Anchorhead

(Note: Would-be players are well-served by other reviews; this one is for would-be authors.)

There aren't really that many works of horror IF. Well-known works are fewer. Award-winning works pretty much come down to a handful, with Anchorhead being the first and only for at least a decade.

What makes it so hard to write horror IF? My usual argument is that it comes down to the problem of controlling pacing, which is critical to building the player's mood, and which is extraordinarily difficult to manage with the toolkit of interactive fiction. Control it too much, and the player is likely to feel "railroaded" and thus cheated of the promise of interaction. Control it too little, and the player will inevitably dawdle and poke about in the world you've built, which has the effect of constantly draining away the tension that you're trying so hard to keep on the rise. The player may enjoy bits and pieces of the experience but will not come away with the whole you envisioned.

Mr. Gentry seems to have very consciously grasped the challenge here and created a number of subtle innovations that go a long way towards overcoming both it and other obstacles to translating the methods of horror into IF. It is well worth examining these innovations in detail to try to understand what they solve, how they work and how they might be improved.

Anchorhead is patterned after the works of H. P. Lovecraft, which typically feature a protagonist who, beginning in a relatively humdrum setting, discovers previously-unsuspected horrors and subsequently struggles (often unsuccessfully) to retain his sanity as he grapples with the redefinition of his reality. In following this formula, it is first necessary to establish a starting point of normality, and Mr. Gentry clearly went to great lengths to do so. The "normal" presented in this work differs significantly from what is typically found in interactive fiction -- it's closer to actual reality in several ways.

First, as Emily Short notes, Gentry's prose offers players a multidimensional sensory experience that is far above-average in its quality, and which is delivered with amazing grace and economy. Not just sight, but sound, smell, touch are all intertwined throughout the room and object descriptions. The work that went into all of this writing was enormous, but with it Gentry achieves an important goal: As a player, you feel much more immersed in the environment than you would in most games.

Second, there are nuances of interaction that faithfully mimic the mechanics of reality in ways surprising to long-time players. Most notable here is the implementation of a model of the PC's hands -- the game keeps track of how she's holding her inventory and interacting with objects, causing failure of some actions when neither hand is free. While this level of realism has the potential to be a major annoyance, Gentry's coding skills ensure that, for the most part, you won't have to worry about it, as the PC will automatically shift things around on your behalf. The mimesis is somewhat broken here by the presence of a "holdall" object with unrealistically large carrying capacity, but since inventory limits are anathema to most players, this is an acceptable tradeoff. From time to time, the lack of free hands or pockets asserts itself in a realistic manner, once again reinforcing an underlying normality that brings you another step "into" the game world.

Third, again surprising, is the implementation of the weather. The game's storms are almost as annoying in Anchorhead as they would be in real life, prone to interfering with your inventory in ways which, though not hyper-realistic, manage to catch the essentials of the situation(Spoiler - click to show). That hurricane lamp you just walked outside with? It's out. That box of papers you had? Well, you still have the box. A well-implemented umbrella, working in conjunction with your hands, deals with most of the hassle, but Gentry has cleverly managed to make it just real enough that you have to worry about it as a player, elevating it above mere background description and again forcing you deeper into the PC's situation.

Fourth is the implementation of NPCs. I agree with Peter Pears that this is an exceptional example of the potential of the ask/tell system in the hands of a good writer, which makes talking to people feel like real interaction. The topic depth here is again evidence of hard work done with great skill; NPCs respond to topics that many players might not think to ask, if they haven't been paying attention to all of the minor details presented elsewhere in the game. This has a positive feedback effect for you as the player in that you are rewarded for making these connections in a way that does not affect the game's playability but once again draws you further "in". (Incidentally, this is a great variation of the "show, don't tell" technique for confirming the player's understanding of the situation, as such connections are rarely noted by the PC.)

Last but not least, the handling of the PC strikes an excellent balance, leaving enough AFGNCAAP-like interaction to allow anyone to project themselves into the lead role while retaining a narrative voice that colors the whole experience in a meaningful way. From time to time, the PC's mentality injects itself unobtrusively into the game, always in a way that reinforces immersion and enhances the player/PC connection(Spoiler - click to show). I am especially fond of the PC's unwillingness to go to sleep with the doors unlocked the first night in the house. Though it means having to get back up, put your clothes on, go downstairs and deal with it, it also makes sense that the PC would be too agitated about the situation to go to sleep without doing so, and I love how it's presented as though you simply forgot to do this -- even though wandering around leaving doors open is perfectly normal behavior in most IF. Again, this is a very restrained and subtle reinforcement of the game world as "real" that is amazingly precise in that it doesn't quite annoy you as a player.

These efforts to enhance reality don't really affect the gameplay very much, but they do affect your experience as a reader. After investing a lot of work to align the player's perceptions and mindset into an expectation of realism, Gentry is able to start introducing the surrealism that is the backbone of Lovecraftian horror. Gentry's success in this effort springs from the insight that underlies the Lovecraft quote which opens the game: "The oldest and strongest emotion of mankind is fear, and the oldest and strongest kind of fear is fear of the unknown."

Mr. Gentry's first key perception was understanding that the right place to develop tension is in the mind of the player, not the mind of the PC. Despite the trials of the experience portrayed, the PC has almost no observable emotional reaction -- if there is an emotional reaction, it comes from you, and it's achieved because the player/PC identity alignment has been so carefully managed. As the situation becomes more desperate, the PC becomes willing to do things that either explicitly or implicitly would have been balked at normally(Spoiler - click to show). Examples: stealing her husband's faculty card, spying on her husband, "hacking" his computer, stealing the mechanic's key, crawling through sewer pipes. Since many of these actions are necessary to advance the plot, in effect, the way the PC's reactions are modified to suit the mood that has been targeted almost acts as an emotional puzzle structure that ensures you feel the way Gentry wanted you to at each point(Spoiler - click to show). I say "almost" because not all of these actions are necessary to "win" (though they are to achieve maximum points).

Gentry's second vital intuition was in understanding that the way to keep the tension from dissipating is, unintuitively, to build it very slowly. Since no number of exclamation points is sufficient to induce a surprise reaction in the player, Gentry instead uses the technique of scattering numerous small clues to the central mystery throughout the game world. As Peter Pears phrased it you build your understanding "piece by piece" from these brilliantly interlocking clues in a way that makes your uncomfortable comprehension seem to well up from the dark recesses of your own subconscious instead of being handed down from above(Spoiler - click to show). I particularly like how this technique interacts with some of the "red herring" ideas introduced during the library research portion. As a player, you're not sure which to expect to materialize in-game. Notably, there are multiple clues for key information, making these realizations easier to achieve for the player and reinforcing the realism style. Even more notable is Gentry's craft in writing some of them. The "visual" clues (Spoiler - click to show)(i.e. the paintings in the gallery) are so well-written that I can recall them to my memory as though I had seen an actual image.

Overlaid onto the plot is a well-formed "scene" structure that divides the game world both chronologically and geographically. While the division of time into day and evening cycles is a bit too crude to be completely believable(Spoiler - click to show)(see Brian Uri's Augmented Fourth for a similar but more granular and thus more effective treatment), large portions of the game world are only accessible during certain times, giving a very dynamic feel to the story compared to games that depend solely on spatial barriers to enforce the plot structure.

In addition, there are a few timed or "action" sequences sprinkled throughout the game to add variety to the pacing. With respect to these, I found very effective Gentry's technique of giving the player the opportunity to explore certain spaces in advance of action sequences that would take place in them. The first time you are in an area, your exploration (unrestricted by time) advances your understanding of the plot. The second time there, the application of timing restrictions seems perfectly fair, as you've had a chance to develop the knowledge needed to "survive" them and your attention is not diverted by the need to explore the environment(Spoiler - click to show). My personal favorite example is the slaughterhouse scene, in which the two modes occur back-to-back in the same area. It is a vividly cinematic sequence, though it is marred by the rather ludicrous (if effective) presence of the crayon drawing and inconsistent use of the verb "hide".

As a last note of praise, I admired the way that the author found a couple of interesting ways to discomfort long-time players via subtle manipulation of expectations(Spoiler - click to show). Example: The fly in the real estate agent's office is a persistent presence in the prose, but can't be interacted with as an object. It's irritating and disquieting since generally for IF prominence in the text equates to prominence in the object structure. Example: The inability to explore the house due to darkness on the first night. A touch of pseudo-realism that doesn't quite fit in the typical IF experience -- having gained entry to the house you, as a player, expect to get to check it out. I think it is small details such as this that left me not quite knowing what to expect from the rest of the story while still feeling grounded within it. This slight disorientation is the mark of encountering something new (which is very, very rare for long-time players), and that, more than anything else, is what makes this work stand out in my mind.

All of the above is not to say that Anchorhead is perfect. I actually felt that the introduction (pre-arrival at the house) was quite poorly done. I had tried this game before and put it aside after 50 moves a couple of times, but this time I gritted my teeth and powered through it -- and I'm very glad I did. In addition, there are quite a few small bugs and places where the polish wears off towards the end of the game(Spoiler - click to show). For the nitpickers interested in a tour of these inconsistencies in the otherwise very high implementation quality:

* There seems to be an unintentional "last lousy point" issue due to a sensitivity to the order-of-events between researching birth and death dates and reading about the Verlach family in the library book. If you read the dates first, you make a connection and gain a point when you read the book, but not the other way around.

* Messages about flute resonance can sometimes call both columns the "right-hand column" in the mound.

* The madman in the asylum mimicking your voice doesn't seem to work correctly. I got garbled text that I am fairly sure should have been repeating back what I had typed.

* The way the magic word "ialdabaoloth" is handled is problematic; quotes don't work and the failure of commands like "say ialdabaoloth" and "door, ialdabaoloth" make it an unintentional guess-the-syntax puzzle.

* Examining the lighthouse after it is destroyed shows it still "there" from multiple vantage points.

* Trying to push William off the bridge gives a default politeness-based refusal that definitely does not fit with the situation.

* The bum's corpse still seems to be treated as animate after his death; you get default NPC responses for many interactions.

* Michael's corpose seems to be absent as an object.

* The luggage default message stays the same no matter how crazy the situation gets. So does taking a bath.

* Automatic key logic doesn't take into account keys not on the keychain -- very noticeable in the madman chase scene.

* There are a few disambiguation issues in conversation topics, e.g. "the book" or "the professor".

Beyond these, there are some places where design choices seem antiquated today even though they are closer to the norm for 1998:

* gratuitous mazes, though small and at least one can be bypassed

* darkness in the hallway during the madman scene; this turned into an annoyance for me and screwed up the pacing of the scene because I didn't have a light source, though this doesn't seem like an intentional "puzzle"

* the torn square of canvas being semi-hidden though it would clearly have been visible to the PC is strange and requires a careful search in a sequence otherwise oriented around a fast escape

* the climactic puzzle with the mirrors has many problematic details (Spoiler - click to show)(Why can you only mess up a replacement? Why doesn't Michael/Verlach notice the label on the replacement mirror? Why can't you "touch mirror" with an oily finger to get the same sabotage effect?) and definitely took a walkthrough for me

. Most likely, this is due to the scale of the work being so large that a) Gentry's skills in writing and coding improved over the course of its development and b) playtesting to perfection would take more hours than were available from volunteers. Space constraints may also have been a factor -- this work was developed pre-Glulx and must have stretched the limits of the z8 format.

Perhaps the greatest criticism I can muster is that Anchorhead very nearly succumbs to the pacing problem that kills so many attempts at IF horror. This is most obvious during Day Three, where I wanted STORY, not puzzles, and my patience for them was wearing thin enough to start consulting the walkthrough.

My natural rating for this work would have been 4 stars, or "exceptional" by my scale. I'm compelled to give it 5, however, because, in my experience, it is the king of the genre, far surpassing its Infocom-produced cousin, The Lurking Horror.