Wednesday, 1 September 2010

Puzzle Tweaking - Tale of a Wooden Stick

Once upon a time there was a wooden stick, a stick so firm, a stick so stuck it could not be seen upon without the touch of many small, busy hands.

And so this tale beings.

Today we released a new gameplay video. Keeping to the Amnesia tradition it is one single clip, showing a sequence from the game. This time it's about exploration and puzzle solving. It is the puzzle part that this blog post will discuss, as it took quite some time for us to get that puzzle right.

The puzzle is that the hatch is too heavy to lift by hand, but by using a crank and pulley contraption the player can open it. Unfortunately something is wrong with the crank/pulley and the player has to figure out what it is. It turns out that a small piece of wood has been jammed into one of the pulleys, making it impossible for the crank to pull the rope going through the pulley.

This particular piece of wood took quite a lot of tweaking in order to make it noticeable. For each test session we found that there was always one or more testers that got stuck on this puzzle. They simply could not find what was wrong, most important, they did not notice the wooden piece at all. Over and over we tweaked the puzzle, we added descriptions and interaction messages, we changed the texture of the wooden piece, we added some faint light to show it clearer in the dark, there are some particles coming down from the wooden piece when you pull the crank, we made the model larger and we changed the angle it is stuck in. We even added an alternative solution to the original solution (not saying what the solutions are, in-case you like to solve the puzzle in another way than what is shown in the video).

Finally, we hope, for the love of all puzzles, that this puzzle has been tweaked to perfection and that none of you will get stuck on it*.

*If you watch this video, it would be very strange if you did. :)


  1. Looks like another example of what Valve designers love to mention in their commentaries - it's incredibly hard to get players to look up in first-person games. For whatever reason, that's just not something that a lot of people do.
    Curiously, other people (such as me) look up frequently and don't get stuck on stuff like that (we get stuck on other stuff, of course).

    I guess that just goes to show how important it is to playtest with a large number of testers. People play differently, no?

    Oh, and it looks like you guys did a good job fixing up this puzzle. It's great to hear that you've put so much effort into everything!

  2. Did you consider adding some unrelated elements to the ceiling to make it more interesting to look at, instead of making the object larger?
    I still remember the ice ceiling in the office rooms at the beginning of Penumbra: Overture. I was gazing at it in awe; it was simply beautiful, especially after the dark and terrifying mines.

    But perhaps that's hard to pull off without making the puzzle too obvious.

  3. Interesting... I started thinking of alternative ways to solve this and I think I got it. Have to try if I'm right as soon as I get the game :)

  4. intgr:
    There is always a fine line between treating the player like an idiot and making things clear enough :)

    We thought of having some more interesting details to draw attention, but then the player might miss the wooden piece because the other detail was so interesting. We never really tried to though so I cannot say for sure that it would not have worked.

    There is also the problem of making the player realize that the wooden piece is significant and not just part of the backdrop. There is another puzzle a bit later in the game where we had a similar problem. Will be interesting to see if the final tweaks worked well enough once the game is released.

  5. @ Thomas & FG:
    QUOTE - Thomas said:
    "There is also the problem of making the player realize that the wooden piece is significant and not just part of the backdrop."

    However, I think I would know what to do to make it work, but it goes a bit against your engine.

    The eye is attracted to movement, so the addition of the dust particles was a right move, although I can't see them in the video. Another important point is: although I can see from the video that the cable/rope model changes or straitens after the player turns the crank, you must admit that the cable/rope is actually pretty static. Now, if you made it swing sideways, but only the section between the crank an the jammed pulley, I'm sure most players would notice where the problem is. However, since the player, in your engine, needs to *look at* the crank in order to turn it, he might miss both the dust particles and the swinging rope. An interesting approach would be to force the player to look at the hatch while operating (or trying to operate) the crank, which would put the puzzle-related cues on the upper edge of the screen - so he/she wouldn't be looking directly there. That would require some trickery related to how the player should perform a penumbra-like interaction while not looking at the object he/she is interacting with. Maybe helper screen area utilizing a render-to-sprite approach? A few problems with that though: (1) the puzzle could turn out to be too obvious if done that way, (2) it may or may not require some modifications to the engine, and (3) this method could go against the overall look and feel established for Amnesia.

    Another problem is, of course, that the game is done, so too late for experimentation now... But I wanted to say this for the sake of game-dev-related discussion - maybe it will be useful in a future game.

  6. P.S. How about posting an entry about what were the ideas your blog readers gave you that made an impact on the game as it was developing? I know a lot of people here had some very interesting suggestions and other general gaming-related things to say, so it would be nice to sum up what you tried, what worked out, what didn't...

  7. hehe.. And I can honestly say that I got stuck on this puzzle for a while when play-testing while doing the Mac and Linux ports. And I did find an "alternate" solution than what was presented here to solving it.

  8. I think the wood piece is a little awkwardly large. I mean, how realistic it is for such a large thing to be stuck in a place like that?

    About looking up. That is quite common, even in real life. Most ppl don't ever look/notice things that are above their line of sight. That's why surveillance cameras are always put above your head and not on the wall (where it would be easier to put them) - so you don't notice them ;-).

  9. I love reading this blog :) especially when im trying to become a game designer myself..! In my oppinion i would not make the stick glow like that. I know that you guys had several of testers getting stuck, but what the frigg! First of all it's a puzzle. There's nothing more than "The hatch" "The rope" "The lever" and "The pulleys". You can basicly go through every detail of the puzzle in less than 20 seconds, and find the problem. By this i personaly dont think you need the glow to the stick. Maby change the color a lil' bit so it doesnt blend with the background.

    But in all... I'm not the game designer here am I? no not even close :) But these are the stuffs i would like concider doing. Am I doing wrong by commenting it like this? In that case, let me know. Thanks for an awesome blog!

  10. Or you could just add Daniel's opinion.
    While trying to turn the wheel for the first time, let the player know what Daniel is thinking, for example: "Something must be blocking the rope".

  11. QUOTE (Anonymous [2 September 2010 12:36 ]):
    "... for example: "Something must be blocking the rope"."

    Now THAT would make it too easy.

    QUOTE (Cyber Killer):
    "I think the wood piece is a little awkwardly large. I mean, how realistic it is for such a large thing to be stuck in a place like that?"

    That's what I was thinking, but in game design you need to make some compromises. Anyway, a reasonable explanation would be that someone (or something?) found a wooden stake and jammed the damn thing for a purpose? But, coming up with what that purpose might be could prove to be a bit tricky - unless the game features some semi-intelligent creature that could manage to turn the crank, but if it wasn't working, it could never figure it out? But, then, why prevent the thing from going down through the hatch?

  12. QUOTE (Joakim Lundström):
    "I know that you guys had several of testers getting stuck, but what the frigg! First of all it's a puzzle. There's nothing more than "The hatch" "The rope" "The lever" and "The pulleys"."

    When designing puzzles, you must never look at things from your own (== developer's) perspective. What if the player assumes that the crank is jammed, and that the game requires an object to be found that could help turn the crank? Or that all the stuff is just decor, and that the road is blocked forever and that a new path needs to be found?
    I mean, only God knows what the player is thinking - some sort of subtle guidance is required.

  13. When I was watching the video, I assumed that, at least once the boxes were being stacked, the player would hold onto the rope and drop down. In retrospect, how would the player then get into the door without it closing back on them (unless the door swung open, no longer needing the pull of the rope)?

  14. Ah, I'm sure this is a familiar problem for any level builder, professional, modder and student (like me) alike. That balance between being obtuse and being blindingly obvious is always a trouble.

    Also just wanted to say that the lighting in this video looks fantastic.


Note: only a member of this blog may post a comment.