Recently, over on ifMUD, a bunch of us had a discussion about what makes games replayable, and why. We tried to consider actual games we had replayed, to keep us on track. You can read the entire transcript (slightly edited for content) at:
http://spod-central.org/~lpsmith/IF/replayability.html
Here's my summary of what we came up with, along with some more of my thoughts:
First, there are two types of replaying that can happen with a game. The first type is when you replay sections of the game during your first game 'experience', defined as the amount of time stretching from when you first start playing the game to the point at which you consider yourself to be 'done' with it. This can involve any amount of the game, from one small scene to start to finish. To make a cross-media analogy, this would be like rewinding a tape to re-watch a particular scene in slow motion, or even finishing a book (as I did once) and immediately starting over from the beginning, because I knew there was so much I had missed the first time. For ease of reference, I shall call this type of replaying, "Iterative."
The second type of replaying happens when you come back to a game after you're already 'done' with it. Again, you may play the whole game, or you may play just scenes (though the latter is harder if you haven't kept your saved game files). Again, for reference, I shall call this type of replaying, "Re-visiting."
It is interesting to note that while the former type is much more common in games than this latter, the reverse is true of more static media such as films and books. More on this later.
Both Iterative and Re-visiting replaying can happen for a variety of reasons. Here are some categories we came up with, along with the definitions we gave them. Some of them overlap, naturally, but each emphasises a different aspect:
Replaying for Mastery: The player's motivation in this type of replaying is to assert their dominance over the game. This desire can take many forms: to know every last nook and cranny, to have read every last interesting response, to complete the game in the least number of moves, to figure out a way to solve puzzle X *before* solving puzzles Y and Z, to get the 'last lousy point' of the game, or even to find and exploit every potential bug the programmer has left open. It was pointed out that people are particularly taken to patterns, and that they will replay a game for mastery sometimes to master a pattern--sometimes even if there wasn't one!
Replaying for Completion: The player's motivation here is to see the entire world the author has provided. In our discussion, we made a distinction between 'Thoroughness' and 'Completion' but I'm no longer convinced that there's a useful distinction there not already covered by the Completion/Mastery distinction, so I'm throwing them back together here. Whether you could have seen something and just didn't the first time through, or whether some of your decisions prevented you from seeing certain other sections, then, we'll call that 'Completion.' Replaying 'Grip' or 'Tapestry' just to see the parallel paths, then, would fall under this category, as would replaying 'She's Got a Thing for a Spring' just to see what Bob did all day.
Note that some of the same things can be accomplished in replaying for both Mastery and Completion--the difference is in the player's motivation to do so. One would harldly claim to have 'mastered' Tapestry before playing through all the paths, but aside from that draw, you could experience in Tapestry a sense of not being 'done' until you had played through all the possibilities.
Replaying for Impact: The player motivation here is to see how they can manipulate and color the world the author has provided through their actions. This is often attempted through dialogue: Whizzard mentioned replaying the dialogue in Monkey Island to see what different changes he could effect (until he discovered that the answer generally was: not much). Adam mentioned how in his work-in-progress (Pantheon) and, to a lesser extent, I-0, the way conversations progressed early in the game could make a difference in the way things turned out later on. I personally remember how, in 'Spider & Web,' the feel of the game was greatly modified by whether I tended to answer glibly, lie a lot, or refuse to answer the questions posed. The impact here must be noticable, but it may be slight--the main story can progress along the same major plot, just be colored differently.
Replaying for Experience: This is the most different from all of the above options. While the motivation in the rest is for the player to do something different this time through, the motivation when replaying for experience is just to see the same stuff again, because you liked it. The clearest example of this we came up with was re-playing an old game for nostalgic purposes. I've re-played the old Zorks several times, just because they were my first games. For Whizzard it was Wishbringer. For Jarb it was Dungeon on a LA32. Other examples are more rare, but do exist. In a later conversation (not in the transcript, unfortunately) Ventura mentioned that he replayed Mercy several times over the course of a month, for no specific reason he could pinpoint--it was just an experience he needed to relive. I recall a newsgroup post claiming that they played AMFV about twice a year, just so they wouldn't forget its message.
Interestingly, when we compared replayability of games with rereadability of books, the same aspects emerged--but in a different ratio. Books, in general, have a very high 'experience' factor, and a much lower amount of the other three. Also interesting was the fact that 'Re-visiting' replaying usually happened for experiential reasons-- and again, that's the kind of reaction people often have towards books. Part of that experience is encoded in the stories, characters, and ideas therein--all aspects that are either hard to do in IF, traditionally neglected, or both.
The other factors, while often catering to the same desires (Mastery and Completion, mostly), are often met in very different ways. This is usually found in aspects of the book that you didn't 'catch' the first time through: unnoticed foreshadowing, cross-references, allusions, and even missed jokes. It may be a particularly effective technique to fold some of these aspects of traditional literature into standard IF motifs that cater to the Completion desire: cross-referencing one path in its parallel path, for example, or foreshadowing future events in an out-of-the way location that the player isn't required to visit to finish.
Of course, in any static media, you can never Impact the story yourself. This is the greatest strength (as well as the greatest challenge and pitfall) of interactive media. I find it very interesting, though, that people claimed to not need *much* impact on the story--just *some*. Enough to flavor it to taste. And even with all our puzzles, I don't think we've begun to scratch the surface of our potential here, and it bears some further investigation.
In general, I believe it will be profitable if we, as authors, look to what brings people back to games as well as to more traditional literature. I hope the classifications above might help people organize their thoughts about some of the issues involved. We might even have an official MUD forum on the subject in the future, possibly some time after the competition. Until then, as we say, "Wave!"
-Lucian