The Game Design Forum

7 Lessons & Citations

If you're looking to get the print or eBook versions of this text, which have an entire extra chapter, you can find those on the books page.

Lessons from Diablo 2

At the end of every Reverse Design, I summarize the major lessons of the book. I want to be clear that I do not believe that the lessons of videogame design can be broken down into rules of thumb. One of the overarching themes in this series has been that execution is far more important than theoretical ideas. The point of the Reverse Design series is to provide extensive detail about how the great games executed their ideas, in particular detail, from beginning to end. But inasmuch as books like these do need endings, a summary of the lessons seems useful enough.

(1) A great variety of videogames, Diablo 2 included, mix ideas from different genres that came before them. Diablo 2 borrows from roguelikes, traditional tabletop RPGs, and console action games. Sometimes it reuses an idea wholesale: maps and item drops in Diablo 2 are very close to those in their immediate ancestors, Moria and Angband. Some ideas are modified: permadeath is optional, but all players lose experience and gold when they die. The same is true with character classes; every class plays differently, but they don’t follow the traditional roles of the tabletop. In other words, Diablo 2 picks and chooses which ideas will work for it, which won’t work, and which just need a little adaptation.

(2) There are many different ways to look at randomness in game design. All randomness might be mathematically similar, but most players (and the designers who seek to entertain them) don’t perceive the complex mathematics. Instead, it’s useful to use different lenses to look at randomness from a psychological perspective. Different systems in Diablo operate by set randomness, randomness within a range, deck-of-cards randomness, and slice-of-pie randomness. If you’re implementing a procedural system in your game, try prototyping it by using a deck of cards, or a pie graph, or a bag of Scrabble tiles to which you assign special values.

(3) Remember that if you’re going to rely heavily on randomness through procedural generation, the game will have a very “copy/paste” feel if you don’t have enough cards in the deck or tiles in the bag. Remember that the component presets in the procedurally generated maps of Diablo 2 required the full time work of three people for several years, in order to feel as “random” as they are.

(4) The fundamental design structure of console action games, Nishikado motion, has been amazingly durable across the history of videogames. Even in genres where it isn’t necessary (like stat-heavy RPGs), it can be a useful way to keep your game from being too easy or too frustrating. But remember that no game is ever perfectly balanced!

(5) If you’re trying to emulate Nishikado motion, your game is going to need gear and enemies which are sometimes inappropriately strong for the area where they appear.

(6) If you want help balancing your gear, try using LEQ. It’s not a perfect metric; there is no perfect metric! But if you’re designing an RPG, it’s useful to think of all power in the context of character levels, since you’re going to have them in the game anyway.

(7) Anticipate the possibility of accelerations, and use them in the right places in your game. The prevailing wisdom about RPGs is that the need for grinding in a game will repel large parts of the audience. Diablo 2 relies heavily on players repeating the same content over and over in order to farm EXP and gear, and yet it is one of the most popular RPGs of all time. That’s because acceleration flow turns otherwise boring repetition into an exciting process of exponential growth.

Special Thanks

As always, I’d like to thank the Kickstarter backers who made Reverse Design: Volume 1 possible. Without their support, this project would not have happened. I would also like to give a big hearty thanks to Diablo 2 developers David Brevik, Max Schaefer, Stieg Hedlund, and Stefan Scandizzo, who were all very helpful and gracious. Finally, I’d like to thank the book’s editors, Amanda Lange and Nicole Kline.


  1. Gygax (Player's Handbook)
  2. Gygax (Monster Manual)
  3. Gygax (DM's guide)
  4. Scandizzo
  5. Brevik
  6. ibid
  7. Schaefer
  8. The Arreat Summit: Classes
  9. Clarke
  10. ibid
  11. ibid
  12. ibid
  13. Diablo Wiki: Blocking
  14. ibid
  15. The Arreat Summit: Skills
  16. ibid
  17. Arreat monsters
  18. The Arreat Summit: Skills
  19. Arreat Weapons
  20. The Arreat Summit: Skills
  21. ibid
  22. Diablo 2 Attack Speed Calculator
  23. Brevik
  24. The Arreat Summit: Skills
  25. Vertical Slice
  26. Hedlund
  27. Brevik
  28. Diablo Wiki: Move Speed
  29. Brevik
  30. Hedlund
  31. Diablo Wiki: Affixes
  32. The Arreat Summit: Monsters
  33. ibid
  34. Arreat weapons
  35. Arreat weapons
  36. Scandizzo
  37. Brevik
  38. Hedlund
  39. Diabloiinet: Nodrop Rates
  40. Diablo Wiki: Treasure Classes
  41. Fenriradramelk
  42. Diablo Wiki: Treasure Classes
  43. ibid
  44. Brevik
  45. Diablo Wiki: Magic Find Diminishing Returns
  46. Diablo Wiki: Magic Find Percentages
  47. Scandizzo
  48. ibid
  49. ibid
  50. ibid
  51. Hedlund
  52. ibid
  53. Scandizzo
  54. Brevik
  55. Diablo Wiki: Area Level
  56. The Arreat Summit: Monsters
  57. ibid
  58. The Arreat Summit: Experience
  59. The Arreat Summit: Monsters
  60. The Arreat Summit: Items
  61. Ibid
  62. Diablo Wiki: Affixes
  63. The Arreat Summit: Skills
  64. Chris Clarke
  65. The Arreat Summit: Items
  66. Diablo Wiki: Affixes
  67. The Arreat Summit: Monsters
  68. Chris Clarke
  69. The Arreat Summit: Characters
  70. Diablo Wiki: Affixes
  71. The Arreat Summit: Monsters
  72. Ibid
  73. The Arreat Summit: Items
  74. Schaefer
  75. Brevik
  76. Hedlund
  77. Brevik
  78. ibid

Works Cited

Bibliographic Note: Three out of the four interviews done for this book took place during the writing of it, in 2017. The interview with Max Schaefer took place much earlier, in 2012, as part of another project. His comments during that interview--specifically that he wanted game difficulty in Diablo 2 and Torchlight to operate like a "sine wave"--inspired the thesis of this book.

(1) Brevik, David. Interview, January 25, 2017.

(2) Clarke, Chris. “To Hit Calculations.”


(3) Diablo 2 Attack Speed Calculator.

(4) Diablo Wiki. “Magic Find Diminishing Returns.”

(5) Diablo Wiki. “Affixes.”

(6) Diablo Wiki. “Area Level.”

(7) Diablo Wiki. “Attack Speed.” (

(8) Diablo Wiki. “Blocking.”

(9) Diablo Wiki. “Magic Find Percentages.”

(10) Diablo Wiki. “Move Speed.”

(11) Diablo Wiki. “Treasure Classes 1.10.”

(12) Diablo Wiki. “Treasure Classes 33-60.”

(13) Diabloiinet. “Nodrop Tables.”

(14) Fenriradramelk. “Magic Find Mechanics Guide.”

(15) Gygax, Gary. Monster Manual. TSR Games, 1977

(16) Gygax, Gary. The Dungeon Master’s Guide. TSR Games, 1979

(17) Gygax, Gary. The Player’s Handbook. TSR Games, 1977

(18) Hedlund, Steig. Interview, March 7, 2017.

(19) Scandizzo, Stefan. Interview, March 4, 2017.

(20) Schaefer, Max. Interview, October 15, 2012.

(21) The Arreat Summit. “Experience.”

(22) The Arreat Summit. “Monsters.”

(23) The Arreat Summit. “Skills.”

(24) Vertical Slice. “S03E18: Graybeard Games' David Brevik.”

Schaefer Variation and Acceleration Flow - Back

Site questions?

All material copyright by The Game Design Forum 2017