Fall Comp 2008
Fall Comp 2007
Fall Comp 2006
Fall Comp 2005
Fall Comp 2004
C32 Comp 2004
Older IF News
The IF Archive
The IF Wiki
By Christopher Huang (Writing as “Hugh Dunnett”)
Played On: October 29th (3 hours 30 minutes)
Platform: Inform 7 (Zcode)
Merk’s Score: 9-
Impossibly, the fiddler-shaped weather vane up on the top of the house twists around and begins to actually scratch out a pleasant, simple tune on its little fiddle. As you stare up at it in astonishment, you completely fail to notice the hordes of stampeding Cossacks which trample you to death as they roar off on their way to yet another pogrom.
*** You have died ***
(Edit: The author tells me there are other XYZZY responses in other areas -- which makes sense, now that I think about it -- but I haven’t gone back to try it yet.)
Be warned. While I don’t intend this review to be particularly spoiler-filled (even disregarding that the game is constructed in a way to minimize the effect of spoilers), I do discuss the plot and the design mechanics a little later.
Here’s a fairly well-polished entry from “Hugh Dunnett” (also the protagonist’s name). You play the part of an Inspector with the Dundreary City Police. A wealthy investor for an upcoming stage musical (“Twisty Passages” -- clever), one Frederic Sheppard, has “fallen” from the window of his study, and it’s up to the Inspector to uncover clues and interview the witnesses in an attempt to identify the culprit.
The opening seems sort of confusing, although I have a hunch the author probably wrote it this way for the opposite effect. Initially, Dunnett is hurried along by one of the witnesses (a relative of the murdered Sheppard), through several rooms and with brief introductions to a couple of the other suspects. He is then handed off to the victim’s accountant (one of the two men who found the body), who leads the way to the murder scene. It conveys a sense of urgency and was probably intended to introduce important points in a logical way. Perhaps I just wasn’t paying close enough attention, but by the time I made it down to examine the body, I forgot who I had met and the layout of the rooms I had just passed through to get there.
Luckily, the confusion is short-lived. The Inspector carries a notebook in which he makes note of clues, statements, and suspicious activities. I kept additional notes of my own (including a map -- although the game covers only a small area), and this helped as well. Only five suspects are present in the house, and that helps keep the mystery compact and solvable. Before long, I felt as “in charge” as the Inspector himself should have been, and that’s when the game really started to shine.
An Act of Murder requires some specific detective-like behavior. In retrospect, it’s all very logical, but it’s possible the author has given players a little too much credit in figuring it out. I checked the beginning just to see if the Chief had briefed me on what to do, but it seems very general and pretty open. Specifically, it’s important to ask suspects for their alibi, and about the murder. Asking about the dead body isn’t the same as asking about the murder (which, in effect, takes the guest’s statement), so it took a while (and possibly the hints) before I realized I should be doing these things. I believe this is probably because I had come to think I could only quiz the suspects about physical objects -- clues, which are also important. So, even though it’s logical in retrospect, some clear direction might have helped.
The same goes for reading entries in the notebook. “Read about...” and “Look up...” do make sense (in fact, I’ve done this in other games, including another similarly-themed IFComp entry this year), but without a nudge or some in-game instructions, it took a while before I realized these notes could be reviewed individually as well.
It’s a well-constructed game, but a few minor bugs are noted in my transcripts (yes, this is the nit-picky section). Duffy is supposed to arrive at 2:00 AM, but in my first play-through, it was well after 4:00 (and even then, only after I called). That did give me more time to investigate, but once before when I tried calling him directly, I was told by someone else that he was already on his way. That was a couple hours before the second call, in which Duffy himself answered and said he would be there in five minutes. I had some difficulty with the walking stick, in a guess-the-verb sort of way. Climbing the rocks at the beach will lead Hugh up, instead of northwest. Asking suspects about “me” gives a quirky response.
Here’s an amusing and obscure one, tried on a whim:
>ask body about me
These are all minor issues. An Act of Murder shows obvious polish and attention to detail. In a competition against many games that are poorly tested and haphazardly implemented, An Act of Murder stands out even more.
The author has given the player a mystery to solve, not just a collection quest disguised as a mystery. This is why I found my own note-taking to be of more help than the in-game notebook. There are logistics to work out, motives to prove or disprove, and yes, clues to gather. This comes closer to the feel of solving a mystery than other games I’ve played (although, admittedly, this isn’t my favorite genre), because the PC doesn’t clue the player in on his reasoning. It’s up to the player to put the pieces together and work out a solution and that’s how a mystery game should be. Chief Inspector Duffy does offer additional insight near the end (and this can sometimes come from a brute-force presentation of gathered evidence), but I think the author rightfully expects that if you’ve gathered the right evidence and perhaps summoned Duffy, you’ve worked it out for yourself already.
The game begins with randomized plot elements. This isn’t obvious from a single play-through (even though the author warns of it from the beginning), so it doesn’t take anything away from the experience. I’m not sure if this was done to provide a higher replay value, or simply to thwart spoilers and potential walkthroughs, but I’d probably have been equally happy with a single “right” answer.
Calling them “randomized plot elements” may be misleading. I played through twice (with a different result the second time -- and I made it through much quicker after the first). It seems that each of the five suspects is given both a motive and a counter-motive. Some can be cleared because their stories coincide during the potential murder window (and if the author includes a dual-murderer scenario or a suicide scenario, I didn’t see it), and some can be cleared because they’re given a strong counter-motive for wanting Frederic to live. The timetable of events is adjusted as well (I restarted a couple more times just to see that in action).
The only reason this seems to escape the Clue movie cliché is that only one suspect can be the murderer on any given play-through. Unless more complicated endings do exist, it’s impossible to gather all necessary evidence, reach the end, and still be left with more than one suspect that fits all clues. It stands up well to a single play-through. It’s interesting a second time as well (especially with the existence of new counter-motives and the absence of key counter-motives found the first time). I suspect that after a couple times, however, it would probably begin to feel more and more game-like.
This kind of design has an impact on what’s possible story-wise. It means that either the author has written five different stories (potentially) in the same scenario, or else it’s written in a way that downplays the story so that these adjustments are kept credible. Here, it seems like more of the latter. It’s an interesting mystery, but it co-exists with other variations to the same mystery. The story doesn’t seem to suffer as a result, but I suspect it might have been even better if it focused on a single “right” possibility.
Mystery seems like a really tough genre, though. It’s probably just as possible to write one that’s unsolvable as it is to write one that’s solvable only half-way through on a strong hunch. An Act of Murder is a mystery that works, even though the details are variable. It’s one of the strongest entries I’ve yet to play this year, and is bound to be a high-ranker in the IFComp results.
Interaction among characters is pretty minimal during the game, but they do make a few comments about each other. It’s fitting and sometimes enlightening, but it lacks a strong “punch” (I’m thinking of Sting of the Wasp here). The writing is effective but not flowery, with a good flow that doesn’t draw attention to itself (and that probably comes from being well proofed -- a quality sadly lacking in some other entries this year).
In addition to five (potential) winning endings, five additional “losing” endings (in which Dunnett can play the fool by accusing the wrong suspect) are a nice touch. Also, one odd but funny bit stands out, primarily because it seems like a pun that would be clever if only I knew what it meant:
I voted the game a reserved “8” at two hours (after voting a little too highly on a couple other games that ran longer), but it’s too well-constructed and fun not to rate a point higher. So, my non-vote review score is “9-” The “minus” is mainly for sorting purposes (I enjoyed my other two 9’s just a tiny bit more), a few very minor issues, and a lack of initial guidance on interview techniques. I highly recommend An Act of Murder, and I expect it to finish well in the competition.