Open Gaming Deja Vua clear-eyed view of open from 1999
In 1999, open source was riding high and role playing nerds were catching on. Small creators were already releasing content under “open” licenses. And Wizards of the Coast, the king of gaming companies, was coming around. They announced a plan to extract and release the “operating system” of their prize property, Dungeons & Dragons, as d20, an adaptable base for others to write and play on. This interview with Ryan Dancey, then Wizards’ VP for RPGs, sums it all up.
Most of what I know about open software I learned by reading, listening, and talking about software. But most of what I think I understand about open software I learned reading up on other fields, other industries, other movements.
The Dancey interview was, at its time, the story of a businessman and his business catching up with software. Reading it now, twenty years later, with current software industry controversies ringing in my ears, the games guy seems amazingly prescient. Dancey got “open”, or at least the business case for it. And he was able to reflect that case back, in his own words, relating the high points of the software story, but without carrying its baggage. In that distilled, streamlined form—translated from software to gaming, in the terms of business—it’s amazing how clear a mirror it makes.
It’s Good to Be the King
[W]e believe that there may be a secondary market force we jokingly call “The Skaff Effect,” after our own [game designer] Skaff Elias. Skaff is one of the smartest guys in the company, and after looking at lots of trends and thinking about our business over a long period of time, he enunciated his theory thusly:
“All marketing and sales activity in a hobby gaming genre eventually contributes to the overall success of the market share leader in that genre.”
From the business point of view, this is the fundamental insight. Control over products, services, and marketing often proves necessary in carving out a market position. But if you achieve dominance, war-born instincts of command can limit your reign. From the heights, you don’t need to control economic activity to benefit from it. It may behoove you to encourage activity outside your control, even very close to your own domain, so long as it contributes to your profit.
Wizards of the Coast found this view of its own position. Dancey was more than willing to spell it out:
We’ve got a theory that says that D&D is the most popular roleplaying game because it is the game more people know how to play than any other game.
There follows a very serious note, apparently added between interview and publication:
This is a very painful concept for a lot of people to embrace, including a lot of our own staff, and including myself for many years. The idea that D&D is somehow “better” than the competition is a powerful and entrenched concept. The idea that D&D can be “beaten” by a game that is “better” than D&D is at the heart of every business plan from every company that goes into marketplace battle with D&D game. If you accept the Theory of Network Externalities, you have to admit that the battle is lost before it begins, because the value doesn’t reside in the game itself, but in the network of people who know how to play it.
From the point of view of an upstart, like the Dominion Rules that went open before Wizards did, this doesn’t sound so great. But from Wizards’ point of view, the implications aren’t hard to deduce:
We make more revenue and more profit from our core rulebooks than any other part of our product lines. In a sense, every other RPG product we sell other than the core rulebooks is a giant, self-financing marketing program to drive sales of those core books.
You begin to see your “secondary” or “non-core” projects as feeders. You hire people to do that work internally. Others you bring in on contract. From there, it’s a short jump to contracting out, or “outsourcing”, entire projects.
Why supervise those projects at all? Why incur the costs of finding people to do them, negotiating terms, and so on? As long as the market sorts the wheat from the chaff, without besmirching your good royal name, the more peasants on your estate, the better.
Separation
Rather than controlling “open” creation, the concern becomes fencing it off, so growth accretes to, rather than detracts from or competes with, the commanding business advantage. Dancey takes this on, as many software companies do, through trademark:
The result is that D20 becomes a rosetta stone for making products that will be compatible with Dungeons & Dragons, without requiring us to issue a blanket license for the D&D trademarks. In other words, we want to use the trademarks of D&D to hold the value of the business, rather than the rules themselves.
We’re going to establish “D20” as a recognizable mark, like “VHS” or “DVD”. It will appear on all D&D products.
They invented a new name—the d20 System—so they could share that name without sharing “Dungeons & Dragons”. That way, companies that were’t Wizards of the Coast could feed into Wizards’ D&D phenomenon, without being able to put “Wizards of the Coast” or “Dungeons & Dragons” anywhere on their work.
This is classic business trademark strategy. Create something new that’s adjacent to your crown jewels, and share that. Keep the crown jewels separate. It’s the reason Harley invented “HOG”, the Harley Owners Group. Members run HOG patches on their jackets, even if they buy those jackets from Harley, even if the club isn’t so much an independent organization as a peer marketing program.
People knew how to play “Dungeons & Dragons”. Controlling that association, that name, so people follow it to the register to buy the core books, was the key. And that is a job for trademark:
Patents are great. They can lock a game mechanic away behind a legal wall very effectively. Trademarks are great. They can stop a lot of commercial products dead by refusing to allow those terms to be used. Copyrights—eh; not so good. Copyrights cannot protect an idea—just a particular expression of an idea.
Wizards’ license for copyright, the Open Game License, was fairly “open”, a riff on the GPL, a line that Creative Commons would follow with its ShareAlike license shortly after. That copyright license explicitly did not cover trademarks. Recent-ish software licenses, like Apache 2.0 also explicitly don’t cover trademarks, which is part of the reason businesses with trademarks like them so much. But the Open Game License went further, explicitly prohibiting use of trademarks to express compatibility, other than under the rules of a separate license. The license for the d20 trademarks, also published, came with more restrictions, including prohibitions on character creation and experience rules that might compete with with Wizards’ core.
However, the trademarks of the D20 System are licensed by a separate document, the D20 System Trademark License. The terms of that License are substantially more restrictive.
The license was also revocable. Wizards revoked it en masse on release a further fourth edition of D&D in 2008.
Permissions
Dancey’s interviewer uses the end of their time to list out various use cases as either inside or out of bounds.
One of my fundamental arguments is that by pursuing the Open Gaming concept, Wizards can establish a clear policy on what it will, and will not allow people to do with its copyrighted materials. Just that alone should spur a huge surge in independent content creation that will feed in to the D&D network.
The primary use case is obvious: creating adventures. Running with the earlier DVD analogy, adventures correspond to the movies you can buy and play on your DVD player. Wizards sells the player—the handbook with the core rules. Other people creating adventures that rely on their rulebook is a win for Wizards.
Dance was less interested in, but nonetheless open to, campaign settings, creatures, spells, magic items, and even games outside the genre of fantasy. They’re accretive to d20 and D&D overall, but their connection to the bottom line is less direct.
The line, as ever, is anything that says “D&D”. That’s what Wizards owns. That’s what makes the money. That’s what’s not open.
Time
Companies running an “open” play often shoot themselves in the foot over time. They’ll rush to release everything they can under a broadly permissive license, only to regret it later, when they’re unable to revoke the old license for their old code. There’s a delicate balance between releasing enough to have a meaningful effect and letting the leash out a bit at a time.
Dancey’s more careful. He describes sharing initial terms and content with select folks at an upcoming conference, under nondisclosure agreement. He’s also open to revisiting the initial set of “permissions” afforded to creators in the d20 playpen:
At some time in the future, after we’ve gauged the effects of this activity, we may loosen those restrictions.
In the end, they didn’t. They reverted to more control.
I found out after publishing this post that Law Insider has done one of their “contract teardown” interviews about the Open Gaming license. Worth the candle!
Your thoughts and feedback are always welcome by e-mail.