I’ve come up with 4 puzzles in my game so far just by creating a world with random objects and details and figuring out ways to use those objects in puzzles. But I’m kind of stuck now. I can come up with rooms, but I have no idea how to come up with puzzles. It’s easy to come up with rooms.
So what process do you use in your head to create puzzles? What do you come up with first? How do you suddenly find that you have a puzzle idea?
Well, if we just posted puzzles, they wouldn’t be very puzzling, now would they? 
Now the best puzzles are part of the plot–think about what the goal is, and then think of what impediments might logically and naturally arise.
However, a few standard puzzle types are not amiss in a game, so here are two lists:
http://tvtropes.org/pmwiki/pmwiki.php/Main/StockPuzzle
http://tvtropes.org/pmwiki/pmwiki.php/Main/StockVideoGamePuzzle
The current version of The Inform 7 Handbook (musicwords.net/if/i7hb.htm) has a whole chapter on types of puzzles. You might find that helpful.
To me, a good puzzle is one that requires some lateral thinking on the part of the player. For example, in “Lydia’s Heart” you need to enter a dark room, but your flashlight batteries are dead, and there are no fresh batteries anywhere in the game.
However, there’s a laptop computer nearby. When you switch it on, its screen will glow: It’s an effective portable light source.
There are many other ways to construct puzzles. That’s just one example. What I feel is not so good is to create a puzzle by describing an object badly. If the description was well written, the use of the object would be obvious, so the author tries to force the player to figure it out by forcing you to try to figure out what is being described. This is a legitimate technique if the object is one that the player character would not instantly recognize, but if you’re tempted to describe a screwdriver badly … just don’t, that’s all.
–JA
OK. Thanks for the advice.
I think I will create goals and then come up with obstacles.
What kind of game are you writing? If we knew the general plot, maybe we could suggest some puzzles?
I’m new to writing IF, and the style of mystery game I am writing is not really puzzle intensive, it’s more about logic and detection. That said, I did think of an interesting small obstacle for the prologue. I set a clear but simple goal for the player during the intro scene and then introduced a character that would basically get in the way of that goal; and there are several different solutions to get through this gateway to the next scene.
The puzzle really came from thinking about the characters. What would the player detective character encounter? What is the NPC character like? What would the NPC want from the detective character? I set the kind of small logical obstacle that would get in the way of the detective protagonist that would fit in a traditional narrative, like a short story.
Maybe this will spark some additional ideas for others. (And I will need to come up with more such ideas later on as I get to later scenes.)
I also find that sometimes coding in Inform7 is its own puzzle! 
–Zack
www.z-machine-matter.com
I especially like puzzles that are designed around a unifying core interaction concept: Enchanter’s magic spells, say, or the word puzzles in Ad Verbum, or the gadget-oriented puzzles in Spider & Web. Instead of having disparate puzzles that might not be related at all, you can teach the player to approach problems in your game in a specific way, and ramp up the difficulty gradually as you go – making the player feel that he’s mastering the system and giving him the satisfaction of feeling competent.
Having a core concept to your puzzles also makes invention easier, because you can brainstorm around the specific type of thing that you’re focusing on for your game, rather than starting from a wide open field.
If you’re interested, I talk about this idea a lot longer here: emshort.wordpress.com/2009/08/13 … g-infidel/
Some other resources that might be useful:
…ifwiki collection of craft articles: ifwiki.org/index.php/Craft#Puzzles
…index of discussions from the past: ifwiki.org/index.php/Past_raif_t … _1#Puzzles
There are all kinds of ways to come up with puzzles. The simplest puzzle is to require a logical action by the user. In “Philosopher’s Quest”, to solve non-existence you think. In Colossal Cave, to solve the lamp problem you buy new batteries. To pass demons in Zork, you’d use the bell, book and candle.
The next-simplest of the puzzles games is the logical maze. Level 9 games often used things like resistor colour codes to identify locations. This would be a tough one to do in the authoring systems I’ve seen, as I don’t know of any support for the virtual room concept. However, the notion of a maze that isn’t based on physical barriers but logical barriers is perfectly doable even on a classic system. An example of a logical maze might be a chessboard with passages linking rooms in the manner of a chessboard knight, where you may not revisit a room you have already been in and must do something in each room.
A good place to get ideas would be anything on the Internet about “lateral thinking” - a great source for inspiration. Mythology is also a good source - there’s often plenty of riddles, descriptions of challenges/tasks, etc.