I sometimes have trouble motivating myself to finish my games, and I am least likely to play a game through to completion when, by the end of the first few levels, I feel like I've already seen all that the game has to offer. To judge a game in this manner is not exactly fair; for who can say for certain that the fifteenth level of Metal Storm Five: Silent Thunder: The Darkening will not turn the entire game on its head via some unforeseen, potpourri excursion to a realm where things are neither metallic, stormy, nor dark?
It has happened before, to be sure. Remember how surprised you were by the lone-commando mission in the original Command & Conquer? (I do not admit of the possibility that you never played it.) That mission was the most brilliant thing about C&C. Not only was it a blast to play, but it also represented a completely different approach to gameplay than every other mission in the game: it placed the player in control of a single, vulnerable soldier, as opposed to the hordes of expendable troops that, even at that early point in the real-time strategy genre's history, had already become entrenched as the standard and expected play mechanic.
It is hard to overestimate the influence that this single C&C level has wrought upon the course of the gaming industry. I propose that it is a more proper progenitor of modern stealth-style gameplay than the later Thief: The Dark Project, which many regard as the seminal game of this type (though by no means do I deny its import). The C&C commando level challenged the player in a way that no prior game of which I am aware had ever done. Rather than tasking the player with adapting to a new enemy, different level design, varying mission objectives, or the like, it instead required that the player abandon all precedent and, in effect, learn to play an entirely different kind of game altogether. But unlike the common and generally intrusive "minigame", the C&C commando mission did not constitute a hiatus of any length from the normal interface or controls. The mission therefore feels contradictory, in a way; innovative in terms of theme, but consistent with respect to player input and control.
And therein lies the goal, for the game designer: to get the player accustomed to a certain way of playing, to establish a known system of play, and then to blow the player's frickin' mind -- not by replacing the old system with a new one, but by exposing previously hidden layers within the established system. We may liken this method of design to the setting of seemingly immutable rules, which the player subsequently discovers allow ample room to wiggle. The difficulty from the designer's perspective is at least twofold: first, in designing a basic game type that is interesting enough to have multiple layers at all; and second, in unraveling the onion skillfully enough that the topsy-turvy period of transition and re-indoctrination remains fun for the player.
Some authors of fiction adhere to a similar method. One of my favorite SF authors, Larry Niven, excels at setting forth rules for his fictional universes, and then breaking them. He does not violate the rules outright, but instead uncovers unexpected loopholes by which they can be circumvented. For example, several of his Known Space stories revolve around certain spacecraft that are indestructible by design. Niven uses his familiarity with physics and astronomy to expose various loopholes in the rule, and, by subjecting the indestructible ships to exceedingly outlandish (and, when the stories were written, still broadly theoretical) physical conditions, he destroys what had at first seemed invincible.
Niven learned this technique from Isaac Asimov, who compiled the famous Three Laws of Robotics:
1. A robot may not harm a human being, or, through inaction, allow a human being to come to harm.
2. A robot must obey the orders given to it by human beings, except where such orders would conflict with the First Law.
3. A robot must protect its own existence, as long as such protection does not conflict with the First or Second Law.
It turns out that Asimov's robots aren't always quite as benign as the Laws might indicate; yet, crucially, Asimov doesn't ever violate the laws outright. The fun of reading Asimov's robot mysteries is in trying to anticipate where the wiggle room will appear.
If there is any one game which we should recognize as exemplifying this same method, then it can only be 1995's Chronomaster. Written by SF authors Roger Zelazny and Jane Lindskold, Chronomaster is a superb point-and-click adventure game in which the player takes on the role of Rene Korda, a retired creator of "pocket universes" -- small universes designed by contract artists according to the specifications of various wealthy patrons. (Pocket universes are not necessarily governed by the same physical laws as our own, and this fact by itself allows for much in the way of creative design.) Korda is called out of retirement because somebody is deactivating certain pocket universes, which has the effect of freezing the progress of time within them. As he investigates these eerie, frozen realms, Korda carries with him "bottled time," which extends a bubble of temporal activity around his body. So, if Korda stands near a landslide frozen in progress, it will become unstuck, and may pose a hazard. If he approaches a bird on the ground, it will come to life and fly away, only to become frozen again in mid-air when it departs Korda's sphere of influence. Under these strange circumstances, all objects -- animal, vegetable, and mineral alike -- behave differently, and the player must constantly assess how the new rules will affect familiar, everyday scenarios. More than any other game, Chronomaster revels in the playful juxtaposition of the unfamiliar and the familiar.
The pertinent difference between Command & Conquer and Chronomaster is that whereas the former offers the player no early indication of the major change in gameplay style that accompanies the commando mission, Chronomaster's setting and story make it quite clear from the outset that the player should expect many situations in which past experience will not suffice as guide, and in which prior "knowledge" will be called into question.
Kudos to you if, when reading the previous sentence, you detected an intersection between games of this kind and one of the more basic problems of epistemology: that of the surety of knowledge. The notion that knowledge is ultimately fallible, that prior experience is unreliable, and that skepticism is irrefutable, remains the most utterly intoxicating thought of which I am capable. By exposing us to situations in which everything we thought we knew is shown no longer to pertain, games that freely exchange what is known for what is not serve as an analogue to this frightening epistemological problem. And, in brief but all too terrifying moments, I fear that they may presage some further revelation yet to come.