Guess what time it is! GAME DESIGN TALK TIME!
And it's happening because I'm stuck at work.

Who here wants to talk about Adventure game design?! :D

Adventure game design is an interesting little bastard of design. At its core, it's basically a thematic grouping of puzzles with intermediary media connecting them all together.

Really well done adventure games reinforce the world with good elements, and help bring you into that world that the game exists in. Myst, Quest For Glory, and Day of the Tentacle are my personal favorite examples of these: Some are just puzzles, others toss other elements in.

Poorly done adventure games have puzzles that have a 'doesn't follow' logic or actively hide the solution. Three good examples are the mustache puzzle from Gabriel Knight 3, the Yeti and Cream Pie puzzle from King's Quest 4(?), and the Fan Switch in Gabriel Knight.

This sounds really simple, right? Just make good puzzles.

Except, what -is- a good puzzle, and what is an adequate solution to a puzzle?

Basically, it's a puzzle that is fun and adequately rewards the player for their effort.

You can find great puzzle design in Portal: They go to extreme efforts to adequately teach you how to solve their puzzles, and provides excellent space with different versions of solutions.

With Myst, you can break down a puzzle different ways and mostly get the answers you need. You can get to the solution different ways.

With some other adventure games, you cannot solve the puzzle with reasonable trial and error.
You basically have to bruteforce the solution, which isn't...great...design. Especially if you must reload because your solution breaks another puzzle later.

Solving is not the same thing as bruteforcing in game design, imho.
With bruteforce, you are forcing the solution with no careful analysis of the consequences to an action.
You aren't solving a puzzle, just jamming a key different ways.

You need to give the player information, and make that information understandable enough to be processed and used later (Gabriel Knight's Fan Switch puzzle versus Myst Journals)
So, what is good Adventure game design?

Good Design has:
1) Puzzles that are solvable and adequately reward the player for their effort. (Example, Myst 3 Amateria Roller Coaster is a good example).
2) Puzzles the reinforce the story and world you are in (Quest for Glory 2: W.I.T. puzzles are a great example)

3) Puzzles that have logic 'which follows' the theme.
(Day of the Tentacle, Toonstruck)
4) Adequate cause and effect that is observable and learnable.
5) Possible external information that will help in the solving of the puzzle

Bad Design:
1) Puzzles that don't fit, don't follow the world logic, and break the flow of game (Such as using the wrong item to solve one puzzle, making another puzzle unsolvable)

2) Actively hostile with clues or parsable information (Pixel hunting BOOOO)

3) Puzzles that exist for the sake of existing. Don't bloat your game time with weird ass puzzles. You'll just piss off players.

These are by no means complete guidelines! These are just little examples I've found!

Good design means going out and finding good and bad design decisions and figuring out why they do and don't work, and why you do or don't like them.

Then experimenting with and tweaking with them!