Foiling the Information Age, Part Two

by Shannon Appelcline

July 3, 2003 - Last week, in Trials, Triumphs & Trivialities #124, Foiling the Information Age, Part One I posed a question. How do you prevent players from using both in-band and out-of-band communication to reveal the secrets of your online game?

I've got a number of possible answers for this problem, largely arrayed around the three central concepts in this conundrum: information design, communication design, and information value. This week I'm going to look at some of the cruder ways to change your information and communication design; be warned, they're blunt and could cause trauma.

The Design of Secret Information

To start off with, you should question the core design of secret information, which tends to involve static puzzles. These quests or puzzles with repetitive methods that can be used to achieve success were originally conceived of during the single-player era of adventure games, and do not have to be the basis of multiplayer games. Presuming that you still do want to have puzzles and quests as an important part of the design (and, as we'll discuss next week, you might not want to), there are a number of ways that you can move them away from the static, single-player design.

The simple and most obvious answer is randomization. If you constantly change the connections between your rooms, the items needed to complete your puzzle, the passphrases, or whatever else, there will never be any problem will players revealing to each other the secrets of your game.

One of the most crucial problems with this approach, however, is timing. When, exactly, do you choose to randomize your secret information? After all, you don't want to do it while someone is in the middle of solving your mystery or there will be serious frustration, gnashing of teeth, and eventually a player exodus. There are, however, a number of ways that the randomization can be organized more rationally (and, as we'll see, some of these are close kin to the issue of when you reset games, as discussed way back in Trials, Triumphs & Trivialities #28, The Dynamic Dilemma, Part Two):

1. Timed Randomization. Rerandomize your puzzles at a set time. As long as you announce this to your players, they'll be well aware that they're out of luck if they don't solve the puzzle/maze/quest/etc. by this time. It's perfectly possible to integrate this randomization into some in-game event (e.g., "Every 7 days the Waves of Chaos wash over the shores of the Elven Forest"). And note that this randomization doesn't have to be rapid. Every week or even every month might be sufficient to keep your secrets fairly secret.
2. Completion Randomization. Rerandomize your puzzles every time they're completed. This way, attempting to solve the puzzle turns into an inter-player competition and having to start over is just a consequence of not being the "winner".
3. Empty Randomization. If your game doesn't constantly have people on in every zone you can randomize when the local area is empty. Players will quickly learn not to wander off until they've either completed a puzzle or are done
4. Player Randomization. Perhaps both the most and least elegant solution. You make a puzzle different for every player. This means that you never have to rerandomize the game area, but also means that players can't really properly communicate with each other about what they're seeing (unless you also do some work on communication design, as discussed below). This is actually the possibility that we're considering for dreamquests in Lovecraft Country, and I think it works great for that particular setting, but might work less well for a more realistic genre.

The idea of randomizing secrets also has one other disadvantage: it could hurt the "reality" of your gameworld. As I said you might be able to explain some of it via in-game reasons, but eventually those excuses will grow thin, and your players will be at least somewhat knocked out of the gameworld by its lack of reality in this aspect.

Is the tradeoff of being able to truly hide information versus having a less realistic gameworld worthwhile? That's just one of the questions you'll have to answer when you're creating the game.

Rather than changing puzzles randomly you can instead make puzzles dynamic by letting players change puzzles. Imagine that when a player successfully completed a quest he was given a new one: he must now set up the puzzle components to make it hard for the next player to solve the same puzzle. You could do this within the constraints of a game world much more easily ("'You have proven yourself worthy by retrieving the Sword of Lightning. Now, you must once again put it back into the Earth lest unholy forces discover it. I shall use my godly powers to break it into three pieces.' ZAP! 'Now hide those pieces where only the holy might reunite them.'")

(And I'll note parenthetically that this introduces the whole idea of player creativity into the game, something that I think is worthwhile in and of itself, as I've discussed previously in Trials, Triumphs & Trivialities #29, The Dynamic Dilemma, Part Three and Trials, Triumphs & Trivialities #67, Creativity & The Online Gamer.)

This of course brings up a whole bunch of other problems. You need to figure out how to incentivize a player to actually reset the puzzle; what to do if puzzles are constantly too easy; and what to do if a puzzle is too "hard" and not solved for a long time. I think they're largely solveable, but they need to be considerations if you choose to do this type of design.

Randomization or player arbitration: either one can do a good job of occasionally resetting your secret information, and thus keeping it truly secret. However, before you undertake these possibilities, you need to think about whether you actually want your puzzles to be dynamic, because in using this method you're inevitably going to create less meaningful puzzles.

And thanks to FogLeg for some thoughts on this topic over in the forums. In particular the thoughts about player arbitration originated from him.

The Design of Communication

Now, let me move on to the more controversial topic: communication. In other words, how can we as game designers make it harder to communicate secrets?

The simplest answer is that you can try to control information dispersal. This is clearly possible in-game. Say you had, as a piece of information in your game, a secret phrase which let you get pass the guards to get into the castle. You could modify your text output engine so that it randomized the word every time someone tried to say it, thus making it impossible to transfer the secret by in-band means. If you wanted to be real clever you could create a two-way function. If player A says the pass phrase, player B hears alternative-phrase #32 (which doesn't work for him). If player B says alternative-phrase #32, player A hears the pass phrase. By this type of tomfoolery you might even keep players from figuring out what you're doing... for a little bit. (And see Trials, Triumphs & Trivialities #34, The Power of the Medium: Individualized Output for some other thoughts on how this all works.)

This clearly only goes so far, however, because players will also (perhaps primarily) be using out-of-band communication to transmit secret information, and you have no control over that. If you really wanted to try and push this element of game design you could try and co-opt out-of-band communication by building "ooc" or "channels" or "chat" or whatever straight into your game, and then you theoretically have the ability to censor some out-of-band talk too.

Frankly, I don't think any of these will really work and I also believe that some of these ideas will introduce various ethical challenges about manipulating the communication of your players. I've discussed them mainly to introduce the idea of looking at communication as a solution for the secret information problem, because it's clearly one of the topics to consider.

However, I think that by flipping the whole idea on its head you do have a somewhat interesting solution. Consider, instead, that you've adopted the randomization of data solution I suggested where each player sees the puzzle differently, even when they're looking at the hidden information at the exact same time. As I mentioned this brings up the issue of how players communicate.

Now, imagine that your in-band communication automatically translates discussion of puzzle components. If a "red violin" is in important object in my view of the world and a "silver fish" is in yours, when I talk about the red violin, you'll hear me mentioning the silver fish. In-band revelation of information is still possible, but that might be OK, because it's part of the world. However, if I then try and talk in any forum that isn't moderated by the game, my discussion becomes meaningless.

I don't know if this would work in a full-out roleplaying game, with free prose available for communication. However in a limited strategy or puzzle game I think it's an excellent solution.

And I need to thank Quigg for his thoughts on this topic, also in the forums.

Conclusions

This week I covered what are, I think, the least elegant methods for keeping your secret information secret in an online world. By manhandling your secrets themselves or bastardizing the communication paths used by your players you can clearly keep some stuff hidden. Is it worth the cost? I dunno.

Next week I want to try and look at some classier ways to solve the problem, and suggest that perhaps we might be looking at the whole issue in the wrong way.

Recent Discussions on Trials, Triumphs & Trivialities: