(Continued from my previous posts.)
I’ve made enough progress for an update, although the game is still high on difficulty.
The cart format of the game, via Acorn Electron World.
I managed to clear out the “Jungle” first to the west, which turned out to be both more straightforward than I expected but also more irritating. Dropped items disappear into the undergrowth, but I figured I would try a bit of wandering anyway to see if I could get something to happen at random:
A few steps in I realized the room descriptions were different, and there are in fact four different rooms with very similar text.
Room 1: You’re in a jungle with exits in all directions
Room 2: You’re in the jungle, with exits in all directions
Room 3: You’re in the jungle. There are exits in all directions
Room 4: You’re in the jungle. Exits leave in all directions
This is intended to be another repeat of the all-different maze from Crowther/Woods; you simply use the text differences to map things out.
At the end of this you end up on the east side of a river (no swimming, remember) with some thermal goggles. I think this is meant to be the only prize.
You find yourself standing in an open patch to the west of the jungle. A wide river lies west, with more jungle beyond. Crevasses block progress along the river bank
There is a pair of infrared goggles here
I could just make grumpy noises and move on but despite this being the sort of maze designed on a spreadsheet, I wanted to test something out: is there a mathematical way to determine how difficult the maze is? I’ve mused about this before with no good result but finally came up with simply cranking the maze through a Monte Carlo Simulation. I have a program that starts the player at the Landing Area and has them wander for 100,000 turns, while keeping track of how many times the player arrived at a particular room. The results:
That means out of 100 random turns, the player reaches the goal approximately once.
A random walk is not absurd player behavior, even — sometimes I have just started typing directions randomly in frustration, and depending on the maze design, that can lead to good results. Authors did think for and account for this; even Don Woods made a “diagram of the first maze (the all-alike maze), used to check whether any simple repetitive actions would get you out.” Here, at the least, there is no connection going from Room 1 to Room 4 but plenty going the other way, so there’s at least a “trap door” effect so beloved of authors in this time period.
Alas, we still can’t just look at that percentage (“1%”) and call that an objective measure that allows comparing all mazes in adventure games. For one thing, there is a significant difference in gameplay when objects need to be dropped, and especially if the player has an object shortage and needs to shuffle objects around the maze as things get figured out (which is why I’ve sometimes had maps with rooms like “Dagger 3” and “Rope 4”). Even without that, there’s the “repetitive action” that Woods was testing for; with the Jungle if you just keep going NW you’ll get to the destination of the maze. This doesn’t get accounted for in a random walk. Tricks like the diagonal exit in the All Alike maze of Adventure (where every exit is north/south/east/west, except for the very end) don’t get calculated in either. Additionally, if the author keeps sticking to particular patterns, even if they use the “most extreme” design possible, by staying consistent to a pattern it can make it so the player has an easier time puzzling it out.
For example, the most extreme example of a maze is the “all or nothing” style, with a sequence of rooms where
a.) only the right direction makes forward progress
b.) all other directions are wrong and will always send the player back to the very start of the maze
(For an example of this in the wild, see Adventure 500.) Testing this configuration in my Monte Carlo setup using the same number of rooms as the jungle maze, I get that the player will reach the final room only 1 out of every 100,000 tries (rather than 100). However, the majority of players will catch on to what’s going on and start to just search for the forward-exit in each room and ignore the other ones; it additionally makes for an easy to chart diagram.
A (nearly) all-or-nothing maze I saw recently from Magic Mountain.
One of the issues here is how comfortable the author is with the player hitting a solution “at random”. Some seem offended by the idea of luck playing a part in puzzle-solving, while others are more relaxed about it. An author who hates wins attained at random might be more likely to pick the all or nothing structure (even though the effect in gameplay is unrealistic, on top of making the act of mapping more mundane).
Speaking of hitting solutions at random….
…I was mostly wrong about the swamp last time. I had, by sheer luck, found not only that I could step onto the swamp by going northeast, but also all steps after went to the north. This is generally not what’s supposed to happen. The thermal goggles from the jungle are meant to aid in figuring out where the safe spots are, and it isn’t always going north. Otherwise the result is the same (getting a black hole and a can of acid).
Besides that, the progress I’ve made is back at the mysterious alien dome.
Entering, again, requires dropping all items, so this section is entirely self-contained. You start by following a path eventually leading to a dead-end, getting a sequence of shapes that is randomly determined out of square, triangle, hexagon, and pentagon. Pressing one of the symbols (it doesn’t seem to matter which) drops you in a maze, where the direction you are facing is now tracked, and the only commands are AHEAD, BACK, RIGHT, and LEFT.
This time my map is not fully-made out because it seems at least the shapes have some randomization, and I confirmed after enough testing the only things to find were a.) a set of four symbols on the floor, that can be picked up and b.) a hole leading down further.
The next level has a sword and a nuclear reactor, followed by a robot (with the solution I mentioned last time still holding — it doesn’t seem to be randomized — the word is always FLEZZ). The robot is followed by a giant rat. Last time I killed it with the sword, but this time I just ignored it. This turned out to be a wise choice (I’ll explain why in a moment). Last time I also didn’t realize the symbols on the floor were portable, but this time I took them along, and after some struggle realized THROW PENTAGON (etc.) would get them in the machine. I had to follow the same sequence as earlier:
If you’ve killed the rat, you get a different message:
You have passed all the tests, earthling, save one — you killed without reason. Begone!
The intent seems to be to lampoon fantasy adventures and their tendency to kill rats with swords on sight. (SIDE NOTE: There’s a little more lampooning in the verbs. I singled out RUB and WAVE last time; apparently Killworth thinks along similar lines, as if you apply the verb to an object like the sword, the response is “This isn’t a fantasy game you know — doing that won’t help.” Rather than roaming into sci-fi as simply a different window dressing to the adventure form, Killworth specifically is trying to root things in a marginally more scientific world. Marginally.)
And that’s essentially it! Going down there’s still exits with the four symbols, but this time you can pass through safely (although again, you need to repeat the same sequence as before).
Once to the end you can go through an exit out the back, leaving the artefact behind for good.
Still remaining to reckon with: the desert which causes the compass to go haywire, the monster at the icy crevasse, the volcano tube which shoots up lava when trying to climb inside, the blob which eventually falls off a cliff, the hover-platform without an easy shutoff, and the mysterious large cube which you can enter.
You are inside a three metre cube of metal that vibrates slightly as if it only partly exists. The exit is north
(If I remember right this allows time travel somehow? The game might suddenly get much bigger. I have yet to get the cube to acknowledge my presence, though.)
Renga in Blue 
Leave a comment