This is a great article about integrating puzzles into action and/or adventure games. As a fan of very puzzle-y games, I do take exception to a couple of points.
The first idea to jump out at me in this article was the following principle: “Limit the investigation area around the puzzle.” Generally I would agree with this principle – it just makes sense to keep puzzles local, in order to build up player trust and keep them from doing a ton of backtracking. Myst games typically violate this rule a bit, and especially in metapuzzles, but I think it works because in a Myst game travel has a very low cost – indeed, it’s usually rewarding instead of risky.
Riven strikes a very nice balance by keeping most puzzles and their clues confined to the same island (or the same CD, in the original format); there’s a clear boundary of how far you should travel looking for a solution. But my favorite puzzles from that game, and the puzzles that are most closely tied to the narrative, completely violate this rule. The large golden dome is tied to smaller golden domes, which are tied to symbols, which are tied to colors, which are tied to numbers, which are tied to animals and their sounds, and these are all scattered all over the islands. Likewise, there are a handful of puzzles that cannot be solved until certain information is collected from a character’s journal in another part of the world. That Riven knows these are the exception to the rule, and uses that to give them more narrative import, is incredible to me.
Another rule that Myst games violate is “Avoid drowning the player in clues.” Myst walks a fine line here. A key player skill is figuring out what information is a clue and what is background information. Since the background information is so important to Myst as well, they don’t hold it back. Instead they go out of their way to point out the real clues in more subtle ways, using color, light, placement, resemblance and proximity to help the player. For example, within a character’s journal there is a lot of backstory, but a real clue will often be inked in a different color or a different hand, as if it was added later (or sometimes even in a different language, so it stands out). Thus, learning to play Myst is about learning a subtle language and reading the conversation between the player and the designer, understanding the reasons for every decision the designer made.
As a designer, I can see why sticking to these rules is important. But I think for a game to be known for its puzzles, there should be at least one puzzle that breaks a rule or two. Otherwise don’t the players always know what to expect?