Reverse Design: Diablo 2

by The Game Design Forum

The most recent entry in the Reverse Design series. This book examines all the design decisions that went into Diablo 2, from its roguelike roots to its traditional character classes and action game mechanics. This book takes as its thesis that Diablo 2 takes the fundamental design structure of action games, and recreates it in the context of a procedural RPG. In doing so, it gives fascinating examples of how to use different types of randomness, how to implement roguelike and classical RPG ideas in real time, and how to keep players entertained (for more than a decade!) when all they're really doing is grinding for levels and gear. It's a fascinating game, and a great help in learning about game design math and player psychology.

119 pages, ebook

Published September 1, 2017

Reviews

[Tim O'Hearn · Rating 4 out of 5]

While I found the concept fascinating, the book felt incomplete. The problem with covering such a classic game is that most people who care enough to pick up a book like this have played for at least one thousand hours, rendering many of these conclusions down to a matter of intuition. Here's what I liked: the book suggests roughly 3 overarching reasons for why D2 was so good and does a good job of explaining each. They are, roughly, 1. A unique blend of fundamental elements from different genres of gaming, 2. "The tension between randomness and familiarity", and 3. Terminal Acceleration (or, a fantastical endgame).

There was special attention afforded to explaining how D2 implemented randomness. Rather than highly variable probabilistic occurrences, D2 capitalized on carefully measured outcomes within a fixed range of possibilities. On top of that, there are many overlapping systems of randomness, from drop probability, to the gear's own stats, to map layout, to monster strength, and more. This worked well, and I can't help but think that one of the reasons it worked so well was that there were certain limitations on how much gear you could stuff into a game (and your loot chest) in 2001. In comparison, Borderlands 2 claimed to have something like eighteen million different weapons, which were all procedurally generated. Late game D2 had only several dozen different Unique, Set, and base items filled with Runewords (okay, rare and crafted items had a high degree of variability and were sometimes immensely powerful, but they weren't crucial to every build), but then had a multitude of modifiers adhering to the randomness mentioned above, so there was much more intimacy and charm when it came to procuring loot and building characters. I couldn't put it any better: The tension between randomness and familiarity.

The section on terminal acceleration was the best thesis I've ever read on why addicting endgames are so addicting. It gives the example of World of Warcraft, which is perfect, because I knew quite a few people who became significantly more dedicated to the game once they had effectively beaten it and reached the level cap. The significantly diminishing returns when honing in on certain statistical bonuses was useful, as was the graph showing how the game's difficulty fluctuated somewhat with character level and how much of an effect powerful gear could have on a player's experience

I would have like to have read about how Battle.net, the ladder system, and frequent patches also contributed to the game's sustained popularity. Also, an attempt to cover the in-game economy and trading mechanics would have gone an long way in helping establish a lasting case for the game's greatness. While somewhat disappointed, it's only because I'm a megafan. I look forward to reading future books in this series.