Fixing a Puzzle

Getting started on this game was a bit of a struggle. I had greatly under-estimated how much I had forgotten about using the game engine and its map creation tool. Making the first parts of the game, the ones that make up the demo, took a lot longer than they would now I'm back up to speed. Who knew that much rust could build up over 8 years, eh?

In the excitement of the possibilities, I wanted to re-use a puzzle type I'd used in both Lost Temple and Iron Tower. It's a switch door puzzle, where each switch toggles the open/closed state of one or more doors. The puzzle is, of course, to toggle the doors into such a state that free travel is possible. Some switches can only be accessed with the doors in a particular state, etc.. Although it's possible to solve this puzzle through observation and logic, in practice this puzzle is also amenable to random experimentation. That's not necessarily a bad thing, but I wanted to make a version with more switches and doors, where it would be more laborious to brute force, to encourage a logical approach.

I'd already settled on having a big mansion as the 'home' area, and having the doors connecting the rooms be the puzzle seemed like a splendid idea. However! This then means that the doors are spread out over a number of rooms, so there's no immediate feedback to the player on what the switch they've just thrown actually does. I had planned for there to be a room where you could see the status of the doors, so you could use character switching to review what had changed each time you throw a switch. Again, however! If multiple doors change when a switch is thrown, it's quite opaque which status light corresponds to which door. It also turned out that where I had multiple doors between two rooms, that caused confusion. I knew that they were essentially a single door, and their status was linked, but it's not obvious to the player. What proved worse was that all these problems were exacerbated by the larger size of the puzzle.

It's not actually necessary to 'solve' this puzzle to complete the demo, and neither will it be necessary to solve it to complete the full game. I can definitely improve it, though.

Firstly, I've labelled the doors, and the status lights, so the correspondence is clear. The paired doors have the same label, and also two status lights above each other, to hopefully make it even clearer what's going on. I've moved an item into the room with the status lights, so the player encounters that area earlier on in the puzzle sequence, to draw their attention to it. The item you use to make the status board work is also encountered earlier on, rather than halfway through solving the puzzle. Another step would be to label the switches, so you know which doors will toggle, but I'm hoping that won't be necessary.

The lesson to learn here is that it's better to put all the parts of this kind of puzzle where the player can see them unfold, even if the logic of it seems obvious to you. Fortunately, I haven't used this puzzle type anywhere in the 3 realms I've finished so far, so I don't have to change anything else. A quick review of the puzzle premises I've used seems to indicate that they're not prone to this problem (although, of course, all puzzles have the possibility of leading the player up a garden path of misunderstanding).

What I can do to harden them against this kind of issue is provide explicit feedback - Dizzy games were notorious for providing no feedback if you tried to solve a puzzle the wrong way. If you tried to use an item and dropped it instead, all that told you was that you hadn't solved the puzzle. You had no way of knowing if it was the wrong item, if you weren't quite in the right place, if you needed to do something else first, or if you were barking up entirely the wrong tree. The more hints I put in, the better it will be.

The latest version of the demo has these improvements.