Ratings and Reviews by OtisTDog

The Owl Consults is a game that I first came across while exploring the history of the XYZZY Awards but which I had never really heard about elsewhere. When it ended up on the chosen field for the Great Play Marathon, I couldn't resist the chance to check it out.

The game placed 6th in the 2017 IFComp, outshone only by Absence of Law, Will Not Let Me Go, Harmonia, Eat Me and The Wizard Sniffer. It went on to be nominated for four XYZZY Awards, contending for Best Game, Best Puzzles, Best PC and Best NPCs. It was clearly enjoyed and appreciated by many people, and it is still enjoyable almost a decade later. Per IFDB data, it is the 17th most-played title from its year of publication.

The game starts off very well, beginning with the blurb, which hooks via its unique premise. The title screen of the game primes the pump with a T. S. Eliot quote that evokes a serious air, though the reason for this particular allusion is opaque. (The quote is from a poem that T. S. Eliot wrote while converting to the Anglican Church.) Opening text quickly establishes a 1930s world with "superscience" elements of the type found in early scifi from that era, plus a player character who reveals almost nothing to the player. There is little time to get situated in the role of this enigmatic protagonist -- the phone is ringing, and we know from the blurb that answering it is what the game is about.

Doing so connects you to one of the Owl's clients, a jaded English catburglar named Amelia Derringer, who very much reminded me of Lady Thalia (though of course this character came first). There are significant differences, most notably that Derringer has a superpower: (Spoiler - click to show)the ability to gain the features of any animal that she consumes for a short period. She's been trapped in a cage by that apparently frequent nemesis of supervillains, Rex Dashing, who on first report sounds like a typical pulp-era science fantasy adventure story hero.

The advice that you must give to Amelia is extremely granular in nature and really seems like something she would think of herself, so at this point the structure of the game suggests that play will consist of a series of vignettes, with the player temporarily taking on the persona of a different supervillain in each. This is a clever way to shape a superhero game, giving the author license to introduce various protagonists (with their differing superpowers) in short, focused puzzle scenarios that don't require much in the way of framing story.

Once out of her initial jam, the call with Amelia is interrupted by that of another client: Dirk Radon, an overconfident mad scientist type with a knack for technologies based on atomic energy, whose superpower is (luckily) complete immunity to radiation. The tone shifts significantly as Radon's narration takes over, but it continues in the light comedy puzzler style. Radon's problem is that he's hemmed in by robots aboard an airship, which he built but which was stolen at some point by Rex Dashing. At this point, it seems as though each vignette might involve a villain trying to deal with Dashing, functionally making that character the primary antagonist of the Owl. The interruption of the first call also suggests the possibility of future puzzles that involve juggling multiple lines as the clients on the other end deal with timed challenges.

Soon enough, it becomes clear that both clients are aboard the airship Mephistopheles, and at this point in playing, the gears really began to turn in my head. If both clients and (as I had mistakenly thought) the Owl are aboard the same airship, does that mean that the Owl is Rex Dashing himself?? (Spoiler - click to show)The reason that I thought the Owl was on the same ship is the game's subtitle: "Aboard the Mephistopheles," which I had read as a continuation of the title text, i.e. meaning "The Owl consults aboard the Mephistopheles." Derringer's introduction mentions an airship and robots controlled by Rex Dashing. Radon's introduction is unambiguous that he is aboard an airship in Dashing's control and surrounded by robots. It is soon confirmed that both are in the same place, and this seems to surprise the Owl not at all. As gameplay progresses, Dashing (who always seems to be one step ahead of the villains) looms large as a frequently-mentioned but mysteriously absent figure, and he seems less earnest than one might expect based on the standard tropes. Could it be, I wondered, that the Owl/Dashing is putting one over on both of these "clients" as a way of defanging their plans, and/or romancing Derringer, and/or humiliating Radon for the fun of it? The high quality writing (about which more later) seems to carefully nurture the ambiguity that would allow for such intricacies, but this seems to be either accidental or the residue of abandoned ambitions.

It seems necessary to stop at this point and fast forward a bit: This game is not that complex. In the end, what's delivered is pretty much exactly what is promised, i.e. the Owl, an accomplished supervillian, helps some less capable villains for money. You should be prepared to enjoy it primarily on that level, or risk being disappointed as I was by its unfulfilled potential.

The core game design problem is handled very well. The two clients' subplots interact, each villain contributing to progress using a combination of capabilities and knowledge that makes sense given their revealed personalities and histories. Some additional difficulty is imposed on the player by the fact that the two clients despise each other and must be kept out of the same location, but the geography of the map is simple enough to remember, and it's easy to build the habit of shunting one character into a "storage" location before switching control the other.

The two subplots are coordinated into a double climax that is the top of the rising action, and the authors correctly understood that any obstacles following it should be very low difficulty in order to keep the action moving and present a satisfying ending even the first time through. The countdown sequence that creates a turn limit feels scary and creates some urgency, but the essential solutions are well-practiced by that point, and if the player has investigated everything along the way and bears in mind that there's no actual need to rush while entering commands, there is more than enough time to reach the optimum ending even with some false steps. It's to the authors' credit that you might not want to stop and think much between prompts.

At the level of individual puzzles, things go less smoothly. The reactions of other reviewers suggest to me that players unfamiliar with parser conventions were drawn to this game, and that for some of them their amusement gave way to frustration before reaching the end. Although the current release (1.02) has substantial hinting by the standards of experienced parser players, there are some missteps that I think hamper its approachability. Examples include: (Spoiler - click to show)1. Although its clear that the vacuum flask can be moved, players paying attention to room descriptions are unlikely to deduce that it's possible to push a heavy piece of equipment up a spiral staircase on rolling casters. 2. It's not clear why the ice on the vacuum flask won't melt in other than extreme heat; perhaps it should be described as having an atomic-powered refrigeration unit. 3. Some reasonable commands (e.g. >STAB MONSTER) aren't accepted when trying to defeat the sludge monster. 4. It's not obvious that Radon would want to give his radio speech a second time, especially since points are awarded for the first time and there is no player reward (in terms of new content) for experimenting by trying it twice. 5. It seems like there could be a story-appropriate alternative solution to keeping Radon bottled up while the reactor door alarm is going off: being trapped by the octopus. The trap door control seems to have no purpose, but Derringer can get out of the room (using powers) without closing the tank while Radon cannot. 6. Given the response to trying to acquire animal powers from the taxidermy specimens, it's not clear why it should be possible for Derringer to make use of a "fur-lined scarf." This doesn't seem like it's intended to be a difficult game, so I don't think more consistency and/or hinting would hurt.

The writing is the game's single greatest asset. The fact that both clients are in the same location allows the description of locations and objects (and even their naming) to differ between the two villains, and this is used to great effect for characterization. Whichever author(s) came up with the prose knew how to maximize the value of each word the player is given to read.

As an example that I found striking: When Derringer reaches a place that she calls "The Opulent Room", the response to >X CHAIRS is "The chairs are, much like Dashing himself, overstuffed and uncomfortably smooth." In eleven words, the line accomplishes three goals: giving the player an idea of the chairs' physical nature, characterizing the off-screen Rex Dashing's habitual demeanor, and -- via the judicious use of the word "uncomfortably" -- reinforcing the implication (also found in several other responses) that Derringer is attracted to Dashing. With the demonstrated capacity for such pinpoint accuracy in the prose, I can only assume that the rougher spots found elsewhere, particularly the somewhat disjointed final passage, are the product of a rush to complete the game.

The game's single greatest weakness is its handling of the PC/player relationship. Extraordinarily little information about the Owl is given outside of the final text. The ultrabrief "scene" in which the player mentally inhabits the Owl presents no personal details outside of a sense of self-satisfaction, and the only time that I recall the Owl's thoughts appearing in the narration is during the tail end of a multi-turn rant by Radon.

In actual play, the interactor is invited to "be" not the Owl but Derringer and Radon. By requiring both of these personae to be adopted by the player -- especially more or less simultaneously and in a manner requiring coordination of their efforts -- player identification with each sub-PC can be only partial. Since the blurb and opening scenelet sell the player on identifying with the Owl, and since the Owl is still ostensibly the one issuing commands to the two clients, the player is left trying to primarily identify with almost nothing at all, and can't help but wonder why. Given the strong characterization of the two other villains, treating the Owl as an AFGNCAAP feels out of step, so one instinctively makes an effort to pierce the veil of mystery that seems to have been intentionally drawn. (Even the sound that an owl makes prompts the question, via homophonic onomatopoeia: "Who?")

The main point of including a mystery in a narrative is to allow for a reveal, whether that happens in the text itself or in the mind of a reader fitting clues together until they form a solution. The Owl Consults seems to be unintentionally mimicing the latter type, at least until the very end of the game. The reveal that's provided rather roughly cuts off loose ends and, by returning the player to the Owl persona only briefly and only to lay out future plans, suggests a wrapping up of the Owl's saga before it has really begun. Although the Owl (Spoiler - click to show)admits to prior manipulation of the clients and plans to take advantage of them both in an indirect fashion, the Owl's victory doesn't feel like the player's victory, since it was impossible to anticipate.

What I'm left with is that I enjoyed this game and think it's worth both play and study, but I don't understand why it seems to invest such work into creating narrative possibilities that it has no intention of realizing. Players will find an amusing diversion with some original flair. Would-be authors will benefit from examination of the writing technique for its characterization, which is exemplary, and also the game and puzzle design, which are above average but still leave room for improvement. It's certainly one of the better superhero games that I've encountered.

Sequitur, the first release by author Nigel Jayne, is an unusual work perhaps best described as a one-room police procedural.

The player inhabits the persona of Detective Cochone, stereotypical grizzled veteran of the homicide police. His hard life has not been improved by the multiple murder case that's his current assignment: several people found dead in an abandoned mansion, with the one survivor who might have information currently in a coma. The only clues are various bits of video footage and other information recovered from devices found on the scene, including a video camera and phones. Due to damage to the devices (or some other plot contrivance), the recovered footage consists of only fragments for which the order is not known. Cochone's (and the player's) job is to review the footage and attempt to reconstruct the sequence of events. A tech expert, Jenny, stands ready to assist as needed.

After a brief introduction to the scene, made somewhat chaotic by the attempt to conduct both character introductions and player instructions ASAP, the game proper begins. Individual clips (of which there are only 15) must be reviewed, sorted and rearranged. As this is being conducted, the PC sometimes chooses to interact with Jenny, and various bits of background business unwind automatically to give some life to the framing story. I have to say: Upon initial orientation, I wasn't very hooked by this game, but that changed pretty quickly once I got going with the main task.

The initial review yields a jumble of confusing impressions, many seeming entirely disconnected. (I didn't find out until after the fact that this game was made for ShuffleComp, and it's clear that the rather tenuous framework for the mystery is disjointed at least in part due to the need to try to reconcile the disparate elements of a random playlist serving as the seed.) The victims are clearly trespassers, and they were clearly murdered, but it's not immediately obvious by whom and why.

The game soon communicates that there are two sequences to be assembled, one consisting of ten segments and another of five, but there is a huge number of theoretically possible orderings for the two. Jayne put substantial work into creating quality-of-life features to assist the player in this task, including allowing the player to rename segments, create summary notes for them, and rearrange them in a number of intuitive ways. When the player gets portions of the sequence right, the PC will note this and explain the reasoning, and subsequently the related segments can be treated as a unit group. As a result, much like with a jigsaw puzzle, connected groups of segments can be added to and rearranged as blocks, so each connection significantly reduces the search space and speeds the process of finding more.

This is a very clever mechanic in that it automatically handles pacing in a manner that is self-adjusting to individual players. Making progress through sheer luck isn't very likely in the early stages, when juxtaposing any two in correct order is a low probability. No matter which thread of imagery or action intrigues the player, it narrows the search to a subset of other pieces -- again similar to a jigsaw puzzle, in which the same applies to details of coloration or shape. With each "click," progress is made.

Two things stand out about this to me: First, when I took a moment to think about what the match confirmation mechanism involves, I was wowed by the number of cases that have to be handled. Responses are not always phrased as matches between a pair of individual segments; the PC might notice a run of three or more that fit together enough to be a coherent block. I doubt that this was handled in an entirely brute-force manner simply because the number of cases to cover would be so high, so there must be some way of aggregating links into larger descriptions. Regardless of how it was achieved, it is very well-done -- the description of Cochone's insights does not come across as any kind of mechanically-generated text.

Second, there are apparently different ways to put the sequences together. However difficult it was to pull off the central mechanic, the complexity jumps substantially when that possibility is taken into account. My hat is off to Jayne for this accomplishment, and he's earned my gratitude for the willingness to share the game's source code. (Very fortunately, though the author's personal website is defunct, the complete source appears to have been captured by the Wayback Machine, at least in unformatted text format. An HTML format version is at least partially lost.) I haven't studied it yet, but I am looking forward to it.

The implementation of the central mechanic is not perfect; I encountered a few run-time errors while playing. However, the game is resilient and makes it easy enough to recover from these hiccups, so they do not significantly impede completion. The author has explicitly licensed the mechanical aspects (but not the story text) under the Creative Commons license, so would-be authors are invited to dissect it and adopt any useful elements. (From what I've seen so far, the code is well-documented. Note, however, that it was written for Inform 7 6L38, so some minor rewriting might be needed in places.)

Once the footage sequences have been finalized, the endgame begins, and this is kept interesting through the device of asking the player to demonstrate some understanding of the total scenario revealed by the assembled footage. Cochone, who is (Spoiler - click to show)suffering medically and about to lose consciousness has only a few turns to try to communicate vital information to his superior officer. (This is done with a command prompt to eliminate the possibility of a lucky guess from a menu.) The ending that follows is improved by providing the correct information, but the story resolves satisfactorily even without this cherry on top, so only those interested in solving the mystery need go back if they haven't figured it out already.

My initial inclination was to rate this a solid three stars, but the central mechanic is unique in my experience and urges a rating of four stars on the basis of notability. It would be much more solidly in four-star territory if it were bug-free, the pacing of the introduction were a bit less hectic, the tone (Spoiler - click to show)(which wavers between gritty police mystery and horror) were more focused, and the individual stories created better synergy in their combination. (Although it's not bad given the central challenge of Shufflecomp, the game itself would have benefited from trimming the (Spoiler - click to show)apparent child prostitution subplot, which feels artificially grafted onto the main scenario. Also, the bad guys, presented as (Spoiler - click to show)generic "terrorist" bogeymen, are not very interesting.) This collection of demerits is enough that in the end I'm going with my gut instinct of three stars, which I remind the reader counts as "good, not great" in my book, and means that I'd recommend it as a play experience.

Continuing a string of Western games, I decided to try this one from 1995 -- the dawn of the hobbyist era. This distant origin is immediately evident when one sits down and begins Spur; its comprehensive failure to invite players to actually, you know, play is a shock to modern sensibilities.

The action begins in media res with the protagonist in the middle of a showdown with someone named Roy Heffer. It's kill or be killed (ending the game) in the first three moves -- there doesn't seem to be any way to de-escalate the situation or otherwise escape the necessity of gunning down this man, whose conflict with you is of vague and uncertain origin. This inevitability becomes clear after a dozen quick restarts, so... goodbye, Roy.

As a player, you might feel that at least you have the justification of self-defense, but local law enforcement seems to disagree. Within about ten turns, they'll arrive and haul you to jail (ending the game). Per the walkthrough, in that time you're supposed to figure out that you can (Spoiler - click to show)recruit the town drunk to move the dead guy's body, giving access to the dead guy's wallet. How you're supposed to figure this out without consulting the walkthrough or source code is not at all clear, but it's an essential step.

Following that... wait. You know what? I was getting ready to go through the list of every problem encountered, but there's really no point. As far as I can tell, this work is entirely composed of gameplay problems. It demonstrates just about every bad old school habit I can think of: withholding information from the player that is easily available to the PC, zero flexibility in overcoming obstacles in other than a single prescribed manner, limited time access to critical objects that the player has no reason to expect will be necessary (or even present in the scene), lack of explicit description of a key exit from a room, a described but inaccessible exit from a room that should go to a key location, an inventory limit that irritates due to its pointlessness, moon logic solutions to key puzzles, and more -- and that's all within the first 50 moves. (50 consecutive moves, that is -- hundreds of cumulative moves across sessions.) After using the walkthrough to get through the first handful of mind-reading exercises masquerading as puzzles, I gave up hope of encountering an actual playable part and just followed it through to the end.

All of those flaws are excusable in a game from 1995, since the model of near-constant frustration established by commercial producers was the only model anyone knew or expected. That doesn't, however, make them justifiable -- or fun. (For those who've read my review of Spur's contemporary Christminster: Boy, howdy, does that seem like a player-friendly paradise compared to this.)

Spur is written in Hugo, a language that never seems to have quite caught on. Any system needs its "admirable example" to encourage interest among prospective adopters, but this work does not fit the bill. That said, the source code has been made available, and perusing that was significantly more amusing than the game itself. (Favorite line, which seems exceedingly unlikely to be encountered without knowing about its existence: (Spoiler - click to show)"Clown-killing asshole," he mutters, before wasting you.) It was the source code that prompted the title of this review; for some reason, there seems to be an awful lot of attention to detail paid to being able to kill almost every living thing the PC encounters in the game universe. (Also because (Spoiler - click to show)pretty much everything the protagonist does is more along the lines of scoundrel than hero.)

The feature that stands out most positively about Spur is its innovative response to the >SCORE command, which relates the current state of progress in a kind of summary of the action so far. The way it's phrased seems to imply that there might be other ways to handle certain aspects of the scenario, but that's illusory -- the source shows that very little variation is possible. Still, it's a neat effect that I haven't encountered elsewhere, and it's one that might have broader applicability to story-forward games that don't provide a numeric score.

I can't recommend this game to play, but it is worthwhile to examine as a target-rich environment of solvable problems. The gap between its current state and a potentially fun version isn't really that large, and one could make an exercise out of identifying the shortest path to get there. It's also a solid Hugo implementation of a fairly complex scenario, so if you're interested in trying Hugo you can probably get a good feel for how well it fits your style within 20 minutes of skimming the well-organized source.

I've been on a Western streak for IF lately, and The Song of the Mockingbird is one of the more prominent entries in that genre. Arguably, it is responsible for kicking off the largest wave of Western works in the hobbyist era of IF -- fully a quarter of all titles listed under that genre in IFDB follow Mockingbird's publication in 2021, when it placed 3rd in IFComp and tied for second in that year's Miss Congeniality vote.

Consciously modeled after old-time serial adventures for the cinema, Mockingbird claims to be episode 12 in an ongoing series, and it makes significant reference to events from (non-existent) recent episodes that directly inform the plot arc embodied by this work. To summarize: the hero, "Boots" Taylor, has fallen in love with the beautiful performer Rosa, who has been abducted by the scurrilous Black Blade -- a local villain whose most recent dastardly plot the hero was not quite able to stop entirely -- and he means to free his love from the clutches of this evildoer.

The only complication is that Boots doesn't have his gun, which was stolen by members of Black Blade's gang the night before. Right off the bat this work eschews some of the most elemental tropes of the Western, in the process creating much potential for action in the medium dry goods puzzle-solving mode so familiar to parser IF. It's a clever setup, and it works really well when backed by author Mike Carletta's engaging prose.

After overcoming an initial ambush in an arroyo, the action moves to the abandoned ranch that the gang is using for its hideout. Carletta does an excellent job of conveying a fairly complex space here, one that features important three-dimensional elements. The hero (via the player) must use his wits to overcome the trio of remaining gunmen protecting Black Blade and the hostage Rosa, who is in the house at the center of the ranch, always tantalizingly near but out of reach across a no-man's land of open ground that offers no cover for the hero. This mid-game is on the whole extremely well done; my one quarrel with it is that a key item is "hidden in the scenery," i.e. only disclosed after the player chooses to >EXAMINE a particular item in the room description. (Spoiler - click to show)(It's the branding iron, found in the stables only after (Spoiler - click to show)>EXAMINE RAFTERS.)

A minor rant for the benefit of would-be authors: (Spoiler - click to show)Although this "search everything" style is a classic old school parser trope, it's really the weakest form of puzzle (if you care to dignify this type of obstacle with the name), since it forces the player to lawnmower everything in sight, usually with at least three verbs: >EXAMINE, >SEARCH and >LOOK UNDER. I don't generally have a problem with this type of activity when there is a payoff in cases where nothing is found, i.e. when the author uses the descriptions yielded by the activity to further exposition, mood, etc., or when it's an integral part of the scenario (e.g. a forensic investigation). (In such cases, the activity becomes interactive storytelling instead of just drudgery.) Otherwise, it's still mostly OK so long as the author has taken pains to cover every mentioned noun with corresponding objects. (I don't exactly like it, but I accept it as fair.) Far worse than either of those approaches is a spotty object implementation that doesn't cover all mentioned nouns but still expects you to be investigating them routinely. Parser errors (e.g. "You can't see any such thing."), even inventive replacement versions, are never entertainment when repeated with frequency. The point of a parser error is to train the player to not do that in this game, and an author penalizes the cooperative player when making the applicability of parser errors inconsistent. If you want to hide items, better to hide them "in plain sight," i.e. as part of some tableau of objects about which the PC reasonably might not perceive every detail at once. (Key word: "reasonably." Having to >EXAMINE a table or the like to see what's on it is ridiculous and thankfully does not occur in this game.) Failing that, some gentle prompting to direct player attention (e.g. in this case perhaps the rafters occasionally creaking in the wind) or eventual automatic disclosure (e.g. the PC "happening" to spot something after so many turns) is called for. An author never ruins any genuine puzzle by directing players' attention to the components they need to solve it.

Once entering the endgame, the plot is resolved in short order -- regrettably, too short of an order. The game hints so strongly about what the player should be doing that it might as well be pure cutscene. The larger problem with the climax is that it hinges on a twist revelation that the player can't meaningfully absorb, resulting in a wedge being driven between the player/PC identity so carefully nurtured to that point. It turns out that (Spoiler - click to show)Black Blade and Rosa are one and the same, and though this doesn't seem to matter to the protagonist (whose undying love for her is proclaimed but not explained) Rosa, apparently unable to live with her recent past, is about to (Spoiler - click to show)commit suicide by dynamite.

This climax could certainly work as drama, but as presented here it's just too rushed to appreciate; the result is a bad stumble at the end of what is otherwise a high quality production. Better preparation of the player's mindset would be the key here: As always, the measure of quality for a twist is how much it makes sense in retrospect given what the audience has been shown, i.e. how well it reorganizes prior perceptions, especially how it grants new significance to details that didn't quite fit the audience's operating assumptions before the reveal. Perhaps the device of (Spoiler - click to show)using each bad guy's dying words as a clue (as seems to be started with the death of Ace at the end of the opening scene) could be leveraged to better set up the intended unveiling. (Even better would be to have released some prior episodes, so that the player would be able to genuinely sympathize with the PC's viewpoint, but let's not get too ambitious with our idle wishes here.)

Ending aside, this was a very enjoyable game to play -- it took me two or three hours to complete since I was unwilling to use hints or a walkthrough (though I did end up needing one clue). That's a testament to the author's skill in crafting a fun and engaging puzzle-solving experience, and I can see why Mockingbird placed so highly in 2021's IFComp. It's certainly among the top few of IF Westerns and Western-adjacent games I've seen (alongside The Legend of Horse Girl, Like a Sky Full of Locusts, and Hoosegow), though I've tried only a small portion of Westerns listed on IFDB. I'd recommend The Song of the Mockingbird for the puzzles, if not the story, and I'd certainly tune in for another episode of the adventures of Boots Taylor.

This game is about a child "inventor" who plans revenge against a trio of bullies that ruin a favorite project. Physically outmatched by the antagonists, the PC instead devises a series of traps to use on them during a planned after school showdown.

It's a terrific premise, but one difficult to deliver in interactive fiction: The ideal freeform combination of objects to novel purposes will always be more limited to the imagination of the author than that of the player. Author DCBSupafly handles this by simply having the game suggest the type of combination that will work when the key ingredient for a trap is discovered. This gives the gameplay a scavenger hunt feel that primarily consists of exploring the available environment (i.e. the protagonist's house) in order to locate the necessary materials.

The writing is of the "adequate" class; simplistic and skimpy on details, it suits the voice of the grade-school age player character well enough but offers little other than description to the reader. Even as description it falls short, as it is frequently the case that items apparently in plain sight (Spoiler - click to show)(e.g. a rotten watermelon sitting on the kitchen counter) go unmentioned in room descriptions, and must be discovered via dedicated >EXAMINE commands. It does do a reasonable job of capturing the burbling, excited outlook of the PC, and this gives the game an air of sincerity that is its greatest asset.

As with every ADRIFT game I've tried, I frequently found the parser to be irritatingly inconsistent. Be prepared for this when starting the game, I guess. Fortunately, by the time you get sick of wrangling with the parser's occasionally erratic responses, you'll probably have made it far enough to end the game.

As Lance Cirone notes, the endgame can be reached at any point after three traps have been deployed, but there appear to be many more possible -- I'm guessing at least eight. Neither the walkthrough nor the >SCORE response gives any clue about the upper limit; I managed five (Spoiler - click to show)(a squirt gun trap, a mud trap, a watermelon trap, a pit trap and a sticky note trap) and had materials that seemed suitable for at least three more (Spoiler - click to show)(two kinds of pepper, a spider egg sac, and rotten milk). I couldn't find a decompiler for ADRIFT that worked on it, so those secrets remain for other players to discover (and hopefully share).

This game could easily have been a high two, since it practically cries out for more polish and refinement. The gleeful climax, which plays out like an 80s kids' movie, is fun enough to push it over the edge into three stars, which means I recommend it as a play experience for those in the right mood. Would-be authors are encouraged to contemplate the premise and how it might be better executed... great artists steal, right?