Shades of Gray

Anyone interested in the history of interactive fiction will sooner or later come across references to this relatively famous piece from the "dark ages" of the genre, i.e. the period after the collapse of the commercial market and before the "renaissance" triggered by Graham Nelson's release of Inform 6 and the publication of the Inform Designer's Manual, 4th edition. In this period, the most prominent tools available to would-be authors were TADS 2, a C-like language of considerable power, and AGT, a less flexible and capable system designed to be easier to use for non-programmers.

Critically, the author of AGT sponsored contests (with at least the first prize paying money) for the best game written in the system, which surely served to spur the completion of many works and began the tradition later continued by the annual IFComp. Shades of Gray is among the works submitted to these AGT contests, and it won in the year that it was submitted. It was constructed by a group of seven disparate authors, one of whom was Judith Pintar, author of the well-regarded CosmoServe. Notably, the seven contributors cooperated exclusively through contact via the CompuServe platform, to which they all subscribed.

Based on the final result, it's not clear that there was much in the way of overarching design concept. As others have noted, the game's separately-developed segments vary in quality, but overall they are well-implemented by the standards of the time, and I must say that this was the highest level of command parsing quality that I have ever encountered in an AGT game. (AGT parsing is quite limited compared to TADS or Inform, based on word-for-word pattern matching instead of attempts to identify parts of speech. This creates a much higher burden on the author to ensure smooth interaction, and it also reduces the transferability of learning about what counts as proper interaction. For example, when trying to use a shovel -- of which there are a surprising number in this game -- the player will find that the correct syntax changes across different segments, reflecting each contributing author's own preferences.) In general, the quality of the interaction seems to go up as one progresses through the game, with its disjointed (and somewhat irritating) opening giving way to large portions of relatively smooth sailing.

What the work lacks is any sense of true coherence. While individual aspects can be picked out as high points for quality of implementation (e.g. the (Spoiler - click to show)tarot reading scene that is the structural backbone of the first half of the middle game) or writing (e.g. the various interactions with (Spoiler - click to show)spirits from voodoo mythology that are the backbone of the second half), the narrative is something of a mess -- layers of unmotivated and unedifying twists abruptly transform the story from gothic horror to lazy psychological drama to magical realism to Civil War survival story to medieval adventure tale to cheap political thriller. It's a ride that keeps the player guessing, which keeps up interest, but looking back from the end of it the question becomes: Why?

The title suggests that the theme is intended to be the difficulty of achieving strong moral clarity in the messy real world, but the gameplay does little to support this. The most direct treatment is in the climax scene, in which the protagonist must choose between (Spoiler - click to show)delivering some incriminating documents to either those incriminated by them and (Spoiler - click to show) delivering those documents to members of a law enforcement agency. This is... insufficient. As a clever person to whom I described the plot quipped: "Nothing says 'shades of gray' like a binary choice!" To the extent that this choice presents any kind of quandary to the player requiring thoughtful reflection, the game subsequently undermines itself by assigning one more point to (Spoiler - click to show)turning the evidence over to the CIA assassins threatened by it than (Spoiler - click to show)handing it over to the FBI, whose interest in it may be more about inter-bureaucratic infighting than bringing the conspirators to justice, which implicitly makes the former the "right" choice after all. (To be fair, the denouement section that describes the long-term effects of various events does not seem to put its thumb on the scales this way, and the various interludes of history supernaturally revealed to the protagonist present multiple perspectives... but in the long run that just makes the score's coded commentary less excusable.)

Other aspects of the game relate only weakly to the supposed theme. Robin Hood is a good guy fighting against abuse of power! No wait, he's a forest-dwelling thief and thug who must be punished for breaking the law! (I didn't bother to use spoiler tags for those because the two segments involved seem ultimately irrelevant to the main plot.) The protagonist shouldn't feel bad about (Spoiler - click to show)his father's death; he was just a kid, and it was an accident! (That's ultimately irrelevant, too.) It's probably OK that the protagonist (Spoiler - click to show)has a dalliance with a voodoo love goddess; it was a rare honor, and she'll (Spoiler - click to show)grant protection to him and his (alleged) true love forever after. I get the distinct impression that there were some last-minute adjustments made after the title was selected, in an attempt to better justify it.

Although there are frequent guess-the-verb and guess-the-syntax issues (as is typical for the era and the development system), these are offset by the very good integrated hint system, to which I found myself resorting frequently when my patience wore thin. Hints are graduated, so it's not necessary to completely spoil the puzzles in order to get help, but I recommend that the modern player make liberal use of them -- for the most part, the obstacles that I used them to bypass were not the type likely to be considered as rewarding to overcome unaided. I also strongly recommend that any player reaching the voodoo-themed jungle section reach for David Welbourn's excellent map of the area (available in the download links) -- this whole zone is a nasty and pointless old-school maze, and the game doesn't even have the good graces to provide sufficient objects to use as markers. On top of that, two rooms that are different enough from the others to not seem to need markers both have identical descriptions but are, in fact, different -- a design choice that comes across as pure spite. The hour that I spent trying to navigate the maze "properly" was completely wasted time. (The author of this section most definitely anticipated the difficulty being created; there are three tone-breaking "comic" cameos of other people wandering through that zone that are encountered if one spends enough time there.)

On the whole, I didn't find much to recommend about this piece. It does remain historically significant, and it clearly stands out from the pack when gauged against its contemporaries, but these qualities do more to justify its place as an exhibit in the museum of the history of interactive fiction than they do to earn it a place in the library of classic works worth playing today. One can point to it as an early example of collaboration-at-scale such as would later produce Cragne Manor or note surprising similarities between one of its segments and Adam Cadre's Shrapnel, but if one is not interested in deliberately evaluating it within its historical context, there is little reason to spend the time playing it.