A theory of puzzles...

[Technocrat]

Being the sort with an academic bearing, the combination of reductivist philosophies and adventure games has produced a notion in my head - that most (if not all) puzzles in adventures can be boiled down to several essential concepts.

For example -
- substitution: these puzzles require the use of an object in place of another, where their use might be unusual. Can't find  pen to write with? Use a stick. Don't have a fishing rod? Combine the car-aerial with the string.
- verbal: using conversation with characters to unlock progress.
- pixel hunt: false impression of challenge created by simply making objects hard to find.
- physical progress: puzzles requiring adaption of environmental or inventory items to enable progress to the next room. Door locked? Kick it, find the key, etc. Can't get up the cliff? Make the rope out of blankets, etc.
- external knowledge: where you need at least a degree of knowledge from the real world, or something you've learnt in game. What was that combination the monk told you? What temperature does the severed hand need to be for the machine to think it's real?

I think there are a few more categories that would cover most kinds of puzzle. Do you have any ideas what they may be? Or do you feel I'm being a bit *too* reductive?

[Hernald]

There are a couple of lists of puzzle types out there. Classifying things, or pigeonholing them is something a lot of people like to do, but I am not sure how useful it is when coming up with specific puzzles for your game. As a hindsight excersize you might find you favour certain sorts of puzzles over others, so it would be useful for describing a game. The lists I mention come with opinions about some puzzle types; the writers don't like pixel hunts or mazes, but personally I have found both of these can be enjoyable if done well. So I think they are okay as a critical tool, but maybe not a creative one.

This is the shorter list with a bit of the analysis you are talking about, It's by Scarpio I think. Maybe you can match some of it up with your categories.

Quid Pro Quo/Exchange Puzzles.
Inventory/Combination Puzzles.
Timing Puzzles.
Distract-n-Grab Puzzles.
Maze Puzzles.
Escape Puzzles.
Cryptogram Puzzles.
Memory-based Sequence Puzzles.
Logic Sequence/Device Puzzles.
Repeated-Action Puzzles.
Dialogue Puzzles.
Forced Dialogue Puzzles.  
Riddles and Logic Puzzles.
GUI/Board Puzzles.
Dead Ends, Red Herrings and Faux Puzzles.

[GarageGothic]

Personally I find that thinking within the existing frameworks of what puzzles traditionally have been  like is very restrictive. As Hernald says, there are several lists of existing puzzle types out there (many linked to in this Adventure Gamers forum thread).

However, I think if you want to create something a bit more original you should forget everything you know about games and rather look into the psychology of learning and creativity. Because this is where we find explanations for why we as humans enjoy the mental challenge of puzzles, and also very good reasons for why we - as old-school adventure gamers - tend to become jaded with age, and no longer respond in the same way to solving puzzles. Because the "been there done that" feeling upon solving yet another item-combination puzzle is entirely linked to neurochemistry.

Learning new behavior, learning how to adapt to new environments and challenges, is a fundamental aspect of being human - it's a fundamental aspect of evolution too, to try out new ways of surviving and adapting to our current situation. To encourage this, nature has come up with the wonderful incitement of releasing dopamine into our bodies whenever we find a better, more efficient way of solving a problem. This is the function of the brain we as adventure game designers hijack, to give people pleasure by finding solutions to imaginary problems. We want them to have a eureka moment, we want them to feel the joy of insight, of learning, we want to flood their brains with dopamine when they figure out the "optimal" solution to a problem.

However, as we gain more experience in our lives, as we learn how to cope with specific situations (real or imaginary), our bodies stop rewarding us because we're no longer learning. We are just remembering how we solved similar problems before and applying the same logic to the "new" situation. This is why someone who's just discovered adventure games might find Gray Matter or Black Mirror 2 a pleasurable experience, whilst those of us who grew up with the genre will find them rather mediocre. Those games simply aren't different enough from what we played before to trigger that surge of neurotransmitters associated with learning.

So yeah, instead of looking into how it's been done before, I think we would be much better off investigating new areas that haven't yet been explored in games. Watch some TED talks about creativity, education, motivation, read books on psychology and neuroscience, oh, and talk to kids - they know a lot better what is fun than we do as adults, and they don't yet have the filter of culture to tell them what's the "right" approach to a problem. They'll try everything, and their solutions to problems may very well be better than yours, even when they seem like nonsense to an "educated" adult. Kids always think outside the box, because they haven't constructed the box to think within yet.

Edit: As for your categories, I think they're pretty spot-on. Anything we do in (good) adventure games is pretty much a mirror of real-world problems, and indeed, making tools/finding alternative usage for objects is very close to what we're doing all the time, as is communication and locating useful information/items. This type of goals are at the center of human existence, so learning them in a simulated environment is immensely useful. What we can change, however, is the "how" and the "why" of solving these problems, we can make them more intuitive, more meaningful, easier to relate to everyday experience, and this, to me, is the real challenge of game design - how to teach the player skills that they can use outside of the game.

[LUniqueDan]

If you like it the reductivist way, there's olny ONE type of puzzle in adventure games :

USE KEY WITH DOOR

All the other puzzle variations have the same logical equivalency.

[Ali]

I thing what GG has to say is quite interesting. I also agree that these kind of categories may be more useful for deconstruction than construction.

There is one thing I don't really agree with:

- pixel hunt: false impression of challenge created by simply making objects hard to find.

Searching for things can be lots of fun. In platformers it often involves exploration and physical challenges. It's occasionally good in adventure games too, as long as the player is aware of what they're looking for (or at least knows that there is something to find). I think the term 'pixel hunt' is most appropriate when there is a hard-to-find object/hotspot/location which the player is unaware of until they find it.

[Anian]

I never thought of pixel hunting as an actual puzzle type, it's more of a bad design choice imo.
Also, verbal/dialog puzzles might also be considered as combination puzzles, it's pretty much the same as inventory based puzzles but kind of limited (usually there's less dialog options), although it's very well masked that it's basically the same gameplay, although they can be dealt with trial and error usually.

But yeah, seprating puzzles by type is really restricitve, although it may be faster in some cases. It would probably be good to set up a plot and plot points where potential puzzles should be, then see what kind of an enviroment of puzzle there is, like you're in a sewer and so you make a maze or at reception desk and you dialog your way into a builidng etc., something that feels natural in that place, and make a few options for each.
Then you look at what types of puzzles those are and plan out what game progression a player might do (unless it's very linear gameplay) - based on that you change the variation of puzzles so there's not like 4 mazes in a row or something.
For extra consideration you look at different acts in your plot and reuse the systems (what you mentioned as "external knowledge" type) as a kind of a higher level of thinking. As in in act I you learn how to play chess, but in act III you have to play a 3d chess (stupid example, but I hope it get's the point across) - that way you basically keep the same puzzles but the player still feels challenged and feels the difference, even if it's not that directly obvious.