The Stack

The demo of Addlemoth was apparently enough to get me jonesing for more neodroddism. I last posted about Wonderquest in 2018; it has spent the entire time since then pinned to my taskbar, awaiting further attention. Opening it again for the first time in four years with intent to actually finish it this time, I found myself in the middle of level 9 (out of probably 20, judging by the “% Complete” stat), enmeshed in a Rube Goldberg contraption of a puzzle, with (predictably) no memory of what I was trying to do or why or how. So I started the game over from the beginning, and got as far as level 7 in a frenzied burst, although my progress is slowing down as the puzzles become more complex and difficult.

I won’t rehash what I said in my previous posts on the game. It’s still less polished than DROD (or even Addlemoth, which is still a work in progress), but worth playing, in my opinion, at least once you get past the first few levels. This is a game that thrives on complexity, building its best puzzles out of masses of moving parts, but it builds up that complexity bit by bit, so it takes a while to become interesting. Particularly if you’re an experienced DROD player and already know the basics of roach-slaying, which, in all likelihood, most of the people who have played it probably were. The only place I’ve ever seen this game even mentioned was on the official DROD forums. That’s where the author published his download links. The forums are still online, but the links are long rotted. I don’t know if there’s anywhere at all you can download it today. I really should finish it while my install is still playable.

When I saw Addlemoth mentioned on Twitter as drawing inspiration from DROD, I had to give it a look. Although unfinished, what’s there is already impressive — not just for capturing what I find appealing about DROD, but for forging its own path. I mean, the only other game I really think of as imitating DROD is Wonderquest (which I should really finish up at some point), and Wonderquest imitates DROD so closely and unquestioningly, in style and mechanics, that it made me doubt you could stray far from the DROD formula and still be noticeably DROD-like. Addlemoth proves that it can be done.

Among the ways Addlemoth deviates from the formula: It ditches the “one big contiguous space” idea; puzzles are entirely self-contained and relatively small. The goal in each puzzle is not to slay all the monsters, but to hit one or more magical crystals. (Which usually involves slaying any monsters that get in your way, but they’re obstacles, not goals.) The default weapon isn’t the equivalent of Beethro’s Really Big Sword (although you can obtain that as a power-up in some levels), but an entirely new one with no precise equivalent even in The Second Sky: you can attack instantly in any direction (without moving), or stand still to parry, which temporarily stuns one attacker. This turns out to be fairly rich puzzle fodder when coupled with enemies that know how to move around obstacles.

One touch I really like: Remember how some DROD puzzles have “Challenge Scrolls” that dare you to complete the room under some voluntary restriction, like “never turn” or “don’t move diagonally” or whatever? EVERY room in Addlemoth has this. The “Conduct” challenge has its own spot in the UI, where it only appears after you’ve beaten the puzzle once. A lot of the DROD challenges were invented after-the-fact by players and only later incorporated into the game, and as a result of being thrust upon rooms not designed for them, they were often punishingly fiddly. And to be fair, Addlemoth Conducts get fiddly too, but not to nearly the same degree. It probably helps that the rooms are smaller.

The one part that’s a bit of a letdown so far is the story: in contrast to DROD‘s vividly inventive grotesquery, Addlemoth seems to be fairly standard CRPG fare, taking its cues from D&D and Japanese visual novels. But that’s never stopped me from playing a puzzle game before. I’m looking forward to the full release.

I’m feeling more sanguine about case 8, largely due to progress made as an indirect result of finally doing what I really should have done a long time ago not just in this case, but in most of the cases preceding it: calling on Sherlock Holmes. I’ve been reluctant to do this because it seems like asking for hints, and I’m always reluctant to ask for hints, lest they reveal too much and ruin the pleasure of figuring things out. Indeed, the game itself warns of this: “Be careful, that help could spoil the fun of investigating!” But in fact every time I’ve tried him, his hints have been very gentle, just nudges to get you on the right track. In particular, here in case 8, essentially all I got out of it was to consult the medical examiner and evidence analysis guy who are among your always-available resources, and who I had neglected to visit in this case simply because there were so many other leads to follow. Once you’re advanced in an investigation, it’s very easy to forget that you haven’t covered the basics.

At any rate, these infodumps provided information that should have been informing my investigation a lot earlier, but I’m actually a little glad that I got them only when I had enough context to interpret them. I’ve got most of a pretty good picture of what happened now. There are still some details that elude me, though, and at this point I’m less inclined to just go for the solution until I’ve satisfied myself. At this point, that doesn’t even mean seeking new information so much as fitting together the information I already have. I’m coming to the conclusion that the way I’m playing this — keeping it open over a long period of time, dipping into it whenever I feel the inclination — is the best approach for me personally, even if it isn’t the style of play the designers intended.

It’s been over three weeks now since my last post, and I’ve had case 8 of Sherlock Holmes Consulting Detective laid out on a table for that entire time. The box suggests “90+ min” as a typical session length, but I’ve found that to be a severe underestimate even for the simpler cases, at least in my current distraction-riddled context. Case 8 — titled “The Thames Murders”, the very case that the entire collection is named for — starts with the warning “This case is longer than the other ones. Expecting visiting [sic] more locations and spending more time solving it.” And this has been intimidating.

It isn’t just that I expecting visiting more locations. The content within those locations trends substantially longer, running to multiple pages in some cases. Where other cases start with a murder, this one hands you five right out the gate. There’s been a worrying emphasis on time-keeping, as if I’m expected to form a coherent picture of multiple people’s comings and goings over a span of days. Plus of course the number of newspapers you have access to, and thus the number you have to scan for possible relevance, has been growing steadily as game-time passes. This is the first case where I’ve felt the need to organize information by taking notes on index cards, one for each character. Although even that doesn’t really suffice to show relevant character relationships — I almost feel like I need one of those conspiracy boards with a web of red string linking photographs.

In short, it’s complex enough to have passed a threshold beyond which I just don’t seem to be able to finish it. Throughout the cases, I’ve been keeping track of locations I’ve visited in a lined notepad, so as to easily consult them again later (as the game allows you to do). I’ve previously limited myself to one page of that pad per case, and usually stopped well short of that. Here, I’ve gone over, and still don’t feel like I’m anywhere near understanding anything. There’s a peculiar thing about this game: it largely relies on narratological reasoning, such as assuming that a thing is important simply because it’s mentioned repeatedly, but it also frustrates it. I know that in a mystery of this sort, the first and most obvious solution is never the correct one. So when I suddenly encounter a new character with a motivation to kill, my first reaction is “Aha! This is the twist I was expecting, and this is the true culprit!” But discoveries aren’t strictly linear, and in a case that’s so generous with its leads, it’s very likely that I instead read the genuinely relevant part first, and only later saw the red herring. It strikes me that one of the things that enabled me to solve case 7 with a positive score was that the path to its secrets did largely form a clear narrative line, rather than the cloudy morass I have here.

I’d probably be happier, or at least have a table free, if I just gave up and read the solution. That’s the thing, you can just end a case at any time. But if I were the sort to do that, I wouldn’t have this blog, would I?

It finally happened: I completed a case with a positive score! In fact, as the gods of game willed it, there was a dramatic lead-up to it, in that I completed Case 6 with a score of exactly 0, my locations-over-par exactly cancelling out the questions I got correct.

Case 6, by the way, is the only one I recognized at all from the Icom FMV adaptations. It involves archeologists seemingly falling victim to a mummy’s curse. There are credulous and sensationalistic articles about it in the newspapers, but basically everyone involved in the case, including Holmes, finds the whole idea ridiculous. This basically why Holmes assigns the case to the players; there’s a pattern throughout the game of him fobbing cases off on you that he doesn’t want to bother with. At any rate, even though I had seen the case before, I don’t think that affected my performance, because I didn’t remember the details at all. If I had, I might have scored higher.

But case 7, now. There, I got all but one of the main questions right — and not just right, but firmly certain in my head, answered without guesswork. (The remaining one, I had absolutely no idea about.) Half the tangential questions, too. Possibly this was engineered. The case just seemed very straightforward to me, and I can imagine that this was a matter of the writers saying “Okay, we’ve hazed the players enough. Now it’s time to make them feel like they’ve learned something.” Or maybe it’s just me. I mean — and here I start spoilers — the central twist here, that the murderer was searching for stolen jewels hidden in plaster statues, is something that I’ve been for some reason anticipating, primed to suspect in other cases where it’s far less justified. Conspicuously mention an object of no immediately clear significance, and my first thought is “I bet that’s where the jewels were hidden!”, even if no jewels have even been mentioned. So when there’s a pile of plaster dust at the crime scene and a recent invoice for a statue that isn’t anywhere to be found, I immediately know what’s up. It’s just a matter of pounding the pavement up the chain of ownership until I have a name.

And it must be said that even Holmes had to engage in basically the same process this time, making for an unusually high par. There have been earlier cases he solved by visiting as few as two locations, which just seems like trolling. This time, it plays pretty fair.

The fifth case in Sherlock Holmes Consulting Detective adds a few new things to the game’s repertoire. For one thing, it’s the first case where the critical path actually involves using the map to interpret clues. For another, it’s the first case involving Moriarty. I’m kind of glad he’s showing up here, because him appearing was pretty much inevitable — for all that Moriarty is a marginal figure in the original Holmes canon ¹Fun fact: The Sherlock Holmes fandom was the first to use the word “canon” in its fannish sense. Really, they pioneered a lot of the customs of modern fan communities., dying in his first appearance, he looms large in the Holmsean imagination, and could hardly be left out entirely — and putting him in the middle means the writers aren’t taking the cheap route and saving him for the climax. Now, The Valley of Fear, one of the few canon stories to mention Moriarty, starts off with Holmes receiving a coded message from one of Moriarty’s agents, whose conscience is just troubled enough to make him send Holmes the occasional pseudonymous tip. This man uses the name Fred Porlock. SHCD expands Porlock into one of your always-available resources, just like your contacts at Scotland Yard and the Old Bailey, but a great deal less generally useful, because he knows only about Moriarty’s activities.

And, as in The Valley of Fear, the case starts off with a coded message, although not from Porlock. This is another new element: the static equivalent of a minigame. That is, like a minigame in a videogame, it’s a break in the game’s main activities, a self-contained segment where you have to think in a different way. I liked this at first, not just because it provided some pleasing variety, but because it seemed to be applying it well. To start with, unlike a lot of mass-market puzzlefests, it doesn’t expect that the player is unable to crack a simple substitution cipher without being handed a key. That is, you can go to Holmes for help if you want (Holmes himself is another one of those always-available resources), but recall that the game discourages taking more hints than you need. And anyway, Sherlock Holmes fans should all have some idea of how to solve cryptograms, seeing how Holmes himself explains the basics in The Adventure of the Dancing Men. But then, once you’ve deciphered the message, it turns out to be gibberish, just semi-grammatical word salad. Because obviously no one would encode the real message in such an easily-cracked way.

Getting at the note’s second layer almost necessitates consulting Holmes, and even then, he doesn’t tell you what to do directly, preferring to watch you flail about helplessly. Honestly he’s kind of a jerk in this game. I’m willing to believe that this is one of those puzzles best suited to a team effort, with multiple minds that work in different ways. Working by myself, I did at least find the signature, and that was worth a few points, but missed the actual content.

At this point, I’m really leaning toward the idea that you’re not really meant to solve these cases, that you’re supposed to try to solve them and then let Holmes actually solve them for you. I intend to keep going out of sheer stubbornness, but this isn’t really to my taste. It makes me wonder, though, if a similar format could be used for a different detective with results more to my liking. Someone whose cases are less about the brilliant revelation that makes everything clear, and more like a police procedural, where you keep on finding more confirmation the more you dig. Columbo, maybe. You’d have to alter the rules for this, though, because the current rules discourage the style of detection where you keep digging and digging and find more and more confirmation of what you suspect.

I said previously that the structure of Sherlock Holmes Consulting Detective, and the static nature of the text, means that the basic resources available in all the cases can never answer questions you didn’t have at the very beginning of the case. I should have said “shouldn’t” instead of “can’t”. It took until case 4 for me to encounter out-of-sequence information, but it happened.

The first four cases have something of a progression in Holmes’ involvement. In case 1, he’s active for the entire introduction, handling the initial questioning before sending you out on your own. Case 2 is similar, but Holmes talks less — admittedly, mostly because Inspector Lestrade spends most of the scene delivering monologues. In case 3, he actually rushes out the door before the intro is over, trusting you to handle things while he pursues a completely different case. (I wish his trust were more warranted!) And in case 4, he doesn’t show up at all, just leaving you a note indicating that there’s something of interest in the day’s newspaper.

This leaves some ambiguity. There are two articles in the paper that present apparently separate mysteries: one involving a pair of dead lions that appeared in Hyde Park, the other of a sailor who fell dead in the street for no obvious reason. Now, the case is titled “The Lionized Lions”, so it’s obvious to the player which item is the important one. But the player characters sometimes express confusion about it, and will ask the people they question about either or even both mysteries, as appropriate. And if you pursue both investigations far enough, you can find a suggestion that they’re connected. So I felt somewhat cheated when Holmes’ analysis didn’t involve the sailor at all, and none of the end-of-case questions had anything to do with that side of things.

I feel like this entire extended red herring is the game turning up the difficulty, on the assumption that you’ve got the basics down now and need more of a challenge to keep things interesting. But of course I haven’t got the basics down. This is something that a game on a computer can deal with a lot more easily than a game made entirely of paper. A truly interactive mystery can test you and make sure you don’t progress faster than your ability. Heck, even Pandemic Legacy slows things down when you fail.

Sherlock Holmes Consulting Detective doesn’t simply expect you to solve cases for the pleasure of figuring them out. No, that stuff is gamified. When you’re satisfied with your investigation — and only you can say when this happens — you turn to the back of the case booklet and answer a series of questions, including “Whodunit?” but also including other pertinent details. Each correct answer scores a number of points indicated with the question, with tangential questions scoring lower. Then you subtract five points for each location you visited over par. (Equivalently, you could say that every location costs 5 points, but you get a certain number of points for free.)

In three cases, I have yet to score above zero, and have not once correctly identified whodunit.

It must be said that this may just be the result of a solo attempt at a game intended for a team. More eyes means fewer details go unnoticed, right? Except that the one time I did attempt the game with a team, it didn’t make a lot of difference.

There’s another factor, though. The text casts the player(s) not as Sherlock Holmes, but as sundry unnamed Baker Street Irregulars, former street urchins led by the now-grown-up Wiggins, studying under Holmes to learn his methods. When the optimal solution is presented, it’s presented as the solution Holmes came up with. His solution is always worth 100 points by fiat; it is this that establishes par. So, you can see the player’s failures as an essential part of the Holmes story. The point is not to equal Holmes, but to stand in awe of him. To get it wrong so he can impress you by getting it right. The main difference from conventional Holmes stories, then, is that a conventional story presents you with all the information the author wants you to have, while SHCD lets you miss crucial data entirely, and what’s worse, implies that it’s your own damn fault. Someone will make an offhand mention of going to the theater, and if you don’t take this as a cue to visit that theater yourself, you won’t see the thing that makes sense of the whole case. Sometimes the questions at the end will reference characters you’ve never heard of.

So it’s something of a guessing-game, and something of a maze — each node leading to other nodes, some of those nodes being dead ends. But it’s a peculiar sort of maze, in that there’s no exit, no definite goal. And yet, I do have a sense of progression.

In case 1, I failed because I went for the solution too soon, hoping to maximize my score by minimizing locations visited and falling for what I now see as a fairly flimsy red herring. In case 2, I corrected and even overcorrected that mistake, but still failed to interpret the clues correctly. In case 3, even though I missed a crucial turning, I at least managed to see the significance of all the physical evidence at the crime scene, figure out one of the story’s major revelations, and answer fully half of the questions at the end correctly. The points from those questions were entirely wiped out by visiting too many locations, but I’m less worried about that: once I actually solve a case, I can start thinking about maximizing my score. The point is, I’m learning how the whole thing works. If it’s still too dependent on reading the author’s mind for my liking, at least I’m getting better at reading the author’s mind.

This kind of progression also fits the narrative frame well. You are, after all, supposed to be learning. I’m not sure I believe any of this is entirely deliberate — it seems to me quite possible that the author(s) intended each case to be solvable but underestimated the difficulty and didn’t do enough playtesting. But the possibility makes me wonder what else today’s narrative designers could do with the format.

I’ve had an unopened copy of the first volume of Sherlock Holmes Consulting Detective on my board game shelf for some time now, and decided to finally give it a solo try during the week between Christmas and New Year’s. This isn’t really my first experience with the game, though: the 1991 FMV game based on it was one of the first CD-ROMs I purchased, back when CD-ROMs were new and exciting. Apparently the cases in that overlap somewhat with what I’m currently playing, but it’s distant enough that I remember nothing. Also, I played one case from a later volume with a board game group a few years ago. The group failed to solve the case, and, on reading the solution, declared it to be basically nonsense, full of suppositions that didn’t really follow from the evidence, and so we didn’t do any of the other cases. But the fellow who had brought the game told us that the earlier volumes were better in that regard, and that’s why I bought the first.

Although it’s packaged like a board game, and the instructions describe it as an activity for multiple players, it’s fundamentally more like a gamebook, a work of semi-interactive fiction like a Choose Your Own Adventure or Fighting Fantasy book, based around reading text passages in a nonlinear, reader-determined sequence. The closest thing it has to a game board is a map of London, marked up with numbers for all the locations referenced in the cases, but this just a reference tool, and hasn’t even been useful in solving the two cases I’ve gone through so far. Structurally, the main thing that separates it from other gamebooks is that the passages aren’t even partially ordered. Instead of each passage linking to its own set of what-happens-next, each passage corresponds to a location (most locations being linked to a person) that you can visit any time you want. (In other words, it’s more like Her Story.) This does not mean there’s no sense of progression whatever: gating is done through player knowledge. You start a case knowing that A, B, and C are persons of interest; you look up A’s passage and he mentions person D, who you didn’t know about previously but who you now know to also visit. (Passage lookup is mediated through a lengthy address directory, shared by all the cases, to maintain secrecy about what things each case makes available.) It does, however, mean that, as the game state is in your head, it cannot affect the content of passages. There are certain resources that are always available from the beginning — public records, informants, contacts at Scotland Yard and the Foreign Office and so forth. It’s tempting to consult them about things discovered late in a case, but due to their position at the head of the discovery tree, you can only ever ask them about the very basis of the case.

Although the game is clearly part of the same lineage as the Dennis Wheatley crime dossiers, the emphasis here is, as I say, on deducing from things described in text, rather than on examining physical evidence included in the package. In that regard, it’s a lot like reading the stories that inspired it and trying to solve the mystery as you go along. The closest thing it has to feelies is the enclosed single-sheet newspapers, which are still text-based, but are nonetheless artifacts from the world of the story that you can hold in your hand, providing physical proof of the fiction’s reality. Most of the articles in the newspapers are irrelevant to the cases, giving an impression of a larger world of which the cases are only a part. Best of all, though, they provide interlinking. The cases occur on specific dates, in sequence, and can use information from not just that date’s newspaper, but also all the newspapers from all earlier cases. One of the best moments the game has delivered so far was in case 2, when I speculatively looked back at the news for case 1 just to see what I could find, and learned that a man I wished to find and question, as he had been corresponding with the deceased shortly before his murder, had been declared dead several months earlier.

I’ll have more to say about the experience tomorrow, after I get another case or two under my belt.

A Monster’s Expedition wasn’t the only sokoban-like to get an update with a whole lot of new content recently! Baba Is You got its first truly major update a couple of weeks ago, including a level editor and two additional campaigns designed for people who completed the original. One, called “Museum”, is a collection of unused levels (although I don’t think they can really be called unused now that they’ve been published) and new variations on familiar ones, and it uses the same collections of words and concepts as the original game, except for one addition: signposts containing developer commentary. The other, “New Adventures”, introduces a number of exciting new concepts, and this is by far the more interesting part to me.

New Adventures introduces quite a few new object sprites, including simple geometric shapes (Circle, Square, Triangle), animals (Dog, Cat, Bird), food items (Banana, Donut) and even a couple of new friends for Baba and Keke: the shaggy Fofo, the llama-like Jiji. I haven’t really warmed to these interlopers. Somehow the new “characters” feel like they go against the grain more than all the other objects, even though they’re exactly equivalent in gameplay terms: like all sprites, they’re just empty containers for properties assigned by rules. The only new objects that really have special properties of their own are a set of musical instruments, which can be made to play notes, a pure amusement that doesn’t affect the puzzles.

No, the interesting part is in the new attributes and relations. There are many. The relation EAT, for example, lets one type of object consume a specified other type, which is elementary enough that in retrospect it’s a little surprising that it wasn’t in the original — the closest it came is probably the OPEN/SHUT rules, which were considerably less direct. Another example: BROKEN, an attribute that cancels out all other attributes. This has obvious applications as both a tool and a hindrance, which is always a good sign for a puzzle game element. Sometimes a single BROKEN rule can even be both in succession. I think my favorite addition so far, though, is FEELING, which makes a rule conditional on an attribute — for example, FLAG FEELING MOVE IS WIN would grant flags the WIN attribute when and only when they also have the MOVE attribute. This not only enables all sorts of wacky conditionals, it also makes a place in puzzles for attributes that would otherwise have no mechanical effects. The original campaign had one room with the attributes RED and BLUE, and all they did was change the color of sprites; now, you can have rules where changing an object’s color also changes how you can interact with it.

The main disappointing thing about the game has always been that, because the designer doesn’t like re-using tricks, a lot of concepts are introduced, used in a small number of levels, and never seen again. The new content happily gets some extra mileage out of previously-underused words like GROUP, but the newly-introduced words can fall into the same pattern. There are a few cases where I consider this positive, though. There’s a new attribute that can make turns advance automatically, transforming the game from turn-based to real-time — useful for those musical amusements, but irksome elsewhere, because it transforms it into a different kind of game, one that it’s not well-suited to being. The attribute that shifts you into a first-person 3D view has this problem even worse. It’s a 3D view based on the same blocky and abstract sprites as the regular game, creating a sort of Bard’s Tale-but-more-primitive look, which means it’s difficult to judge distances and you essentially have to memorize the level layout while it’s still 2D and comprehensible. Also, this won’t be a problem for most people, but in my current circumstances, the concept of YOU2 is troublesome. With my right hand still sore from injury, part of the appeal of the game is that it can easily be played left-handed, just like A Monster’s Expedition (and for the same reasons). So what does YOU2 do? It gives you a second player-controlled entity, one controlled by each hand. Here’s a feature request: Provide an option to toggle between controlling YOU and YOU2 with a button press. First-person mode already has a similar ability to switch between YOU instances. Consider it an accessibility feature.

Anyway, if you haven’t started on the new content yet, I can’t fully recommend doing so, because it’s still receiving frequent updates, often more than one per day. I recall the original levels had a similar problem when they were new: some levels even got renumbered, replaced with simpler versions and shifted into bonus levels, or even removed entirely. (You can now play the removed levels through the Museum.) Back then, I ultimately restarted the game just to make sure I saw each level in its final form. The new stuff seems a bit stabler than that: levels change internally, but they don’t move around the map. Or maybe they do and I just haven’t noticed.

Still, there’s gobs of delightful new stuff. The New Adventures could have been released as a standalone sequel, and I would have paid for it, and it would have been worth it. Instead, it’s a free update.