Last Vestiges

This game is an Inform murder mystery by thesleuthacademy, who has written numerous reviews for mysteries on IFDB in the last year or so. It’s nice to see a game by them!

Mysteries are one of my favorite genres of game, so I was interested to see how it plays out here. There are several standard ways to run a mystery in interactive fiction:
1-Have a standard puzzle game that happens to be about murder mystery, with solving the puzzles leading to solving the mystery. This is like Ballyhoo.
2-Modelling evidence and clues in-game, which have to be combined to form a solution. This is how Erstwhile works, and most of my mysteries.
3-Collecting evidence through puzzles and conversation and then having a quiz at the end (where you have to accuse the right person). This is how Toby’s Nose works.
4-Collecting physical evidence and showing it to someone, being able to make an arrest when you have enough evidence.

This game is a mix of 3 and 4. You have to collect enough physical evidence to proceed to a quiz, and then pass the quiz to beat the game.

The storyline is simple. A man was found alone in his room in a pool of blood with no visible wounds. You must examine this single room to discover the clues.

This game boasts a large number of beta testers, which is nice. I struggled with some of the setup, however. Many of the ‘standard responses’ for Inform were not helpful. For instance, there were some ear plugs that I tried to take and it said ‘That is not portable’. Some commands that might have had useful responses didn’t work; for instance, TALK TO didn’t have any message like ‘Conversation in this game is handled by ASKING’ (although that was mentioned in the help system!) and PLAY PIANO had no response.

There are some very helpful responses, though, like SEARCH and LOOK UNDER saying you only need to ‘examine’.

At the quiz at the end, I really struggled with the third question. I guessed it but then decompiled the game to see how I could have gotten there. It seems that the conversation system is a lot larger than I had expected. I had gotten stuck since SHOW (something) TO (someone) often didn’t have a response, so I assumed asking about those things wouldn’t be helpful.

There is one puzzle of a type I haven’t seen before in a parser mystery, involving a grid. I thought that was pretty clever.

Overall, I felt like tightening up some of the standard responses and adding more synonyms and actions like TALK TO and PLAY PIANO would make this an excellent short mystery adventure.

(This is a lightly-edited version of a review I posted to the IntFiction forums during 2023's IFComp).

Would it be braggy to say that I don’t think I’ve ever had the showing-up-for-a-test-and-you-haven’t-studied dream? When I was in school, I had the Lisa-Simpsonish trait common to many people-pleasing nerds of kind of enjoying tests for the opportunities they offered for external validation, and I guess as a result tests, even when they were hard, never generated the same intense apprehension that makes them the stuff of literal nightmares for many (to this day I do still have anxiety dreams about the time in college that I failed to drop a PE class that I couldn’t attend since it clashed with my course schedule – make of that what you will). So it was something of a revelation to me when I entered the endgame of Last Vestiges and suddenly realized I was wildly unprepared for what this locked-room mystery required of me if I wanted to solve it.

After Mayor McFreeze and Death on the Stormrider, this is the third mystery in the comp that decouples the puzzles required to trigger the end of the game and the steps needed to actually solve the mystery – I continue to like this structure, though its recurrence is making me wonder if this is a well-settled design approach that I’m only belatedly catching on to? – but here, instead of Mayor McFreeze’s medium-dry-goods puzzles and Death on the Stormrider’s NPC-manipulation challenges, progress involves solving a series of escape-room style puzzles. It took me a minute to get into the swing of this, since it felt somewhat at odds with fairly-grounded vibe established by the game’s narration, and I felt another moment of dissonance once I reached the end of the chain, since it turned out that the reward for solving all the puzzles was finding the victim’s hidden will – protip from someone who knows a bit of inheritance law, you generally want to make it super easy for the authorities to find your will, not conceal it behind a set of arbitrary barriers. But judged on their own terms the puzzles are solid enough, offering a bit of variety without too much challenge (though one picross puzzle was a bit annoying solve since it displayed oddly on my interpreter – fortunately it was pretty easy to intuit the answer).

The implementation has some nice touches, with the various characters having a robust set of conversational topics, and an integrated tutorial helps get new players up to speed. And while I wouldn’t formally count it as a limited-parser game, it’s aggressive about pruning out unused commands to avoid potentially confusing or unexpected interactions. There are some odd choices (the game’s single location doesn’t appear to have a description, just a list of its contents) and an awkward moment here or there (when I wanted to check whether there was anything behind the clothes in the wardrobe, MOVE CLOTHES and TAKE CLOTHES didn’t work, with the less-intuitive X RACK being the right answer), but I found the game generally quite solid.

So everything was going well, until I found the will and the game told me I could trigger the endgame at any time, and I realized I only had a very vague idea of what had happened and no obvious outstanding leads still to investigate. There were a few small clues I’d picked up along the way by asking my superior officer about various bits of forensic analysis that had already been done, and noting a few suggestive details about the objects I’d turned up, but certainly nothing conclusive. And then when I figured I’d just see if I could bluff my way through the game’s concluding quiz – which works similarly to Antony and Cleopatra, where you need to select your theory via a multiple-choice menu – I was reasonably confident about my guess of suspects, maybe fifty-fifty on my guess about the general cause of death… and then was confronted with this set of choices for the final question on what specifically killed the victim:

1 - drug overdose

2 - self-inflicted cut injury

3 - adverse side effect of propranolol

4 - adverse side effect of antidepressants

5 - drug interaction

6 - gunshot wound

7 - stroke

8 - heart attack

9 - ruptured esophageal varices

10 - stomach ulcers

11 - intractable vomiting

12 - diverticular bleeding

13 – tuberculosis

Thus the anxiety – as far as I can tell, there’s nothing you can do in the game to get a primer on what diverticular bleeding is, or an NPC you can consult to fill you in on the specific side effects or interactions of drugs. So for me the game ended in a decidedly un-Sherlockian fit of panicked trial-and-error guessing; after four or five tries I eventually got there by process of elimination, but rarely have I felt a less sleuthy sleuth.

The game’s postscript explains what I was missing: apparently the author wrote it as a proof of concept for potentially using IF as a teaching tool for people studying forensics or medical examination. Used as part of a course of study where players would have outside resources to consult and a sense of what questions to ask, this final challenge could actually work really well. This focus on pedagogy also makes sense of some aspects of the game that feel underbaked (I haven’t mentioned the plot or characters that much in this review since there isn’t much to talk about on either head), as well as the rationale for the easy escape-room style puzzles, which I suspect are meant to be there as a pleasant bit of sugar to make the lesson more fun.

But with all that said – yes, this is exactly showing up at the test without having studied, and without even knowing you should have studied and there was going to be this kind of a test at all. So as a Comp entry offered up without more warning or context for the player, I can’t judge Last Vestiges as that successful; nonetheless, I’m intrigued by this approach, and could easily see myself enjoying a future game along the same lines that offers lay people an opportunity to access the information they’d need to be successful, and maybe learn something along the way (I told you I’m a nerd, turns out I want IF Comp to be more like school).

Adapted from an IFCOMP23 Review

The classic closed room mystery makes an appearance! Here, a bloody body with no wounds, and a somewhat spare residence to extract clues from. It’s a parser game, so you are X’ing everything you can find, and asking your partner and the victim’s landlord anything you can think of. It’s a fairly quick play, but man is there a twist at the end.

For most of its runtime you are pushing against a Notably incomplete parser implementation where puzzles are harder than they should be due to terse descriptions that hide lower level details. The most egregious will not let you manipulate objects, but if you X an incidental object, only then moves items around for you. Puzzles like these are tough, because when you find let’s call it the Implementation Horizon, the level below which parser commands yield only ‘you cannot’ or ‘there is no,’ as a player you might conclude ‘this is a dead end, there’s clearly nothing more here.’ But this game will pierce that Horizon randomly with super important low level details amidst a sea of gaps.

It also has a few code breaking type puzzles - one of which is clever enough, but the other defies in-world credulity. It is nominally a mnemonic for a password, but it is the most convoluted mnemonic imaginable and nothing anyone would actually use. But it is a puzzle, and it can be solved, so there’s that.

Conversation similarly has a an Implementation Horizon problem - talking to the two NPCs about everything you find yields information, until you start getting ‘no answer.’ (Ignoring you is a curious choice for this messaging, given you are actively trying to solve a crime the NPCs are notionally invested in!) Made worse with incomplete synonyms. If you ask for details about the victim by his LAST name, you get an answer that suggests there’s nothing to learn. If you ask about his first name, hey, info! (Spoiler - click to show)If you ask about an autopsy you are told its not done yet. But lower level medical details are still in the chamber!

Inhabiting this unevenly implemented world are a protagonist and NPCs that are all but ciphers. The most human personality you encounter is through investigating the victim, and even there the details are spare and incomplete. All of this combines to a kind of representational reality, a parser-based psuedo-world, rife with simplifications, non sequitor logic puzzles, and short-hand logic leaps. Like animators only drawing three fingers because that’s representationally good enough.

Thankfully, the game does come with actually helpful hints to point the way through the darkness, at least as far as the parser-search. When you exhaust your environs, it’s time to solve the murder, via answering Who/What/Why questions. The first two questions reasonably trade on what you might have learned from your first half gameplay. But for that third… it’s like you jumped from an abstract cartoon mystery into the middle of a busy emergency room!

That final question encompasses DEEP cuts of biology, science and tragic inferences. Where most of the game was soft and well-meaning, suddenly the last question was gritty, clinical and super detailed! Boy, if you could have seen those options FIRST, before stopping the inquest… because it turns out, that you COULD ask about a lot of that. Based on the light gameplay, there was no hint that you SHOULD, but you could. I ended up taking reasonably educated guesses and hitting it, but it didn’t feel earned. The fact that that last question was so out of place relative the prior gameplay was shocking and kind of subversively fun. The fact that that level of detail was also implemented in conversation (if you go back and try it) was amazing. Sure, all of it deeply unfair, but amazing. I can’t say it justified the unnecessary struggling or elevated it above Mechanical, but I’m glad I saw it.

Played: 10/9/23
Playtime: 40min, “solved”
Artistic/Technical ratings: Mechanical, Notably uneven implementation
Would Play After Comp?: No, mystery solved!

Artistic scale: Bouncy, Mechanical, Sparks of Joy, Engaging, Transcendent
Technical scale: Unplayable, Intrusive, Notable (Bugginess), Mostly Seamless, Seamless