Return to Camelot is a fantasy pastiche game loosely combining hardboiled detective tropes with Arthurian characters. It’s fairly sizable, and I ran out of time before completing it (though others might not).
As of ADRIFT 5, ADRIFT games can now be over the web in a WebRunner. This is a terrific development, and I’m delighted Campbell Wild’s putting the work into it.
The downside is that everything is still new and the runner is still a bit quirky (I ran into a minor but irritating display bug that Campbell immediately fixed, but it meant I needed to restart the game at one point to take advantage of the revision). Also, there’s a significant delay in responding to commands because there’s a reload of (as far as I can tell) the whole page every time you act; and if you go away from the game for a while, your session appears to time out. The cumulative effect of all that is that I spent much of my permitted comp-judging time replaying the opening to Return to Camelot (because of the bug, because I got interrupted while playing and had to start over, and because playing is just slower than it would be on a local interpreter). So I didn’t get nearly all the way through it and had to go to the walkthrough.
Design is fairly wide-open, in the portions I saw — it’s possible to wander around quite a bit without knowing where to look first. That’s not always a bad thing, but I would have welcomed a slightly more vigorous pace in the early midgame. Then again when there is hinting, it is extremely direct, even aggressive. At one point during the early game, a message popped up saying “Narrator interrupting:…” with a clue about what I ought to do next to trigger forward progress. At another point, an NPC directly challenged me to do something I didn’t particularly want to do, and for which I was then immediately punished — but the story wasn’t going to go forward until I did it.
This felt awkward. A tighter design would have let me find the intended path a little more naturally, without this kind of fourth-wall-breaking intervention. The writing also could use a good thorough trimming; there are sentences that look like hedges gone untended for a generation. (More specifics about both those issues follow the spoiler space.)
Overall, from what I saw (the latter part all from the walkthrough), this is a fairly middle-of-the-road offering in terms of puzzle difficulty and implementation depth. The setting doesn’t really grab me, being chiefly a compilation of bits borrowed from other works, rather than a freshly imagined or re-imagined world. So it falls at the low end of my recommended scale, I’d say: it’s playable and has some amusing moments, but there are an assortment of craft issues, and it didn’t grab me with a compelling, coherent vision.
So in the first part of the game — which for circumstantial reasons I replayed I think four or possibly five times, so I became intimately familiar with it — there are a couple of critical triggers to move the story forward. First you have to sit in your chair to trigger the appearance of your visitor; then you have to kiss her in order to get kidnapped into Merlin’s lair. Using triggers to ratchet the story forward is a time-honored technique in games, and it’s useful, but the aesthetic goal is for the player not to experience any stuck periods or to have to go on a hunt for a trigger, and also not to notice the arbitrariness of the triggering event. In this case, I did both: I spent a bunch of time poking around my office looking for something useful to do with the painting after it was delivered until the narrative voice told me to sit in the chair (and then I sat in the easy chair, which didn’t trigger anything, before trying the office chair).
So there are a couple of points here about triggers.
1) It’s good if they’re connected to something the player has a strong intrinsic reason to try even without knowing there’s a trigger there. Sitting in the office chair isn’t necessarily the obvious choice for the player, given that there are appealing objects to explore and a mystery about the freshly delivered painting.
2) Sometimes it’s helpful to have more than one trigger to set off the same event, or to use a count of related actions, rather than just a single action, to spring the trigger. Assuming the aim is to let the player look around his office for a little while and get acquainted with the protagonist’s semi-loserdom, but to move the narrative along before he gets too bored or stuck, it might be reasonable to trigger off, say, interacting with three of the narratively meaningful objects in the room (out of a somewhat larger total set — say the painting, gun, whiskey, files in the desk drawer, phone).
3) If at all possible, it’s cool to disguise the arbitrariness of all this by making the trigger and the triggered event feel somehow related. E.g.: give the player the fake goal of making his office look a bit less pathetic; then as he tries to carry this out, via any of about a half dozen legitimate methods (like say scattering the files on his desk), have the beautiful woman walk in and mockingly notice him in the middle of this embarrassment, but move the plot along. Now we may not have agency in the classic sense that the player accomplishes what he wanted to accomplish — on the contrary, his actions make him look like a bit of a dork in her eyes — but there’s a narrative connection between what he was doing before and what happened after.
As to the writing, here’s what I mean about trimming (and while I’ve picked an extreme example, a lot of the writing shows similar characteristics):
Peacocks, pecking for food, wander around a marble bench which is placed under the protective shade of a small almond tree and a low stone wall surrounds the garden on two sides, the castle wall making out the third and the wall with the arch which you walked through to the east, makes out the fourth.
There are too many ideas in this sentence. The structure means the player may need to read it more than once to work out the layout that’s being described, and by the end the interesting details about the almond tree and peacocks may be already forgotten. Besides this, much of the description has a dutiful quality:
> x stone
The stone is light gray, not that it matters, but that’s the color. It’s polished smooth and is actually quite decorative. There’s some writing on the side, only a few sentences written in glowing letters.
I don’t know about you, but I actually felt a little sorry for the author as I read this. It sounds like a cry for help. “Oh, fine, the player wants me to describe this stone… I guess I have to pick a color for it. It’s rock-colored, okay?”
Many many objects in the game are described mechanically with reference to their color, material (lots of oak things), or general level of attractiveness. The result is often a kind of modifier soup: too many adjectives and adverbs, sentences that go on too long about things that aren’t distinctive or important, and a real wall-of-text experience for the player on some moves. The descriptions would be much stronger — more memorable, more interesting, easier to parse — if the author gave us half or a third as many details, but made sure each one was something that mattered.
What matters about this stone (other than the already-mentioned presence of Excalibur)? It has glowing letters on it. That’s the interesting thing. That’s the first thing I’d notice if I were checking out a magic-inscribed rock. Lead with that, describe it evocatively, and I guarantee the player won’t be thinking “okay, sure, glowing letters, likely magic, has Excalibur stuck into the top of it, but what shade of gray is it?”