Gamasutra is part of the Informa Tech Division of Informa PLC

This site is operated by a business or businesses owned by Informa PLC and all copyright resides with them. Informa PLC's registered office is 5 Howick Place, London SW1P 1WG. Registered in England and Wales. Number 8860726.

Gamasutra: The Art & Business of Making Gamesspacer
Postmortem: Blizzard's Diablo II
arrowPress Releases
March 4, 2021
Games Press
View All     RSS

If you enjoy reading this site, you might also want to check out these UBM Tech sites:


Postmortem: Blizzard's Diablo II

October 25, 2000 Article Start Previous Page 2 of 3 Next

1. Diablo II is still Diablo. A constant theme in previews and reviews of Diablo II was that we didn't change anything; it was more of the same. At first that struck us as odd. We kept less than one percent of the code and art from the first game. We rewrote the graphics engine, changed all the character classes and skills, shifted and expanded the setting, reworked and added to the magic items, brought back only a handful of our favorite monsters, and designed a ton of new gameplay elements, such as running, hirelings, left-click skills, and random unique monsters. Why, then, did everyone think it was the same thing? In the end, we decided just to take it as a compliment. The play-testers and reviewers meant they were having exactly the same kind of fun that they had in the original game.

Both Diablo and Diablo II provide a constant source of simple pleasures, many of which are perhaps too basic and obvious to mention in evaluations and reviews, but which are fundamental to their success. We used the term "kill/reward" to describe our basic gameplay. Players continually kill monsters and get rewarded with treasure and experience. But the rewards don't stop there. We offer a steady stream of goals and accomplishments to entice the player to keep playing. There's always a quest that is almost finished, a waypoint almost reached, an experience level almost achieved, and a dungeon nearly cleared out. On a smaller scale, we tried to make every single action fun. Moving around inventory items produces pleasing sounds. Monsters die in spectacular fashion, like piñatas exploding in a shower of goodies. We strove for overkill in this sense, in that players are constantly on the verge of something great - only a few mouse-clicks away from a dozen interesting things.

Diablo II retained Diablo's randomly generated levels, monsters, and treasure. This obviously allows for better replay potential, but also serves to make each player's game his or her own. Players feel an ownership of their own game experience in that they are actively generating a unique story. It's enjoyable to tell friends about what you have just done in the game, knowing for sure that they have not done the same thing. Simply following an online walk-through won't help them accomplish goals without effort.

Finally, Diablo and Diablo II are easy to play. We used what we call the "Mom test": could Mom figure this out without reading a manual? If we see new players struggling with how to sell items, we look at how they're trying to do it and make that way work too. We strove to make the interface as transparent as possible. You want to open a door? Left-click on it. Want to move to a target location? Left-click on it. Want to attack a monster, pick up an item, or talk to a non-player character? Well, you get the idea. It's amazing how many games have different controls and key combination for all these actions when simpler is always better.

The architecture in Diablo combines aspects of many different cultures in order to arrive at an interesting mix that doesn't look too much like any single one. Here, the buildings of Travical from Act III are based on Mayan and Aztec references.

2. Blizzard's development process. Blizzard's development process is designed to ensure that we make a great game. While our goal is to meet the milestones we set, our process, in terms of design and business, is structured to allow us to wait until the game is as good as it can be before we ship it. We recognize that not all developers have this same opportunity, but many of the methods we use along the way are applicable to any development environment.

First, we make the game playable as soon as possible in the development process. Our initial priority was to get a guy moving around on the screen and hacking monsters. This is what players would be doing most of the time, and it had to be fun. We were constantly able to hone the controls, pathfinding, and feedback mechanisms during the entire length of the game's development. Most importantly, it allowed us to determine what was fun to do, so we could provide more of it, and discover what was awkward or boring, so we could modify or remove it. For instance, it became obvious very early that players would be killing large amounts of the same monsters, and those monsters would predominantly be attacking the players. This gave us the opportunity to plan for multiple death sound effects and additional attacking animations for each monster. If we hadn't experienced the core gameplay as early as we did, combat would have ended up feeling much more repetitive.

Also, we constantly reevaluate gameplay and features. Up until the very end, if we can make the game better we will, even if it means redoing big tasks. For instance, we decided that we didn't like the Bone Helmet graphics for the characters more than a year after having rendered them, but we went ahead and remade them, even though it took a couple of weeks and the collaboration of four artists. Only weeks away from scheduled beta testing, we scrapped our Act IV level layout schemes because they were just a bit too empty and similar. The last-minute fixes turned these levels into some of the best, befitting their climactic function. Diablo II took more than 40 people and over three years, essentially because we made two or three games and pared them down to the best one.

Another gigantic reason for our success is our open development process. We strive to hire people who love games, and we make games that we want to play. Every member of the team has input into all aspects of the game. Discussions around the halls and at lunch become the big ideas that shape the game. A programmer suggested to a designer the concept of gem-socketed, upgradeable weapons, which turned out to be a huge crowd-pleaser. A musician's dislike for the old frog-demon's animation inspired us to redo it. As a team, we don't have to wonder what our audience wants, because we are our audience. If we like the game we are making - especially if, after two years of playing it, we are not bored to death - the game is clearly going to be a winner.

3. Character skill tree. Our most revolutionary new idea was the character skill tree. For a character to attain more powerful skills, he or she must master prerequisite skills. The ability for characters to branch into different areas of the skill tree, and to choose a level of specialization in each skill along the way, provides truly unique characters.

TOP: The player characters have modular armor of three varieties, light, medium, and heavy, which were mixed and matched to provide more individualized character appearances. "Paper dolls" created on paper and in Photoshop allowed mixing and matching of different pieces of armor to see how they worked together on the Barbarian.

BOTTOM: The Barbarian, translated from the sketches into a full, high-polygon model. Each part of a character's armor (the head, the torso, the legs, each arm, a weapon, and a shield) was rendered separately with in-house tools.

At the start of development, we planned to use the model from the original Diablo: characters would find and read books to learn spells and skills. Unlike Diablo, which had 28 spells shared by all three characters, we wanted to create a separate group of 16 skills for each of our five new character classes. This would definitely have been an improvement, but every character of a given class would still end up knowing all the same skills as other members of their class. Another problem was that players would likely be finding spell books for other character classes much more often than for their own. The skill tree solved these problems. The general idea was taken from the tech trees many strategy games employ. In strategy games, players advance by researching new technologies, which in turn open up further avenues of research. We adapted this to have our characters advance by choosing a new skill or strengthening an old skill every time they gain an experience level. Characters can generalize by choosing a wide variety of skills, or specialize by allocating many skill choices into a small group of skills. We also created a strategy element of choosing skills you might not use, just so you can get to one further up the tree later.

The end result of the skill tree is that one player can develop a Necromancer who kills monsters with a powerful poison dagger skill augmented by curses that cause monsters to fight each other, while his friend's Necromancer will summon hordes of skeletons to fight for him, and doesn't use any curses at all. The longevity of Diablo II will be enhanced by the endless strategies that can be debated and experimented with.

4. Quality assurance. The task of testing a game of Diablo II's scope, with its huge degree of randomness and its nearly infinite character skill and equipment paths, required a Herculean effort. We found we could not play-balance the climactic fight against Diablo without actually playing the entire game up to that point, because we could not predict what kinds of equipment a character might have, or what path through the skill tree he or she may have followed. This meant 20 or 30 hours of play for all the different characters, with a good variety of skill sets and equipment for each. Whenever we changed the game's treasure spawn rate or experience curve, we had to test it all again. Further complicating matters were multiplayer and difficulty-mode balance. Would a party of five Paladins, each using a different defensive aura, be untouchable? After more than 100 hours of play, is a fire-based Sorceress unable to continue in "Hell mode"?

The QA team created a web-based bug-reporting database through which we categorized and tracked all bugs, balance issues, and gameplay suggestions. In the end, this list delineated more than 8,300 issues and suggestions. Well-organized teams of testers concentrated on different aspects of the game, divided into groups that would specifically test character skills, item functionality, monster types, and spawn rates, or explore the countless variations found in the random level generation system. The members of the QA team became very good players and astute observers of the progress of the game. Everything worked much more smoothly than our experiences with the original Diablo.

5. Simultaneous worldwide release. In the past, Blizzard's strategy for shipping its game has been to get games on North American retailers' shelves as quickly as possible after the English version of the game went gold. With the original Diablo, we created our gold master on December 26, and some stores had it on the shelves by the 30th. Since Diablo was released, the percentage of international customers had increased substantially, and with Diablo II, we fully expect more than half of our sales to come from outside North America. With such a large number of customers located outside the United States, for Diablo II we decided that there would be significant advantages to coordinating the U.S. release to coincide with the rest of the world, not only to build anticipation for the product, but for the benefit and satisfaction of our customers as well.

Characters and monsters, such as this Vampire, were created in 3D Studio Max. An in-house tool would render the files from many different angles (eight for all monsters, 16 for player characters), and export them in the file formats used in the game.

If we release a game in the United States first, customers in the rest of the world don't want to wait a few months while we translate and localize it for their country. Due in part to the international climate fostered by the Internet, players around the world all know about the game at the same time and want to get it while it's hot. They might buy the U.S. version under the table or search out a pirated copy. Worse, they might lose interest by the time we release a localized version. Diablo II's simultaneous worldwide release also allowed our marketing and PR departments to focus their efforts toward creating a frenzy of interest for the first week of sales. Although the simultaneous release was a logistical headache, it was all worth it in light of Diablo II's superb success.

Article Start Previous Page 2 of 3 Next

Related Jobs

Airship Syndicate
Airship Syndicate — Austin, Texas, United States

Gameplay Programmer
Airship Syndicate
Airship Syndicate — Austin, Texas, United States

Senior Gameplay Programmer
Sucker Punch Productions
Sucker Punch Productions — Bellevue, Washington, United States

innogames — Hamburg, Germany

PHP Game Developer - Grepolis

Loading Comments

loader image