[discussion] Age & Puzzle Design Testing.

General debates and discussion about the Guild of Writers and Age creation

[discussion] Age & Puzzle Design Testing.

Postby Aloys » Sun Apr 26, 2020 7:34 pm

(I had started talking about this in another unrelated thread, so instead of going off-topic there I'll copy/pasta it here)

For all the talk about building Ages on a technical level, we don't talk much about the techniques of designing what goes in those Ages; and in particular puzzles. And one the key part of that is to be able to test your gameplay. When it comes to puzzles all people will react differently, some will miss information, some will understand the puzzle but try to solve it the 'wrong' way, some will try things you will never have thought possible etc.. Which is why it is so important to test.

Now we don't exactly have a lot of Ages with puzzles right now. But part of that may be because of the design/testing workflow. Maybe we can improve that? :)

:?: How do you test your puzzles? Do you test them at all? If not why?

Because it is long and complicated to implement a puzzle (python, SDL, visuals etc) it is better to test a puzzle out of the game. On paper. It is easier, can be done properly. It saves lots of time, and will lead to better puzzles. Take some pen and paper to draw stuff, a few random items to represent your puzzle items, and show it to people. Like an pen and paper roleplaying game.

Because again the best way to do that is with other people. You should never validate a puzzle just by yourself. We need outside perspective. :arrow: The problem we (the GoW) have here is that most of us are designing their stuff alone, 'testing' it alone. (I know: I do. :( ) And part of why we do that is because we want our puzzles to be surprises, we don't want to spoil them, and so it's hard to just talk about them and ask for opinions... But we should. How do we do that?
User avatar
Aloys
 
Posts: 1968
Joined: Sun Oct 21, 2007 7:57 pm
Location: France (GMT +1)

Re: [discussion] Age & Puzzle Design Testing.

Postby Sirius » Mon Apr 27, 2020 4:37 am

I think the lack of resources about how logic works in the game is the first blocker when it comes to puzzle quality. But I agree bugtesting is going to be really important as soon as people start getting the hang of nodetrees and Python/SDL... There are a LOT of things that can go wrong, especially when you start putting in multiplayer.

Well, back in the days... Bugtesting was what the Guild of Maintainers was for. Unfortunately, there aren't enough people nowadays, the maintainers would also be the players :lol: (ha, this might even have been the case back when Uru was more popular...)

At my work, we sometime ship untested applications. But that's because our clients are always in a rush, and they know THEY are going to be the testers, and WE are going to be on the phone with them until the bug is fixed. (And we do our best to keep bugs out in the first place...) In the case of fanmade content, this is problematic because people might stay on the first impression they get of an Age, and think it's always going to be a "buggy mess", and have no idea how to report bugs they find. (Wait, Uru is a multiplayer MMO and doesn't have a bug report dialog ? /sarcasm)

I think Korovev's Fan Age Expeditions can help solve that. The Age is stress-tested with many people in multiplayer. People are less likely to stop playing at the first bug they encounter. And then, all bugs that were found can be reported to the Age author. This is really useful, but still require "spoiling" your Age before it's bug free.

Another solution (although far from perfect) is to "unskin" your puzzle. Make a copy of your Age, remove the visuals, sounds etc that are the "skin" of your Age, and leave only buttons and doors (and a low-poly untextured version of your meshes) which are the "bones" of your Age. You could send this for testing and let people try it out. It still spoils the puzzle itself, but is not too much work and should help a lot with debugging puzzles.
(You could even use Korman's "page" system to have both the skinned and unskinned versions coexist in the same Blender file, but I think even I would not attempt it :lol: )

Aloys wrote:Like an pen and paper roleplaying game.

Cyan actually did roleplay puzzles when developing some parts of Uru. Sounds like a good idea.
User avatar
Sirius
 
Posts: 1506
Joined: Mon Jul 26, 2010 4:46 am
Location: France

Re: [discussion] Age & Puzzle Design Testing.

Postby Semjay » Wed Apr 29, 2020 2:31 am

You are also limited by what the avatar is able to do, don't you think?

You might think of a good puzzle but then find it can't be done. There is no inventory, so the player can't pick things up. The player can't manipulate objects to solve jigsaw or peg board puzzles. You can't push things into place very easily unless you kick them. So sliding blocks puzzles are not possible. You can't type things in so that means no key word or password based locks.

What can you do? I've been trying to think... Some very simple things can be "dressed up" and appear more complex than they really are. The obvious one is that the player can click on things or click in a set order to get a response (eg, Journey cloths, Eder Gira vents puzzle or Kadish vault puzzle). You can walk into and out of volumes to trigger events (eg, the mazes in Kadish, I think that's what happens there). There are those telescope focusing puzzles but I'm not sure how those work.

What else can you think of? What puzzles have you seen that you really liked? And can you explain how they were created!?
User avatar
Semjay
 
Posts: 60
Joined: Tue Nov 03, 2015 6:52 am
Location: UK

Re: [discussion] Age & Puzzle Design Testing.

Postby Sirius » Wed Apr 29, 2020 8:10 am

Semjay wrote:You are also limited by what the avatar is able to do, don't you think?
You might think of a good puzzle but then find it can't be done.

This is definitely good advice. Avatars can't do much besides running, pressing buttons and climbing ladders. Always assume they have only one arm, with no opposable thumb and wearing a boxing glove :lol:
And some puzzles are just too hard to implement properly anyway. Don't design your whole Age/story around a puzzle, you risk wasting a lot of time. Think your Age first, add puzzles later (you will probably get ideas while building anyway).


Semjay wrote:What can you do? I've been trying to think... Some very simple things can be "dressed up" and appear more complex than they really are.

Due to the limited interactions, puzzles are usually just button pushing anyway. So yeah, I would say properly dressing them up to hide the solution in plain sight is the best thing to do. Especially when you can seamlessly integrate them in the Age design (easy to say but hard to do), add some red herrings that make sense being there, etc. Riven is the reference.
After spending entirely too much time trying to make Ages (and not releasing a single complete one), I can say puzzles are going to be a pain to design and wire anyway. Properly dressing up a simple puzzle is usually better than half-assing a complex puzzle, IMHO.


Semjay wrote:What else can you think of? What puzzles have you seen that you really liked? And can you explain how they were created!?

I don't have a lot of suggestions for this topic. This may sound weird, but I dislike puzzles. I don't know, I view them more as an artificial lock than a proper challenge for your skills or creativity. And they have no replayability.
Uru's (and EoA's) puzzles are fairly meh overall. Cyan tried a lot of ideas to work around the limited interactivity, to the point that most puzzles are hardly puzzles anymore. I don't dislike it though - most puzzles aren't really challenging, but at least they make the exploration fun.

I play Myst mostly for the environments. Nowadays when I build things in Blender, I focus more on how to make the environment itself a traversal puzzle, a bit like Dark Souls does. Chain narrow corridors with wider open areas. Go up then down. Walk across narrow paths where there is a risk of falling. Branch the path and let the player choose which one he will take. Reconnect pathes together. And secret passages ! Etc. Obviously this doesn't include being ambushed by zombies or walking across deadly traps :lol: but at least it makes walking around more interesting, especially if you manage to provide the player with some nice sights.
User avatar
Sirius
 
Posts: 1506
Joined: Mon Jul 26, 2010 4:46 am
Location: France

Re: [discussion] Age & Puzzle Design Testing.

Postby Tsar Hoikas » Wed Apr 29, 2020 9:02 am

Semjay wrote:the player can't pick things up. The player can't manipulate objects to solve jigsaw or peg board puzzles. You can't push things into place very easily unless you kick them. So sliding blocks puzzles are not possible. You can't type things in so that means no key word or password based locks.

All of these things can be done in Uru... even in the legacy Path of the Shell game.
Image
Tsar Hoikas
Councilor of Technical Direction
 
Posts: 2180
Joined: Fri Nov 16, 2007 9:45 pm
Location: South Georgia

Re: [discussion] Age & Puzzle Design Testing.

Postby Sirius » Sun May 03, 2020 3:32 pm

That's true. With some creativity and some Python scripting there is a lot you can do.
With that said lots of people are not comfortable enough with Python scripting to setup some of those interactions. So when in doubt I recommend starting with something simpler, and improve it later. There are often multiple ways to create something. For instance, IIRC Zahlto's pickable keys are simply well decorated toggle buttons. It has a few downsides, but this still works well enough for the Age's purpose.
User avatar
Sirius
 
Posts: 1506
Joined: Mon Jul 26, 2010 4:46 am
Location: France


Return to General Discussion

Who is online

Users browsing this forum: No registered users and 0 guests

cron