Effective design decisions in Graham Nelson's Curses!

I recently completed a survey of Interactive Fiction, where I tried to play all of the major games of the amateur era. It was enjoyable, and I discovered some wonderful games. After completing it, I decided to go back and play some old favorites, to see if they still held up. Curses! had always been my favorite game, so I booted up. I quickly found myself engrossed by the quotes, the narrative, the puzzles, and the overall structure. I realized that, more than any other interactive fiction game available, Curses! is my favorite.

I decided to write this essay to explore what precisely about Curses! makes it so enjoyable to me personally, and to quite a few others as well. I do so by highlighting 4 design decisions that Graham Nelson made. All of this is my personal opinion, especially my interpretation of the story line. This obviously contains strong spoilers for Curses!, but shouldn’t really have spoilers for other games.

Coherence of narrative

In my opinion, the most effective games are those that tell a single story over and over again, using a variety of methods. For example, the various colored stories in Photopia are allegories for the same thing; Howling Dogs is the same story in different settings; and the different branches and endings of Galatea contribute to a unified picture of the main character. Many otherwise well-made games can be quickly forgotten if they try to do too much at once (like tacking on an environmentalist message to an otherwise unrelated story, or combining goofy space humor with grim and bloody horror). I found Graham Nelson’s own game The Meteor, the Stone, and a Long Glass of Sherbet to be much less effective because it didn’t settle down on the story it wanted to tell.

Curses! tells the same story from beginning to end, a story about alienation and a desire to prove yourself. In many ways, it’s an allegorical rite of passage, similar to the ancient mystery plays. It’s message is communicated by endless variations on the same themes.

Unlike most adventure games, the setting of Curses! is packed full of people, but you aren’t really part of them. You begin the game in a house full of your big, bustling family and your first goal is to keep away from them. You have basically cut yourself off from the living, and the living are not happy with you; the druids plot to kill you, the Gods try to zap you, Sosostris despises you, and your fortune is the Drowned man, the Fool, and Death. Wherever you go, everyone ignores you, looks down at you, or attacks you. Your only friends are those who have also cut themselves off, like the crusty Jemima, the incorrigible Austin, or the self-sacrificing Andromeda.

Not only are you yourself cut off, but your family is one great big failure. The house is full of incomplete tasks and odd jumbles of stuff, and the information you have about your family shows that they always fail at the critical moment. You are proud of your family, but embarrassed at the same time. “Just another Meldrew” seems to be how you feel about yourself.

As the game progresses, there is very strong imagery related to death and rebirth. You narrowly escape death on numerous occasions (such as jumping from the mast of the ship). You disarm a bomb, you are poisoned but take the antidote. This is all fairly normal in adventure games, but it get’s stronger. You sleep in a tomb for centuries, unseen by others. You have to put the rods into a coffin and out again to release their power. You have to anoint yourself, bury yourself, and rise as an immaterial being to progress. In one of the most powerful allegorical moments, you must reconstruct a mechanical knight with no heart, no hand, and no head, bring it to life and teach it to love; like the player, they must be reborn from their impotent, heartless state into a full being.

This death and rebirth are the key to true power. The rods gain power after their rebirth; you gain respect from Sosostris; you can defeat the Kraken by death and rebirth; and after constructing the allegorical knight, you gain access to true power in the form of a high rod, where you yourself select between Love, Life, and Death. In the end, hidden in an underground cave, you discover your true purpose (essentially the divine self of the mystery plays), and redeem not only yourself but all of your family by completing your ancient task. Your ultimate reward is to rejoin your family as a whole, complete person.

None of this is forced on you in the game. The storyline is slowly spooled out through surrounding actions and thoughts. The game rarely tells you how you should feel. Instead, feelings are reinforced by endless repetition of the same theme, and this is the first effective design decision of Curses!.

(One weakness people have noted is the demon in the basement; for many people it seemed out of place. The same is true of the angel in the clouds. I wonder if this was intentional, as they are both hint systems, and as such do not really belong in the game either.)

** Puzzle design**

Curses! is notoriously hard to complete, if nothing else due to its length. But there are several elements of its puzzle design that I think are especially effective.

Nelson has avoided one problem that other big puzzle games have, which is managing inventory taken from all over the world. Muldoon Legacy and Finding Martin both have numerous puzzles where you combine completely unrelated objects from far across the world, and this is essentially unfair.

The vast majority of puzzles in Curses! fall into two categories: puzzles unified by location, and puzzles unified by theme. These help cut down on the work generated by sorting through your inventory.

As an example of puzzles unified by location, consider the garden and surrounding areas. The garden roller literally cannot be removed from the area, making it clear where it must be used. The rod obtained in the maze is used in the garden. The weedkiller is used in the maze. The ball on the mosaic is used nearby.

As examples of puzzles unified by theme, the rods themselves are identified by a common shape. The pollen-filled branch tells you that ‘you don’t get the bird’ when you shake it, and you later use it in a room with a bird. The golden scepter is used in an ancient tomb, and the nuts are used to distract a squirrel. There’s none of the type of puzzles where the shaft of a spade is used to knock a marble into a mousehole.

The puzzles that don’t fall into these categories generally have multiple solutions. There are 7 ways of opening the medicine bottle. The main puzzle of the midgame (how to return from the various worlds) has a variety of solutions of increasing efficacy: using the poetry book (a one-time solution), lagach (a partial solution), and finally the rod of returning.

Some weaknesses in this area are the occasional guess-the-verb puzzles, including two about speech (commanding the mouse and lagaching), and the important puzzle about learning to use the rods (HIT GROUND WITH ROD, HIT ROD, etc. don’t work; only STRIKE ROD works, which is an odd construction).

Classifying puzzles into themes and allowing multiple logical solutions is the second effective design decision of Curses!.

** Information storage **

This is a smaller part of the game, but an important one. Like most plot-heavy games, Curses! has several info dump mechanisms. Babel has its memories, Theatre has its diary pages, and Worlds Apart has big cut scenes and actual, readable books. Christminster has its library.

Many of these suffer from one of two opposite problems: either everything is forced on you at once (like the book in Worlds Apart), or information is only available after extreme searching and can easily be missed (as in Christminster).

Curses! avoids this neatly by having three books (the history, the diary, and the dictionary) where you are very clearly told what can be looked up. Thus, every time you find a name, you can look it up in the history. Every time you hear a year, you can look it up in the diary. And every mythological name can be found in the dictionary.

This lets you reliably discover information at your own pace, which is the third effective design decision of Curses!.

Pacing and difficulty

The last decision has already been hinted at, but its the excellent use of pacing. The game opens with short paragraph or two that immediately sets the scene, gives you a goal, and encourages you to explore.

The initial area is small, allowing you to explore, and exploration gives you several natural mini goals. As you walk north, you disturb a key that falls down a floor. You find a map, but it’s in a jar whose lid you can’t grip. You find a cot that invites you to sleep, you find a pipe that needs tightening. All of this suggest things for you to do.

The game was intentionally designed so that players stuck on a puzzle can always be working on something else. The author has stated that this is why he added so many things to find in the attic; three secret doors, three portals to other worlds, a passage behind a fireplace, a skylight, a darkened area that you are nervous to enter.

He also deliberately removed a large chunk of the game because testers thought it was too much. Many epic-writers like to stuff a game full of as many puzzles as they can. It’s wise to consider player exhaustion when doing this.

Many of the slowest parts of the game are ‘spiced up’ by an appropriate quote or a joke. The quotes are carefully selected and contribute strongly to the overall feel of the game. Some of the jokes are pretty lame (“It’s a wrench, but you take it”, “Let’s call a spade a spade”), but they break up the monotony of the least interesting areas.

Finally, information is given as a reward, instead of having the author gleefully dump it on the player. This is something Midnight. SwordFight., Babel and Spy Intrigue did well, too; the backstory as reward encourages player exploration and is more effective than a purely point-based reward system. However, the player must know that information can be obtained and must not be starved for it; the second Unnkulia game, from before Curses!, literally does not tell you what you are trying to do in the game until the very last move.

Controlling the pace of the game through variations in difficulty and information rewards is the final effective design decision of Curses!


Curses! is not the perfect game. Its conversation system is minimal, there are some lingering guess-the-verb issues, and it is difficult to finish it without a walkthrough. But its clever design choices make it an enduring work, and I still regard it as my favorite game of all time.

You seem not to have commented on a design issue that is crucial to me and prevents me from enjoying all the best that Curses has to offer - it’s not only possible but easy to lock yourself out of victory without realising it for a long time, or even ever. Every time I entered a, well, let’s call it a vignette - every time I entered a vignette I was worried whether I had everything I needed to succeed in it; whether I’d be able to return, and if so, whether I’d be able to leave again; whether I was able to get everything I needed from that vignette. If (hah - when!) I got stuck later on I’d be worrying about all the vignettes I’d been through, unsure whether I’d missed something that I now needed.

This is the one aspect in which Mulldoon Legacy well and truly one-upped Curses for me, and it made all the difference between an enjoyable and an unenjoyable game. I could always trust Mulldoon Legacy, even if I had a few issues with half a dozen of puzzles. Curses I couldn’t ever trust. Progress always felt like a fragile thing, that could come tumbling down at the realisation that I’d locked myself out - probably on more than one occasion in a single playthrough!

Having said all of that, those are interesting thoughts, thanks for sharing.

Have you been able to “lagach” through images? This allows you to revisit many areas.

I don’t remember being able to do that, no. Does it also allow you to leave? I remember a vignette (I think it involved lots of barrels and caskets, or maybe it was a store room) where I could only leave once, so returning to it later would be a walking dead.

If I am wrong about this - if it is possible to leave these vignettes, or to return to them later for something you’ve missed - I’d be very interested in knowing, because it’ll make a big difference and prompt me to revisit the game.

Any area that has an image allows you to leave it and return by saying “Painting, lagach” or something similar. This gets you back to Hamburg, the Lighthouse and the Maiden. But the store room and restaurant are still one shots, each with just one item.

Hmmm. Just those two? And just one item to get? Well, knowing that does make a difference. May be time to give that monster another shot - thanks!

Gah. Now I remember why I don’t like the game.

It’s not that it’s huge - I can deal with huge. But i’s overwhelming! Really overwhelming. The large map I could deal with. A huge inventory list? Fine. One way trips to various locations? Sure. Strange magic words that keep elluding me? Right on. Trying to figure out what is a puzzle and what isn’t? Ok.

Bring them all together and what do you get? You get me bumbling around, brain hurting, and trying to finish the game for the fourth or fifth time - and always quitting more or less in the same place, always feeling masochistic, always swearing to myself I wouldn’t pick up the game ever again because it’s just too much at the same time for me to enjoy.

EDIT - And no, I haven’t been able to lagach through images. What’s worst is, I don’t know whether it’s because I have to solve a puzzle or because the parser wants a strange syntax for it.

EDIT - By consisting of smaller and more self-contained scenes, Jigsaw, The Mulldoon Legacy, and most Andy Phillips games were far more enjoyable to me - even solvable, give or take a peek or two at the walkthrough for the odd puzzle. I just can’t deal with something this sprawling. Which is a shame, because I do like many things about it - individually. The whole is meticulous, ordered, well crafted - it is a non-messy sprawling mess.

EDIT - This really bothers me. Curses is one of those games I very much want to like! That’s why I keep coming back. But I always form the same impression and have to leave, having been thoroughly battered to submission. There are many reasons to want to play IF, and being battered to submission does not rank highly with me.

A couple of decades ago this would reflect poorly on me. Nowadays it may reflect poorly on the game. Depends on whom you ask. (I think the answer lies in-between - how I’d LOVE to see Curses slightly redesigned! Then again, maybe I have, and it’s called The Mulldoon Legacy! Or even Jigsaw!)

Re the magic word thing:


I tried that. “You can’t talk to *** ****”. Amidst all the other difficulties I had, I could never tell whether this was meaningful - all the error messages pointed towards me not using the correct syntax. Not a single message indicated I may have been trying to lagach something unlagachable. I never really knew whether I was phrasing it wrong or trying the wrong target… really disheartening.

[spoiler]IIRC, lagach cycles you among the various paintings/murals/etc. that you’ve encountered so far. So, it won’t do anything unless you’ve discovered at least two.

In the case of the painting, it doesn’t count until it’s been hung in its proper place.[/spoiler]

I wouldn’t say that it reflects poorly on you, but I think it reflects a change in players’ expectations, and not necessarily one for the better. At the time, there weren’t many IF games and most of them were on the longer side. It was expected that one would mess around with a game for a period of weeks or months, maintaining lots of save files, attacking it on various fronts and slowly making progress. It was about facing a challenge and eventually defeating it, and a lot of the fun was in the journey rather than the destination – exploring the environment, poking at the fringes of the simulation, uncovering custom responses to weird actions or in weird scenarios that the author had anticipated. For all that I’ve enjoyed smaller story-oriented or experimental games that can be consumed more quickly, the massive parser-based puzzlefest (Curses, The Mulldoon Legacy) is still my favorite subgenre of IF.

And then when I get into discussions like this, or read reviews, I get the feeling everyone got further than me, which makes me feel dumb and doesn’t encourage me to keep at it. :slight_smile:

I mean, the first time I peeked at the walkthrough and saw mentions, very early on, of things I’d had no idea about… those rods everyone keeps mentioning… the angel, I never saw any angel! I only got to the demon!

Peter, you and I have very different styles; you play without walkthroughs, and I only play with walkthroughs. The first time I beat curses, I used a walkthrough for 80% of the game. Each time I used a walkthrough less, until I finally beat it without hints for the first time two days ago.

I think this might color our different reactions to games; I was frustrated by Muldoon Legacy and Andy Phillips, but not by Curses!

By the way, what do you think of the Unnkulia games?

[I edited my previous post to include a response to something that you’d posted upthread.]

People find different puzzles easier or harder and solve them in a different order from one another. Here’s a Graham Nelson quote from an earlier draft of The Craft of Adventure than the one that appears in the DM4:

Rather than approaching the game as if you’re in a race with others to beat it, I’d suggest that you slow down, relax a bit, and just poke around. Can you find a new room? A new item? Manipulate something in a new way and elicit a new response? Are you having fun? If not, go do something else and come back later. The game will still be there.

I think the one place where this approach breaks down is where a game bottlenecks on a single puzzle that one can’t seem to solve. I usually play without hints and am not a fan of walkthroughs, but in this case, after several sessions with no progress, I’d say go ahead and grab a hint in order to move forward. Luckily, the larger puzzle-oriented games tend to have multiple puzzles open at once during most of the game.

I’ve yet to get to them. :slight_smile: I have played the first one when I was a kid, and remember very little. I’m looking forward to them!

I’m sorry if that’s how it came across, it’s not what I meant. Let me try that again - it’s been years that I’ve been playing Curses! on and off. When I finally relent and look to the walkthrough, I see tons of references to stuff that, apparently, I’ve completely missed on much earlier parts of the game. When I start looking at spoilery reviews, I see people talking matter-of-fact about important items and verbs that I had absolutely no idea about.

This is not what you want to see if you already feel squeezed like a lemon. :wink: I’m still trying to find my way out of the maze in the past (without using the poetry book) and all of a sudden I don’t know whether I’m focusing on puzzles I can even solve! Then again, neither do I know whether I DO have to use the poetry book in that place. So I either use up that resource and play onwards, never knowing whether I’ve made the game unwinnable, or I don’t use it and risk being stuck because I WAS supposed to use it.

That maze dilemma I just alluded to sums up what I can’t deal with in the design of Curses.

You know what, maybe what the game really needs is hints. A proper hint system. Or at least external hints, right now all we’ve got are straight-up walkthroughs. God, I’d welcome a hint system to at least let me know whether I was on the right track, or telling me “quit banging your head against that puzzle, you’ve others to solve first”. A walkthrough for a game like this doesn’t really work. It tells you what to do regardless of what you’ve accomplished already, so if you want to just try solving the thing mostly by yourself… you get frustrated because you see all the crap you’ve missed out on because you were dim. :stuck_out_tongue:

I played The Mulldoon Legacy for two weeks in exactly that way (and most Andy Phillips games as well). That game earned my trust (and so did AP’s games). Curses crushes it every time I venture into it again…

I love Curses and its puzzle design decisions. They’re all mostly local and pretty much clued in one way or another.

I’ve not finished it yet, though. Like most large IF, I usually take several uncompromised playthroughs just to see where each path of action leads to. That’s the right way to play puzzlefests: lower your expectations about finishing it in a few hours or so. These games were meant to show you dead ends, so you can learn what to do and what not.

Interesting discussion either way.

Just in case it wasn’t clear, I don’t have those.

Sorry if that wasn’t directed at my posts in particular, sometimes it’s hard to tell in forums. :slight_smile: If you meant the “your” to refer to a generic reader/player, nevermind me!

I’m quite late to this thread, but I gotta say, Craiglocke I encourage you to write up more from your survey of IF games. If you did an analysis like this on 10-15 IF classics (community or old-school Infocom games) that would make for a wonderful book for would-be game designers. I get that it’s a subjective view point, but I think it would add immensely to the body of knowledge available for designers. Today with sites like Kickstarter for funding and Amazon Createspace for publishing it is much easier to embark down a path of donor funded projects.