exquisitely detailed. One of the heroes, the Rook, is a huge lumbering castle. One of his upgrades adds a trebuchet on a tower above his head. If you zoom in, you can see the gears turning on the trebuchet as it fires! The game runs on the Supreme Commander engines and lets you zoom far out so you can see the whole map, which is really useful in terms of surveying the map and seeing what goes on. It also gives you a good view of the fog of war, which actually looks like fog; very cool.

Now to the heroes! There are eight heroes in Demigod, four assassins and four generals. I’ve played all of them at least once and most of them several times. Unlike DotA, each hero has a tech tree requiring the player to choose how they want to build their hero. As the Torchbearer, do I want to focus on my fire attacks or my frost attacks? This really adds a large amount of depth to the game compared to DotA. In DotA you had 10 skill points, 3 abililties you could put 3 points into, and 1 final super spell. By the end of the game every hero had all their skills maxed out. By offering a large tech tree, you can play a hero very differently from one match to the next. From my experience, and reading online, there are few must have skills or completely worthless ones. I’m sure this will change as people play it more, but for now there is a lot of experimenting. Even though most of the heroes seem decently well balanced, there’s also the issue of how fun they are. Obviously, this changes from person to person, so my thoughts on the heroes are my opinion and even though I may not like one hero, it doesn’t mean no one will. Also, there’s some back story on all the heroes, go read about them somewhere else if you are interested; I’m just going to talk about how they play.


Torch Bearer

This guy is the closest thing to a typical mage. He relies on his spells to do damage and is pretty fragile. He’s one of two heroes who have two forms. His are fire and ice. In his fire form he has three spells (instead of a 4th, he has a “switch to ice” spell): Fire Nova, Fireball, and Circle of Fire. Despite being a mage type character, two of his three fire spells require him to be in the thick of things. Fireball is a ranged “spike” spell and very useful to taking out towers or enemy heroes, while Circle of Fire throws down a circle of flames that do massive damage over time, and Fire Nova is an instant cast that will throw all small units around him into the air killing them instantly as well as doing damage. In his frost form, the Torch Bearer has Frost Nova that does an AOE stun around him, Rain of Ice which throws down a bunch of ice crystals at the targeted location, and Deep Freeze which is the frost spike and interrupts the target. The Torch Bearer is pretty fun. He is amazing against creeps, and can dish out a lot of damage against enemy heroes, but he is super weak and will die if attacked much. None of the items buff your spell damage, so late game he is forced to buy health and armor items it seems, reducing his offensive capability. I don’t know if this is a problem yet or not. I just know I end up with a lot of money wondering what to do with it. Most of my experience is with fire, but I plan on trying out a frost heavy build soon and then a half and half build.

Fun: 4 Strength: 4

Unclean Beast

Your typical… demon dog thing. He reminds me of Naix the Lifestealer from DotA. Small, fast, uses his auto attack to do most of his damage. The Unclean Beast also has a variety of poison and diseases that allow him to harass the enemy from range and is one of the most effective hero killers in the game. If he can close the gap on one of the weaker heroes, it’s an almost certain kill. He does feel a little squishy though, especially at the beginning of the game. He’s one of those characters that late-game, if you got the money to get good items, he is an unstoppable juggernaut, but if you didn’t then he’s worthless.

Fun: 3 Strength: 5


Regulus uses his crossbow to attack from afar and as the squishiest character in the game, he does very poorly up close. If you focus on his hero killing abilities, of which he has several including Snipe an ability that reaches halfway across the smaller maps doing more damage the farther you shoot it, he doesn’t do very well against creeps and generals with a minion army. While I can appreciate his strength, he seems rather boring to me, which kind of surprised me. I love playing as dancing hit and run characters in games like this. It just seems like there aren’t too many interesting abilities for him. Get a good weapon, buff up his passive abilities and just auto-shoot.

Fun: 2 Strength: 4


The big walking castle you see on the box. The epitome of a walking tank, the Rook is very slow, quite powerful, and extremely impressive. He’s got some pretty cool passive abililtles. He can grow towers on his shoulders that shoot arrows and light beams and on his head a working catapult! He can also pull towers out of the ground that shoot enemies, and if need be he can suck the life out of those as well as enemy towers to heal himself. Seems like the Rook can be a decent hero killer with his stun and hard hitting Hammer Slam, but you need the enemy to commit to being close, as the Rook isn’t running anyone down. Instead, focusing on taking out enemy towers is a better strategy. I’m not sure if I like the Rook yet or not, he has potential, but with so many relatively passive abilities he almost falls into the same category as Regulus.

Fun: 3 Strength: 4



This guy is badass. He’s a huge, clanking suit of armor that summons ghosts from every unit that dies around his totems. Very moment dependant, he can be very hard to stop in a push, and with his invincibility shield that he can cast on friends, is very powerful with a teammate. I have played him with more of an assassin build, no minions at all, and he is still pretty powerful. He’s got some decent attack spells and a lot of spells that heal and buff him. Definitely one of my favorites so far.

Fun: 5 Strength: 5

Queen of Thorns

The Queen of Thorns is kind of odd. As the only other hero, apart from the Torch Bearer, with two forms, she’s a little strange to play. In one of her forms she wraps herself inside of a flower gaining improved armor, health and mana regen and can summon minions as well as shield herself. When the flower opens up, she gains access to damaging abilities that make her effective against creeps. With little to no hero killing capability, she seems instead relegated to support and indirect attacks. She can be very hard to kill if played properly and is an excellent support hero, but her lack of offense may turn off some, and she doesn’t feel very intuitive to play. Definitely a character I’m looking forward to figuring out.

Fun: 3 Strength: 3


The healer. There was one on each side for DotA and there’s one in Demigod. Supposedly, she can be pretty powerful and can have one of the strongest minion armies with all of her healing auras, but she just seems boring to me. With really only one offensive spell, Pounce, doesn’t seem like you do much with her other than heal and silence the enemy. She’s no doubt very powerful, but I just can’t get myself to play her.

Fun: 1 Strength: 3

Lord Erebus

Ah, the vampire lord; cheesy and ridiculously powerful and fun. This guy has everything: an army, a damage spell that heals himself, an AOE stun, LOTS of self healing and healing auras, an ability that lets him teleport short distances doing damage and either surprising an enemy or escaping one as well as a superb creep killing spell that leaves in invulnerable while using it. All of his abilities are useful and I’m really looking forward to playing him with as many different builds as possible. His true strength is that he is almost impossible to kill, but he is an excellent hero killer as well. I can’t say he’s terrific at pushing towers, and instead needs an ally or creeps to help with that. Fun: 5 Strength: 5


So there ended up being a lot of 5’s and no one below a 3 in terms of strength with only one character getting a 1 and 2 for fun, respectively. This is pretty good. When I was at GDC, I heard David Sirlin give a talk on balancing multiplayer and one thing he said is that you want all of your characters to feel powerful, and most of the characters in Demigod do. My own personal distaste for Regulus and Sedna probably reflect poorly on their perceived strength; from what I hear they are both very powerful characters. I’m not hearing any huge uproar over one character being ridiculously overpowered at this point.

Demigod is a lot of fun so far. I’ve spent two nights in the lab until 3am playing with some friends and I expect to play a lot more. I’ve been looking for some really good LAN games over the last few years and my friends and I always resort to custom Warcraft III maps and TF2, but I think we have a new challenger for keynote game.

I’d like to post some more on specific heroes and what I’ve found to work and not work, we’ll see how busy I am with the end of the quarter coming up.

Army Ants Download

April 4th, 2009

A download for Army Ants has finally been released!

We demoed it at GDC last week and got a lot of really good feedback, as well as hearing from recruiters over and over that I need a playable demo of any games I’ve worked on. Thus, this last week I’ve really pounded away at a few nitpicky things like GUI, controls, menu and got it in a state that it can be distributed. Please, go download it and try it out and let me know what you think.

Edit: Official site

Capstone Day 3

March 11th, 2009

Yep, development on my capstone project has started and it is the beginning of day 3! I will try to keep this blog updated with our progression as well as some more in depth articles on what I’ve been working on.

On Monday, I began work on my deferred renderer, trying to get multiple render targets (MRT) working but didn’t have much luck. Yesterday was much smoother and I got MRT working, however it became apparent to me that a standard pipeline with some simple scene graph management was more important right now than a fancy deferred renderer so I switched over and started implementing the scene graph stuff I designed last quarter.

I would like to write up a more official post on the system later as I’m quite proud of it, so for now I’ll just tell you that it is a 3-part system.

  • Scene Graph
  • Spatial Graph
  • Render Graph

Scene graph updates hierarchical transforms and passes that on the spatial graph which manages collisions and frustum culling and passes objects in the scene to the render graph which can do render state sorting as well as batching and instancing and does the final draw call. That’s what I’m working on today.

Tank Frenzy (working title)

February 27th, 2009

It is finally done! After a quarter of work, our design doc for our capstone game is finished. Weighing in at 243 pages, this thing is massive! Go check it out.. Be warned, it’s about 8 megabytes.

Warship Armada

October 8th, 2008

In the spring I had a class, 2D programming. In that class we made a game. For whatever reason, I never got around to blogging about it and never even took screenshots to show anyone! The situation has now been remedied.


July 27th, 2008

I was actually going to write a post. An honest to goodness post. But when I logged in I noticed i had about 3000 spam comments so I spent a couple hours deleting all of those and now it’s bed time. So instead of the brilliantly witty post I had half constructed in my mind you get this junk.

I’ve really been wanting to read more science fiction, ever since last fall when I realized I was so poorly read in that genre. In response I read Starship Troopers, The Forever War, and Flatland over Christmas break. Since then I haven’t really read anything. Last night I bought Watchmen, Wanted), and Dune. Wanted you might recognize as graphic novel that was adapted for the recent move with Angelina Jolie. “Watchmen” is probably the most widely praised super hero graphic novel and the only comic book to win the Hugo Award. It is apparently being made into a movie and will come out next year.

That’s what I bought. Two graphic novels (comics) and one of the most important science fiction novels ever written. Don’t worry, I have a long list of books on my Amazon wish list.

That’s all for now. Perhaps I will post again soon and I’ll tell you about what I’ve been doing this summer.

In 2005 Guiding Hand Social Club, in the game Eve Online, executed the great virtual heist in the history of gaming. The mercenary corporation was fulfilling an assassination contract on the CEO of Ubiqua Seraph. However, not only did GHSC destroy her ship, destroy her escape pod, and then capture her frozen corpse, they also infiltrated her corporation in total. Over a year in planning, Guiding Hand operatives infiltrated Ubiqua Seraph from top to bottom which allowed them to not only gain trust but also access to the corporation’s hangers. In a perfectly timed climax, Guiding Hand assassinated the target and simultaneously looted all of their target’s hangers stealing over 20 billion ISK (Even Online money) and inflicting damage estimated at 30 billion ISK. The amount stolen was valued at over $16000.

This is of course an operation executed on an unprecedented scale and rocked the Eve Online community as well as spreading to the entire online gaming community as well as traditional media. They key point to this story is that all actions while “illegal” and “wrong” in a role playing light, are all allowed within the game. Other players responded in a variety of ways, some praising Guiding Hand for their accomplishment, someone glad that the assassination target got what she deserved, and others condemning Guiding Hand for going beyond their assassination contract when they completely dismantled and destroyed Ubiqua Seraph. A few people even pleaded with the game developers to step in and condemn such an action but they apparently refused. Eve Online is one of the few games in existence that is open enough to allow for such events.

This brings to light the issue of self governance in games and how much the developer should step in when mistakes happen or contracts are broken. In World of Warcraft, if you accidentally destroy a valuable item you can often ask a GM (game manager) for help and they will recover the item for you. If someone hacks your account and steals all your gear they will attempt to return your character to his previous state. But if a guild member steals all of your guild bank’s items and leaves the guild, Blizzard has been satisfied to say “be more careful who you let access your bank.” When should player’s actions be governed by players and when should it be governed by the developers? Much of it depends on what actions have been built into the game and balanced. If the developers have attempted to make item transactions as secure as possible and there is a sense that trades agreements are sacred, then taking advantage of a loophole could justify punishment. But a game where theft and heists are expected, one should not expect the developers to police this activity.

Should all games allow events like the Guiding Hand heist? Of course not, but stories like this engage our imagination in ways that other games can’t.

Citations: GHSC Heist Press Release PC Gamer USA Article Wiki on GHSC

I found Burke’s article on self governance in games particularly interesting as I’ve spent time thinking about the possibilities of self governing economies in game, but I hadn’t thought about them from a player’s identity point of view. Nor had I considered the identity of a player in an MMO as a citizen of developer-dictator state but it makes so much sense in the context of community forums. The WoW forums are rampant with “why did you nerf my class?” “my class needs to be buffed” “this situation is imbalanced, fix it!” As citizens of a world where they don’t have a vote about how it is governed, the only thing the players can do is voice their opinion. I imagine a game where much of the hot balancing topics were controlled by players, through a voting mechanism perhaps, much of the complaints would disappear. Players would realize that they are no longer governed with an iron fist, but determine their own fate. Whether or not mob-rule makes for good game design is an entirely different question.

I do think there is oppurtunity for in-game, mechanic relevant self-governance and not only in the realm of economics where self governance seems to be widely accepted as a good thing even if it’s not always implemented. Allow players to form governments and manage the resources of their nation to buy NPC guards or fund quests or subsidize certain crafting professions. The biggest problem is who makes those decisions? A purely democratic approach could work technically but it still lacks the vision of a single leader. Probably the easiest would be to piggy back on guilds that are already mainstay in MMOs and let them get voted into power. Or give anyone who wants, a chance to rule their own city/nation and like the saying goes “if you build it, they will come.” Self governing leaders can be problematic if people are unaware of the cultural situation in the community. The moment you raise some of the players up over other players in terms of importance, people beging to feel they are getting the short end of the stick. Or if players are unaware of who their leaders are, how can they make good decisions when picking them? A developer would have to be very careful in implementing something like this or else the self-governing system could be worse than the developer-governing system.


I found the paper “Tradegies of the ludic commons – understanding cooperation in multiplayer games” to be particularly interesting both from a game theory point of view (an interest of mine) as well as from a cooperative design point of view (a significant interest of mine).

Much of the article deals with trust, when and why do we trust people we are competing against. Obviously the kind of trust we have for our teammate is a different kind of trust we have in our competitor but it can be just as important. In the Warcraft 3 custom map playing community there is a widespread dislike of people who disconnect partway through the game. Unlike the standard game type, there is no penalty for leaving halfway through the game and many maps work best when there are two teams of five each. If one person leaves 10 minutes into an hour long game, the rest of the game has been ruined for the other nine people. This means that people are often very picky about who they allow in their game. If you download the map slowly you are targetted as one with a slow connection and a high probability of disconnecting and booted! It is worth the extra few minutes wait up front to find ten “reliable” players.

In the same way that we must trust our opponents to ensure a good match, we also want to trust our teammates especially in strictly cooperative setting. In World of Warcraft players are careful to ensure that their teammates for a dungeon run are of the appropriate level, appropriate class, and appropriate spec. There is little room for lenience when it might mean an unsuccesful run. However more leniency is given when that player is a friend or a guildmember. This seems to go back to the whole Monkeysphere as well as Smith’s discussion of clans. By only playing with clan members, in a cooperative or competitive setting, the player is more likely guaranteed a positive play experience as opposed to playing with random people where the experience is also random.

Of particular note with this article is the discussion of Ebay’s reputation system and how few or no games have ever implemented something like this. Obviously a rating system in a game like Starcraft would different from a game like Counter Strike which would differ from a game like World of Warcraft, but the idea that players are rated based on their past performances is interesting, especially when that player is an opponent. Would people positively rate a challenging opponent after a difficult, but fun match? Or is a reputation system for opponents only useful for marking deviant play? This is something I will be exploring in my final project and will be writing on later.


Apparently Google has released their Google Social Graph API. I’m not entirely sure how it differs from previous social API’s but the concept and the timeliness of if are what is interesting. I”m still a little fuzzy about the implementation details, apparently you embed some code on your own domain so that other sites can find information about you. This means that Facebook and Twitter and Amazon all need to implement Google’s API right? Sounds good for fast action web 2.0 companies but sluggish old-fashioned sites (looking at you Facebook and Myspace) might never implement this.

Much of the API feels like OpenID in that you own links, not Google. The author recognizes the power of API’s, allowing people to do things the creator didn’t expect. Would it be possible to make a decentralized social networking site? I could see something like this taking off amongst leading bloggers as an alternative to current solutions.

After looking further it looks like the primary purpose of this API is to let Google index it. Simply by putting information on your site specifying who your friends are, Google can index this.

It appears there is some dissenting noise on this and the lack of privacy that it provides. The author seems to think that the API gives users enough control over their privacy, but I’d have to dig into how it actually works before I formed my own opinion. It’s very interesting at the very least. But as it requires someone to throw up their XFN or FOAF text on their site and who but technologically adept bloggers will know how or bother with this?

“The people who exist outside that core group of a few dozen people are not people to us. They’re sort of one-dimensional bit characters.” And yet it’s the good will towards people outside of our Monkeysphere that’s the lubrication that makes communities work. Of course I may drive your kids to school when you are sick because you’ll drive mine when I’m sick, but what about giving that old lady a seat on the bus? She won’t ever return that favor. Perhaps it’s the religious regions mentioned in the article, perhaps it’s a general good will, whatever it is, community falls apart without it. It’s basic game theory with the biggest payoff requiring all of us to work together and no one cheating. Unfortunately this means cheating pays big. For some reason we end up helping each other out just a little bit and things work much smoother.

But what about online where anonymity rules and people seem inherently more hostile? A big problem with MMORPGs in the past has been ganking (killing other players, usually through some sort of sneak attack giving you an unfair advantage). World of Warcraft sort of solved this problem by forcing players onto one of two teams. You can’t gank players on your team, but you can gank players on the other team. This means that half the population (the other team) is outside of your Monkeysphere. Of course 99.99% of the people on your team are also outside of your Monkeysphere. But something strange happens. All of a sudden it’s not every man for himself where we make small groups or tribes that band together for survival. No, it’s us versus them! You may not know anything about PantyMcPants over there, but he’s on your team which means that when the other team comes to kill you, they are also coming to kill him! So through unspoken agreement, you protect each other because it improves your chances of survival. But that doesn’t really fall under the Monkeysphere anywhere does it? No, it’s game theory again. (game theory <=> social interaction?) You’ll do unselfish things for people inside your Monkeysphere and you’ll do selfish things for people outside of your Monkeysphere.

Say I have to kill some uber boss for a quest, right as I’m about to kill him someone on my team runs up so I invite him into my group so he can kill the uber boss too. That seems rather unselfish, after all I gain nothing from it, but the costs are low too. This falls under the community lubrication category. But the very next moment I might be farming Motes of Fire (gathering a rare material) with two other members of my team and I’m doing my best to collect these motes faster than the other two guys because even though they are on my team, they’re outside my Monkeysphere. All of a sudden two guys from the other team show up to farm motes. More competition! But these guys are on the other team which means I can kill them! So my two “buddies” and I team up to force out the rest of the competition. The moment the threat is gone I’m back to killing Fire Elementals before my two teammates can kill them.

So it seems like we behave the same way online as we do offline. Friends still belong in the Monkeysphere and the enemy is still the faceless ganking “shitcamel” lowlife. And everyone else is just looking after themselves. Only sometimes that means helping you, especially if it’s easy and it greases the community gears.

Near the beginning of a game night, a nervous guild member logs in and types ”/gu looking for healer for Shadow Labs run so I can get my Kara frag.” There is silence for a bit, and finally the sole guild priest online says “I will grant you your wish. Some day, and that day may never come, I will call upon you to run my alt through Gnomeregan.”

Guild are like the Corleone family – WowInsider


Clay Shirky’s “A Group Is Its Own Worst Enemies.” This is the first paper we’ve read that specifically addresses designing for communities and it excites me.

Throughout my use of different social tools on the internet it is clear that succesful communities police themselves to avoid being their own worst enemy. Every forum I’ve ever seen that gets used on a regular basis has an Off Topic section because people are going to post off topic and if you don’t have a special section for it, then it goes into the topic section. Shirky addresses this towards the end of the paper where it addresses a meeting where everyone had too much fun talking to get anything done. If a forum has a purpose it has to control the fun talk.

The one forum I’ve spent more time on than any others is the forum on Anandtech is a large hardware review site and its forum is one of the largest English message boards. It is sorted into several software and hardware discussions including a Highly Technical section. This is one of the least used, yet most highly policed sections. Post in there asking how something simple works and you will get jumped. The users know what this mini-group is for and they squash innapropriate behavior. Contrasting this is the For Sale/Trade sub forum. This is also policed, but exclusively by moderators. In this setting community policing is not enough, governmental intervention is required. All this from the same “community!”

I found the analysis of healthy group size on IRC fascinating. I use IRC frequently and while I would only call myself a “member” in two or three channels I idle in about fifteen. Some of these are too small to maintain conversation while others are too large and it becomes difficult to maintain a conversation with anyone.

Shirky asks why social software and blogs didn’t catch on before. What I’m wondering is why they replaced IRC and chatrooms but not forums. Of course social networking sites like MySpace and Facebook are much more appropriate for building and maintaining personal relationships but where is the discussion? By focusing on who I am and who you are and what our relationship is, these sites seem to prevent us from having discussions on topics we are interested in. If one is interested in motorcycle and wants to find an online community, he looks at forums not MySpace. IRC use seems to be restricted to more “geeky” interests.

Perhaps this is why, while I have a Facebook account, I don’t log nearly as many hours on it as some of my friends. I know who my friends are in real life and I don’t need to document it on a website, I’m more interested in conversing with others in certain fields. And since I do this to pass the time or to distract myself while doing something else, I tend to use IRC more than forums.

The last topic that really piqued my interest was Shirky’s insistance of handles in building an online community. There are online communities that purposefully prevent users from identifying themselves through a handle. 4Chan is the one I’m most familiar with. Every user posts as Anonymous and while it is possible to identify yourself from post to post using a built in hash function, the community frowns on this behavior in the same way that some Usenet groups frown on anonymous postings.


Not a lot of thought provoking content in this post, mostly summarizing things I already know like Constance Steinkuehler thinks MMOGs are the new third place and she uses “The Great Good Place” as an argument for that. But the author is a librarian and argues that libraries are becoming less and less of a third place. What? Since when were libraries a third place? After reading part of “The Great Good Place” and seeing Oldenburg’s defining traits I’m not sure I see libraries matching those. Where is the cameraderie in a library? Where are the familiar faces? Continuing with this line of thought “The only thing wrong with the third place concept is that it’s being utterly co-opted by corporate America.” How is Starbucks at all like a third place described by Oldenburg? Oldenburg describes a place where people know you and you are all members of a community, but none of that exists at Starbucks. People come to Starbucks (and coffee shops in general) to be around other people and drink their coffee and read their book and won’t ever talk to anyone else. Yes, this points to people needing proximity to others in a public place, but none of the community that is so vital to Oldenburg exists.

Of course libraries today are not what they used to be. Perhaps there were more book clubs in libraries I don’t know. The libraries I go to seem very detached from the “community” around them, but that could just be me.


Tonnies book introduces the idea of community vs. society. That they are not equivalent. I would like to redescribe them in my own words. Both offer a feeling of familiarity, the difference is whether this familiarity is personal. You are familiar with those in your community and you know their name, or you share enough in common that you consider yourself a group or a community. In a society, you share a familiarity with others in the society, but they don’t necessarily know your name and there is nothing betweeen any two individduals that justifies further intimacy.

While one might bowl frequently and be part of the community at this bowling alley, when on a trip if he were to visit another bowling alley, he would be familiar with the culture of the society, he would feel no community with the people at this bowling alley.

“The Great Good Place” doesn’t really go into details of community vs. society, it does emphasize the importance of community through a third plane, that which is not work or home. People need to have a place, or a community, they can be a part of. This I agree with and I can see this in my own life. But either because the author doesn’t see the importance, or it’s simply not the topic, he doesn’t discuss the importance of relationships outside of the third place. One thing that is very important about the relationships at this third place is you don’t have great intimacy. Two people might meet every day at this third place and not know the other’s age or the names of their wives.

People are spending more time in their home and to combat loneliness spend more time on their phones and on the internet. But what of the person who only has relationships in this third place? What of the person who has no one he can invite over to dinner? Perhaps not as widespread a problem, but the lack of personal relationships a third place cannot offer could be just as destructive.

Citations: “Community and Civil Society” – Tönnies “The Great Good Place” – Ray Oldenburg “Bowling Alone” – Robert D. Putnam