Project Horseshoe’s design papers confirm decades of MMORPG lessons


Not everyone involved in game design is trying to squeeze every last penny out of you. Yes, a lot of them are, but many others are in it for the love of design, and that’s the takeaway from the papers emerging from Project Horseshoe 2018.

Project Horseshoe is an invitation-only “think tank” for game developers to exchange ideas and build on them – think of it like GDC, only much smaller and much more intimate.

“Project Horseshoe is a conference that brings together a group of highly skilled and experienced games industry professionals, who have a primary or strong background in game design, with the goal of using the combined knowledge and skills of these people to positively influence the art and science of game design.”

Dig around on the website and you’ll see that MMORPG studio ArenaNet (of Guild Wars 2 fame) is a key sponsor and in fact sends multiple developers to attend.

The most recent junket came to our attention thanks to MMO developer Raph Koster, who pointed out two papers from the event that are relevant to the MMO industry in particular: Towards Consistent Social Game Design (Cammarata et al.) and Design Practices for Human Scale Online Games (Youngblood et al.)

The former discusses how social features like “guilds, gifting, trading, friend assist, raids, leaderboards, events, even friend points” impact engagement with the game – they’re glue, as MMO players know. But they’re also difficult to design and test. The authors also make heavy use of Koster’s Trust Spectrum, which was published last year. The latter paper focuses on the “massively” part of MMORPG, wondering whether bigger is always better – that one relies Dunbar’s Number, the concept that humans can’t maintain more than 150, meaningful relationships at a time.

Of course, while it’s fun to see some of these revelations in published papers, they’re also not new, as Koster himself is fond of pointing out. “I want to underline how much of this MUDs and early MMOs knew at the level of practice, if not science quite yet; and how much of that design practice has been neglected since,” he writes.

“From basic things like the size of the typical MUD population, guild, and party; to the common practice of greeters or mentors for newbies; to all the lessons painfully learned when [Ultima Online] suddenly provided greater scale and slammed headlong into the issues of scale; these issues have served as basic design patterns ever since. This is why we fought to have sitting in UO, towns in [Star Wars Galaxies], character customization and dyeing clothes, reputation systems, player-formed guilds instead of dev-made ones, moods and say alts, emotes, head tracking for speakers, variety of roles that the game actually rewards, cognitive downtime to provide time for bonding, level design with ‘cozy world’ pockets and spaces for ritual, systems with shared risk, larger-scale collaborative tasks that could be contributed to asynchronously, and even abandoned tools like in-game mail systems and forums so that a sense of social center inhered within the game rather than solely on external sites. I could go on and on, but I would highly encourage designers interested in specific practical system examples to go back and look at online game design predating 2004, more or less – basically, before the larger game industry jumped in and copied practices without understanding the forces that shaped them.”

Maybe instead of reinventing the wheel for the fifth time because we didn’t read the instructions, we could be building spaceships.


No posts to display

newest oldest most liked
Subscribe to:
Loyal Patron
Bývörðæįr mòr Vas´Ðrakken

Many of these things have proven out in both single player games and the massive sprawling mmorgp games.

We had color options in the adventures of zelda, but even back then in a single player game the option to change the color paled after a day or so because it was too limited. People associate colors with different things than the designers are focused on. Lucus wanted the earth tones to represent the jedi so they would feel more natural as toys, while the bright shiny colors were supposed to feel like cheap plastic. Only when the toys were made the red dye came from fire works and the earth tones had to be ink washes over beige undyed plastic resin. So the red storm troopers ended up feeling like they were the more eternal option and the plastic ink washed colors faded too fast. It impacted all the star wars properties after that. Which resulted in the Jedi not wearing alligator skin and exotic leathers but boring terry cloth and cotton bath robes accented by leather patches like on a smoking jacket and the empire getting the more desirable individualist bright colors, that you see on the fighter pilot helmets but not the normal players. He tried to fix it with the volumes one through three but for star wars galaxies the developers were limited initially from making the jedi feel like and appealing character, as the light sabers on the screen show great contrast and you can tell an actor to pretend they do not look ridiculous but a player wants to feel their character has both an impact on the story and is appealing to their idea of who their character is.

Video games as a whole tend to miss that concept in the execution as the artists focus on crating the npc who the characters take their style from while assuming if a player looks too appealing everyone will simply want to look like that player. Yet you can see where players have dye systems that are unlocked, players tend to strong bright colors contrasted with the absence of color black. which is why modern games are treading toward shader systems that created patterns of colors that work well together with the options of materials so that as people are able to find the color and texture they feel is appealing, they chose that over red/pink, blue/purple and black.

Many times these concepts are left out of small projects because they are often considered trade secrets as when companies figure them out they make everyone sigh five year nondisclosure agreements to have time to implement a system before their competitors for their monthly entertainment budget implement a better system. Epic is working toward providing the source code to be the company that gets the ball moving. They charge for supporting the engine more than the use of the code at this point, but the systems they do not have written is the in house database systems that create the social interaction because companies do not want outside companies to figure out how to access their databases from a client.

The data base warehouses are something companies with less revenue struggle with resulting in the large companies having a larger advantage with the amount of assets and individual space to hoard stuff. A good data base manager makes six figures because they can make that or more in the financial field. This creates the assumption that the supervisors have to be managers to justify the same kind of income.

The truth is a well functioning company has an ownership bracket that represents the people that own the company or corporation and take responsibility for all the employees well being below that. Below that are two paths one of management that makes sure everyone has something to work on and supervisors that are responsible for making sure that the task that is done meets the quality standards that the ownership bracket thinks is necessary to conduct business. So when you merge the management or create it as a level above the supervisors you end up with a company that fails. The management has to be the check on the supervisors by assigning tasks and making sure people are working while the supervisors can focus on training their assigned people and when they can only handle so many people and have too many worker bees, they find themselves assigned to a totally new team to make sure that the worker bees do not end getting blamed if the supervisor feels they have less power. Good super visors work on getting people trained to the level they need and focus on the new project, bad ones spend more time writing reports of how they are not getting enough people to do the job. At a small company when that happens people quit and often times the projects fail because the cash reserves to see a company through problems do not exist yet. It is why companies that are well tuned hate working with publishers because it means that when the publishers want to be able to eliminate the problem developers, the teams that work well do not want the swapping on projects because the appeal of a team is diluted by working on more parts of the project and keeping a target audience is harder. It is one place were hollywood still excels over most video game developers agile development. An artist should be able to work on the good guys and evil villians with the same amount of skill. Many teams in video game development want to only paint the heroes.


I want to point out the greeters part. Many times I tried an MUD I felt immediately welcome to the community. To start with on most there is a message that has joined the world. If you figured to talk in help you were immediately responded, usually from more than one helpers. If you didnt usually some helper will message you to find out, without disturbing/pushing, if you needed help. Always usually first in person before going OOC. When you make a new char on any modern mmo you are literally alone usually till high lvl when any guild lacking members notices you. By the way this wasnt the case for older mmos too. Every single mmo I tried pre – 2005 I got greeted by one or more people before even trying to talk on world chats. Also people run cities-states and guilds helping on the next steps people to integrate on these structures. I think at some point the mmo industry shifted to taking social power/tools from players and increasing content turning them slowly to semi-single player experiences. Because we got like 15 years in this situation people nowadays even find it normal for mmos. There are sure many positives as well to modern mmo gaming but we slowly lose the RP/RPG part and the Social part in them

Raph Koster

I feel like every tweet becomes an article! :)

The one from Cammarata et al is a follow-on to the Trust Spectrum — Aaron Cammarata was the instigator and leader of the Trust Spectrum work.

Game design often reinvents things, because it’s still pretty loose as a discipline, and there’s not a lot of going back and reading old writings, I think.

Loyal Patron
Cosmic Cleric

Raph Koster
I feel like every tweet becomes an article! :)

Love your work! Wanted to get your opinion on something please.

In my lowly layman’s opinion, the issues we are having these days is that artists have to take orders in game design from ‘max profit’-focused executives. In essence, the old days of just “build it and they will buy” is gone, replaced by designing game systems to maximize profit regardless of the impact to the core game play, or health of the player.

Is there a way of designing game that are fun and healthy to play, that will allow for the type of profits that the executives/stockholders demand?

If not, then these kind of meeting end up being IMHO a waste of time, ultimately, and, unfortunately.

Kickstarter Donor
Loyal Patron
Patreon Donor
Tobasco da Gama

Just wait until they start turning every one of your comments into an article. :D