Diablo 2: The human cost of making a classic
At Devcom, David Brevik recalled the "terrible grind" required to finish Blizzard North's evergreen masterpiece
Diablo 2 is one of the most admired and influential games ever made, but the story of its production is a cautionary tale for every developer attempting to build on a prior success.
Speaking at Devcom Digital this week, David Brevik gave a post-mortem of Diablo 2 on the 20th anniversary of its release. As president of Blizzard North, he was in charge of the team that created the IP back in the mid-'90s, and while the first game was a success, its production was not without problems.
According to Brevik, the Blizzard North team crunched for "three or four months" on Diablo in an attempt to make a Thanksgiving 1996 launch date. In the end, the game was released in January 1997, but those hectic final stages of development left the team far from enthusiastic about starting work on a sequel.
"We didn't even want to think about making Diablo 2 or anything like that for a couple of months," Brevik recalled in his talk at Devcom.
"We just kept on designing more and more, and guess what? It kept taking longer and longer and longer"
After taking some time off, however, Blizzard North evidently recognised that it had unfinished business. Diablo was not perfect, but it was also flawed in one vital respect; it featured an early attempt at online multiplayer using an embryonic form of Battle.net, and the peer-to-peer system that underpinned it allowed for "rampant" cheating within its community.
"We wanted to make a legitimate online version of Diablo, where there's a real economy and you can trade items and it means something," Brevik said. "That was one of the biggest motivations; seeing how people loved Diablo, but being very critical -- and rightly so -- about how things had been going with the online part of the game."
This was the most pressing of many improvements the team at Blizzard North wanted to make with Diablo 2, all of them building towards an overall goal of creating something "bigger and better" -- a common phrase in the development of sequels, and one that allows ample room for scoping problems and feature creep. However, with double the amount of people working on the game -- 40, versus the 20 that worked on Diablo -- and a number of key tools and processes already established, Brevik was confident that Diablo 2 could be ""at least twice as large" as its predecessor.
"We would order meals, we would give people sleeping bags and toothbrushes, and some days people would sleep in the office"
Blizzard North wanted to add more classes, a new skills system, twice the number of movement directions, higher resolution graphics, more detailed character models and environments, and dynamic streaming of the game world to eliminate loading screens. As anyone who has played both games will know, the progress made in the sequel was substantial, but Brevik warned the Devcom audience that such ambition came at a cost.
"All of these things led to the final grind," he said. "Because we had decided to make more classes, more levels, streaming worlds, all these kinds of decisions... what we thought was going to be a two-ish year project turned out to not be that. We just kept on designing more and more and more and more, and guess what? It kept taking longer and longer and longer."
The target for Diablo 2 was Christmas 1999, and by the spring of that year it was apparent that making that deadline would involve a period of crunch.
"Nobody wanted this, but it was the way things were back then," Brevik added. "It's not a good decision. I don't recommend it. It cost me dearly. It cost everybody dearly. But it was what it was. We crunched."
Brevik recalled that crunch hours started in late April, early May of 1999. For him, that meant working every single day, sometimes upwards of 14 hours a day. Over the entire period of crunch on Diablo 2, Brevik estimated that he worked an average of 12 hours a day, seven days a week.
"It wasn't just me... Everybody was working on the weekends. Some people took time off -- I recommended that some people take some time off -- but most of the time we would work every day.
"We would order meals, we would give people sleeping bags and toothbrushes, and some days people would sleep in the office... It was a terrible grind at the end of this project. We were working, and summer became fall, and it was pretty obvious that we were not going to make it."
"It was an incredible grind on myself, my relationships, and life, and my soul"
In October that year, a call from Blizzard's main office came in, acknowledging that Diablo 2 was still too far behind to hit its original launch target. It was agreed that the game would be delayed, albeit with a general directive to finish development as soon as possible. At that point, Brevik said, "the grind became a little bit more relaxed," but he and many others continued to work every day to reach that gold master.
"I ended up taking three days off throughout the entire process of this crunch, which lasted up until the gold master, which was June of [2000]," Brevik said. "It was a little over a year that I crunched... It was an incredible grind on myself, my relationships, and life, and my soul.
"In the end we finished the game, and it came out great, but it was a terrible, terrible grind."
Of course, this is a story from a different era of the games business, when crunch conditions were not as widely acknowledged as an issue to be eradicated. It is nevertheless important to place the myriad creative and technical achievements of a venerated product like Diablo 2 in the context of how it was made -- the kind of thinking that leads to insurmountable workloads, and the huge sacrifices people are then required to make to reign in those ambitions and ship a finished product.
Time heals all wounds, of course, and that is evidently true of Brevik, who was open about both the rigours of Diablo 2's creation and the sense of pride he feels at having been part of it at all. Just over 20 years later, what Blizzard North achieved lives on, and the sacrifices that made it possible have been largely forgotten.
"Diablo 2 today is still very, very popular," he said. "There is a vibrant modding community... and it's very vibrant on things like Twitch. The other day I went on Twitch and over 2,000 people were watching Diablo 2... on a morning, on a random week day. It's incredible that there's this many people watching a 20 year-old game.
"The legacy continues, and it's amazing to see something we worked so hard on -- myself, the team, many people sacrificed and made a huge contribution to this product... For it to be so beloved is a wonderful experience. We're very blessed to have that."
GamesIndustry.biz is a media partner of Devcom Digital