I'm Skaff Elias, a Senior Vice President of Research & Development. I started working for Wizards of the Coast offsite from Philadelphia when Magic was released, and soon thereafter I moved to Seattle with Jim Lin and Dave Pettey to help with design and development on location. Through the years, our group designed or developed every Magic set through Alliances, and worked on R&D for every trading card game released by Wizards of the Coast until 1997. In that period I also served a short stint as Magic Brand Manager, during which I began designing the Pro Tour and -- along with Rick Arons and many others -- developing our Organized Play programs to be world-class in their ability to support our games.
Since Randy Buehler hadn't even touched a Magic card at the time Legends was being developed, and this column needed to be about Legends, I'm "pinch-hitting" for him this week.
The Development of Legends
After its initial conception, Richard Garfield's invention of Magic: The Gathering spawned a huge group of playtesters in Philadelphia as well as groups at Wizards of the Coast in Seattle and associated others in the Pacific Northwest. I started working with Richard on Magic in 1991 while in graduate school at the University of Pennsylvania.
In 1992, the Philadelphia group, which included Jim Lin, Chris Page, Dave Pettey, and I, started creating Magic's first "stand-alone" expansion set -- Ice Age. We also helped Richard edit the initial release of Magic and, later, Arabian Nights. At this time we -- along with Joel Mick -- were also the design team for Antiquities.
Lots of submissions for Magic expansions were coming in at that time, and since WotC had no trading card R&D department, the submissions were passed on to Richard who made sure they were in shape to be published. In late 1993 or early 1994, Wizards CEO Peter Adkison told Richard that the next expansion set to come out was going to be a set designed by Steve Conard and his group of Canadian playtesters. Steve, who headed up one of the Northwest groups, was an original founder of Wizards and an avid game player. Richard reviewed their set, and while it was filled with brilliant ideas for new cards, the wording and costs for the cards were often in pretty bad shape, as the Canadians were unable to benefit from the proximity of Richard and all his design knowledge. Before long, Richard realized that the task of "developing" Legends was beyond his ability, because (unlike us students) he now had a real job as a professor at Whitman College in Washington. Steve's group had done a great job at producing playtest cards in a clean format on the computer, but the cardfile was, I'm guessing, sorted either by order of creation, or completely randomly. To this day, Richard's biggest complaint about working on Legends was that, even after multiple requests, the cards were never given to him sorted either by color or rarity. If the cards had ever been given to him sorted, we might never have been asked to develop the set.
Legends playtest cards were modeled after Richard's "Gamma" version of early Magic.
The communication problems with the Canadian team were horrendous for geographical reasons. We hadn't really worked out systems to handle the difficulties, and so any card set changes or comments were sent between four different locations by a devilishly sub-optimal mixture of email, faxes, snail mail, and phone calls. Once Richard left the project, Peter asked Jim, Dave, Chris, and I to do the development work, which was quite a shock since we thought Ice Age was to be the next set for Magic. The central office of Wizards, while wanting to do stand-alone sets, was scared by the idea of repeating existing cards. It wouldn't be for another year before they understood that there were too many basic effects that needed to be in a set before it could be played as a stand-alone. Try playing Legends by itself and the difficulty should become clear. Instead of creating inelegant versions of basic effects, designers either wanted to repeat cards or create completely new ones. The Legends team took the latter route.
Working on Legends was unbelievably hard for several reasons, all of which sprang from inexperience. The entire development process was a negotiation; the designers were supposed to have the majority of the say in the product, but the developers and editors were accountable for its final quality. This dichotomy led to a long and complicated process by which the cardset was completed, even though the entire schedule for the project -- from the initial design file through development, templating, final playtest, and editing -- was only two months long. All this happened while the designers had full-time jobs, and the developers were in the middle of graduate school in math or physics, and with editing, design, and development dispersed among three sites.
We had to get the approval of the Legends designers to change cards, but often we didn't even know what the cards were originally intended to do. For instance, the card "Tendolas Affect" said: "Creature is able to lock onto another creature to force combat. Creature locked onto can not band and must fight." This was an enchant creature spell. Which creature does it enchant? Does it grant an ability to "duel?" Is it enchanting two creatures at once? What happens if one dies? Can the "Tendolassed" creature "lock on" to another creature? Is this merely a forced block, or is it completely outside of combat? The amount of rules vocabulary and understanding shared by the three different sites was negligible. Each group had its own internalized rules system (true, one of the three systems had been published six months before, but that didn't seem to matter at the time). As you might imagine, the difficulties were so immense that tempers flared and tensions rose past the point where any work could get done. Eventually, I had to fly out to represent the development team to the design team so we could engage in actual communication.
Only a face-to-face meeting would get us out of this situation, and so I flew to Seattle and then drove to Vancouver with Peter. Anyone who tells you the US/Canada border is undefended has read too many books and not spent enough time watching Mounties tear his car apart, and probably has never spent an afternoon sitting in a border station while the guards calculated exactly how much duty there should be on five boxes of Magic cards. The answer, by the way, when the products are not for sale and the duty is based on the manufacturer's cost, is under $1 Canadian, but that took two border guards almost an hour to calculate.
After being released by the border patrol, we were finally on our way to meet the Legends design team. Once Peter and I got to Robin Herbert's apartment, all the problems we had been having (other than an incredible time crunch) disappeared. While there was still a certain amount of animosity from the big build-up of the previous month, it dissipated steadily over the course of the evening. This was partly due to the fact that the design group was a great bunch of guys with a lot of creative ideas (albeit with no rules knowledge or templating skills), and partly due to the fact that Robin had a multi-gallon cask of rye set up like a water cooler in his home. Many elements of the original Legends set were laid bare by that one appliance. I'm not sure if Robin (or any other Canadian) has ever used a shot glass due to its inefficiency -- at any rate we got by just fine with full-sized drinking cups.
In addition to the fact that they were fun people, Steve and I shared a lot of common ground in that I now hated Canadian border guards, and he was kidnapped by a Canadian woman (whom he married) and forced to live there for many years against his will. We slogged through the cardfile and I was able to get an understanding of what exactly they looking for. Eventually they understood that compromises had to be made because a simple description of what was intended on the card wasn't enough.
"It's, like, GASEOUS, eh?"
Sometimes, the design team even had disagreements among themselves about what certain cards were supposed to do. Take Gaseous Form, for example. When I offered that maybe it could prevent combat damage, an argument broke out. One of them thought that Lightning Bolt shouldn't be able to affect the gaseous creature, because, "No way would that hurt it -- it's, like, GASEOUS, eh?" I said maybe it shouldn't take any kind of damage, but some thought Fireball should hurt it while others didn't. When I asked one of them exactly how the card should work, he said it was supposed to turn the creature into a ball of gas.
Discussions like these led to a general acceptance that the cards would probably not be able to act exactly like they had in the playtest in the interests of concise wordings with universal meaning.
I left Vancouver with a list of intended effects that the developers could begin to try to simulate. I don't mean to imply that the difficulties were over; there was still a lot of long-distance arguing, but the major understandings necessary had been reached by both sides and the project was on its way. One huge factor that allowed the cardset to be completed on time was that, by-and-large, the Legends team didn't care about casting costs. If a card's effect was too powerful or out of flavor for the color, instead of trying to get an agreed upon alteration, or explaining concepts of color theme, we could just overcost the card by a mana or two -- a lot of this can be seen in Legends today.
My biggest regret from working on Legends is the time constraint we were under. Too many of the cards would simply have taken too long to playtest, develop, and word properly. The set is packed with good ideas, that, even though humorously worded, are usually the basis for a great Magic card. Most of the cards in the initial playtest file are interesting in one way or another, and one day someone will probably be able to go through the old list and dig up some new ideas if they ever get stuck looking for a card or two...
Randy sits next to Skaff, so send your comments his way at email@example.com.