Too Large Game

This discussion reminds me of Adam Cadre’s blogged thoughts last month on “Endless, Nameless”, which was both his first public IF release in many years and a rather large game that took him a long time to make: adamcadre.ac/calendar/14/14559.html (Scroll down past the comics discussion until you get to the graphic of the computer and the “Endless, Nameless” header text. Afterwards feel free to return and read the comics discussion anyway, because Adam’s deep criticism is quite delightful, but that’s neither here nor there.)

If I understand him correctly, he writes that his big game in 2012 was downloaded and played by a much larger contemporary audience than anything he ever published in the late-90s/early-00s era, and yet he received only a relative trickle of post-release feedback – despite an enthusiastic writeup by a very popular PC gaming blog. I get the impression he was looking forward to the rich public online discussion that would follow his (and many other authors’) new games back then, and found himself feeling disappointed when this did not really happen. This though everyone who has played it through seems to agree the game is excellent, perhaps even his best work (!), with a 5-star average on IFDB.

I suppose my point here mirrors what others have already said on this thread. If you have to sink a few months into making the game exist because it’s punching its way out of you, and you envision yourself being proud of it for being part of the world from then on, go for it. It’s a great time to write IF, and you may be surprised at the attention your work will receive. Perhaps for more than one definition of “surprised”, and perhaps all at the same time.

On the more technical topic of testing enormous games: I find this interesting because of the difficulty inherent in accessing and testing things that happen later rather than sooner in a game. (Imagine a book whose final chapter would let you proofread it only after reading all the way up to it at normal speed from the beginning, with every new proof copy…)

This phenomenon isn’t limited to IF; I have encountered talk among triple-A studio videogame developers that they expect many-perhaps-most players to never see the end of their games, and that’s unfortunate enough as it is. But with noncommercial IF, when your QA team is all volunteers who often simply can’t invest the time needed to slog their way through every pathway offered by every build of a game, it seems natural to be that you’d end up with rather more dust and disarray the further you travel from from the front door.

I imagine that there perhaps exists some ad-hoc parser-IF convention for giving testers a set of “warp verbs” that zip them ahead several chapters, setting up the game state and packing the PC’s pockets appropriately, and perhaps presenting them with a story-so-far synopsis of everything the player (versus the character) would know by this point…? (And pairing this with an earnest request from the author to please make use of them, maybe especially later in the testing process?) I don’t know, man, all my games are like 20 minutes long so I’ve never thought to ask…

There’s an Inform 7 extension, “Checkpoints”, which lets the author define the game state at certain points in the story: an error will be thrown if the player gets to that point while in the wrong state, and a debugging verb can be used to jump to any checkpoint in the game.

Heck, I do that for myself.

I don’t generally give the debug verbs to testers, but I usually don’t rely on testers for multiple play-throughs.

I made heavy use of the I7 Skein while building Warbler, as both a test framework (“Did this change I just made break anything?”) and as a way to warp freely around the story, with world-state changes included. But I didn’t make this or anything like this available to testers, who I expected to always play from the start, I suppose. (Or be all hardcore and manage their own paste-ready transcripts, or even use Zoom on Mac to make client-side skeins…)

I have on occasion sent testers (well, one particular tester) debug versions, with which the player can use verbs like purloin and goto in order to save time. What I found, however, was that using those shortcuts sometimes puts the game into a wonky state, a state in which bizarre things show up that would never happen in actual game-play. But yeah, if used carefully, such a technique can save tremendous amounts of time.

The other factor militating against it, though, is that some testers just want to play the game. They will go through the game and send you transcripts, which is great, but they’re not actually trying very hard to be testers. A good tester (and this is far afield from the original topic of this thread…) really tries to break stuff by using commands that few if any players would ever think of.

oops, er…

Jim! Please make the game, it sounds delightful.

It’s kind of funny, because I read the first post without noticing who the author was, and then I was mentally composing a reply, along the lines of “I like long old-school puzzly games, like Muldoon Legacy or, what was that one by Jim Aikin about a ballerina? I play the heck out of them and then I don’t ever replay, but I really enjoy them the first time, so please make your game.” Then I got a few posts in and realized who I was mentally replying to.

So yeah.

Thankee kindly. You may like this one (assuming I finish it – big job ahead). It’s an explicit sequel to “Ballerina.” Ten years have gone by, and little Samantha is now 17. Tonight is the senior prom, and Sam’s little brother Stevie has spilled black ink on her prom dress, so Mom has to return to the fashion boutique in Stufftown to get a replacement prom dress … which is not going to be easy at all, for bizarre and convoluted reasons. Some of the same shops, some new shops, a couple of puzzles you might recognize (in strangely altered guise) … but no mazes this time. There might be something that looks like a maze, but it probably won’t be.

Let me just say that my biggest turnoff in Ballerina was the puzzle that got power back to the mall, that strange math puzzle. Even when I want to replay it, I get to that and I just think “what if the rest of the puzzles are like this?”. I can’t understand it even when it’s explained in the hints.

My point is, do think twice before including puzzles like that in the new game, please. :slight_smile:

The purpose of testers is a lot broader than just finding bugs. Finding bugs is obviously a huge part of testing, and the kind of tester who really likes to poke around and break stuff is obviously valuable. But there are many, many questions I want answered during the testing process beyond ‘where are the bugs?’, and some of those questions are better-answered by different kinds of testers.

I expect you’re right. Thanks for the comment. I do like having a variety of puzzle types – using physical objects in unexpected ways, noticing bits that seem at first glance to be only scenery, figuring out how to influence an NPC, getting past barriers, and so on. Logic and math puzzles need to be part of the mix, I think, but that one was probably too intimidating – certainly when it appeared so early in the game.

Please make it, I’d like to play it!

Now, I hope I haven’t missed anything by not reading the last 5 pages of posts … That doesn’t bode well for me making it through your super-long game, does it? :slight_smile:

But I’d like to play it anyway.

I have enjoyed this thread and its suggestions–and I hope that, if you need more encouragement, Jim, this post helps you to go through with it!

I have to say that I was disgusted with my own large-scale entry (Spring Thing 2013) and figured, why did I bother? It was buggy, but someone wrote me May 6th saying “Thanks for writing this!” He helped me weed out a ton of bugs & put out a less wonky release 2.

I’m glad someone else said something & I’m glad (for myself) I’ve written and expanded it and made it more sensible. Maybe you don’t need that motivation, having already written some good-sized games like The White Bull, but hey, if you’re doing something new to you–go with it!

Unless you’ve found an even more fun project in the meantime.

Haven’t been working on it this week. Trying to control a complex sequence of events (to put the story in a sensible order) while simultaneously allowing the player to choose any of the actions that she might realistically be able to take at that point in the story … it just started to seem overwhelming, and not fun at all. I have this really annoying NPC, you see, and the downside is that she started to annoy me as much as she will annoy the player.

But no, I haven’t found an even more fun project … unless you count the three magazine article assignments I just received this morning, for product reviews of three extremely cool software synthesizers. That’s going to be fun.

Maybe you can integrate the synthesizer reviews into the game; seems like you’ve got everything else covered. :wink: Good luck with the project. Seems like a lot of people are excited about this.

I did actually integrate synthesizer music (not reviews) with IF in “The White Bull.” That game had some issues with plot structure, but it was a serious attempt at integration. Alas, one reviewer faulted the game because he didn’t happen to like synthesizers. He wanted to hear bouzouki music or something.

With respect to the new game, I’m starting to think maybe I need a collaborator – if for no other reason, to keep me from getting distracted. If anyone out there is proficient in adv3Lite (or would like to become proficient) and thinks it might be fun to work on a too-large game, please let me know. The design is largely but not entirely complete, and there’s certainly room for creative input. Up to and including adding a few more rooms and puzzles, or suggesting that I’m out to lunch in certain areas and need to rethink the design.

Inform authors might want to be assured that adv3Lite really is (in my opinion) the newer, better TADS 3. It’s easier to work with, and has some very Inform-like features (including Scenes and Regions).