This is kind of a side tangent but one of the next 4-5 games I have in mind to work on is a game in the model of Zork and Adventure. One thing both had in common was a kind of optimization game (especially Adventure) with light and other resources.
So I’d like to make a game with several simultaneous puzzles, where there are three phases of gameplay:
1-Figuring out the early, individual puzzles
2-Getting to the harder areas/puzzles one at a time but running out of time (or resources) after just one or two
3-Optimizing the game until you win
Adventure did this, and so did Zork (to an extent) and Planetfall. Hadean Lands does it with resources.
I think unwinnable states can be fun in a game like those where tons of puzzles are accessible right from the beginning, so you can easily focus on different parts every time you start. I think unwinnable states are awful when you have to copy the same actions every time.
I think a particularly bad example is Enchanter. It requires Food, Water, and Sleep, and only one of those actually contributes to the plot in a meaningful way. I love the game but if food and water were cut out I think it would still be just as fun as before.
I think it depends on what you mean by experiment. To me, the Babel Fish was an experimental puzzle – i just had to keep trying something until I found the answer. So it led to some experimentation. Anyway, just a thought.
Yes, I agree with what seems to be the consensus forming here: that how bad unwinnable states are, is proportional to how much tedious repetition they force you to do. Hadean Lands can’t be played without encountering dozens and dozens of unwinnable states but it makes resetting nearly painless.
People are averse to lots of things that are worth doing. Exercise for example. You could use this argument about permadeath in roguelikes. Maybe making players replay an unrandomized game is good and maybe it isn’t, but not because people are averse to it.
The thing about unwinnable states is that they’re boring. There’s only really one type of game that can make them fun (apart the Hadean Lands): and it’s the optimisation game. The one where you can restart and go, “oh, yeah, my run-through will be different”. And those are all the conclusions that we’ve reached.
But coming back to the original decision, they might be unfun in the original Infocom games, but should we remove them? No. Many of the games rely on them or use them to their advantage: Deadline, Suspended, H2G2, and so on. Some feel a bit annoying with them. But those are the design choices that were made back then, and it feels weird to change it. It was an active choice they made, unlike some of the bugs.
Speaking of which, I found a bug in the code of Bureaucracy.
I can’t disagree with any of this, but at the same time I also love Angband, where a successful playthrough is probably 30-50 hours depending on class…
Though it’s notable that it’s probably the easiest and fairest of the classic roguelikes - after a dozen or so hours and learning maybe three or four tricks, you really only die post-opening if you’re extremely unlucky or intentionally playing in a risky way.
Games are supposed to be fun. If they are merely an exercise in tedious repetition, not many people are going to put themselves through it. As I said, repetition in some types of games is tolerable due to variability and skill-based results. Using Roguelikes as an example: I do enjoy Nethack playing as a human Archeologist, even though I’ve never gotten very far. The game is different every time, and sometimes interesting and unexpected things happen, so I tolerate the long replays. Infocom games mostly lack any kind of variability in replays except for a few random elements. People might tolerate retyping all the commands they need in order to get back where they were after dying, but I doubt anyone actually enjoys it.
Many terps already lets you paste in command. I almost always use Windows Frotz for z-machine games and keeps a text-file with the tedious command sequences (paths through a maze, for example), and only paste it into the terp.
True, but they stuck around for a very long time, even in games where they had been rendered completely useless.
Though I have to wonder how much the walking death states in classic IF are genuinely intentional versus a limitation of the times versus there not really being any standards of good game design in those days(the 80s are lousy with games that are frustratingly hard and at times downright unfair as much because of technical limitations and because pretty much everything was an experiment in something new as deliberately making the games hard.
You can say that design decisions in Colossal Cave and MIT Zork were unintentional. The designers were throwing together ideas as fast as they came up with them.
By the time Zork 3 was released, you can no longer say that. Infocom was thinking very hard about their game design by then. They had standards of good game design, and what we call “walking dead states” were good game design. In Enchanter they designed the entire game around an interlinked web of walking dead states.(*)
Any modern analysis has to start with this fact and work forward.
(* The KULCAD scroll, which can be used to solve many puzzles, all but one of which has an alternate non-KULCAD solution.)
Limited lives were just a design holdover from arcades, where there was financial incentive to make the player lose quickly. They make no sense in other contexts and eventually most game designers realized this.
Maybe that was true originally but, in most linear platformers, lives serve as an incentive to collect coins, etc. This basically gives you two ways to manage risk.
You can go slowly and collect coins as a buffer against losing progress. Or can speed through levels without collecting anything, and if you die, you can try speeding through again because you didn’t spend so much time collecting stuff.
I think this model works best when there are significant limitations on when you can save and how often. Donkey Kong Country, which required you to to reach a save point in each world, did it best I think. Though Donkey Kong Country is a bit too unforgiving.
Platformers and non-platformers that aren’t linear don’t really need to take this approach because there are other reasons to dump time into exploring what they have to offer.
However, the “undo” command in IF/text adventures is similar to lives systems in way. Whether the player uses undo (and whether the game designer allows it) determines to what extent the game can be played through trial and error or requires full engagement within the limits of the game, recognition of what you’ve done, and understanding of the gameplay loop.
I’ve used limited/absent undo and save systems in some of my IF games. I’m sure there are more but they’re not labelled on IFDB and nothing comes to mind.
Beer by the lake sounds awful. Granted, I’m not a big fan of the outdoors and every sip of beer I’ve ever tried made me think “How can anyone choke down enough of this stuff to get drunk and why would they even try when their are alcoholic beverages that actually taste good”.
But yeah, if a game stops being fun, that’s usually an indication something has gone wrong.
The outdoors is great in small doses, with air conditioning and computers nearby. Beer is just plain awful, but is protected by some sort of weird conspiracy. Same goes for coffee.
Ugh, coffee is worse than alcohol. I’ve found mixed drinks I like, there are several flavors of Smirnoff Ice I like, and I’ve yet to try a hard lemonade I didn’t like, but I’ve yet to find a coffee-based beverage that didn’t make me cringe… and on the subject of bitter, I’m not a fan of dark chocolate either(milk chocolate is okay, but white chocolate is my preferred variety(too bad so much of the commercially available stuff skimps on the cocoa butter in favor of cheaper vegetable oils or omits cocoa butter altogether).
