There’s presently a bit of an argument going on about a LARP called Empire. Okay that’s a white lie, there are always lots of arguments, but let’s focus on this one: During an event generals submit plans to move their armies, between events these moves are resolved and during the next event they hear about how many of their people died and so on. In theory this adds a lot of roleplaying opportunities to the game, in arguing about strategy and excommunicating/executing/exonerating generals who make controversial decisions. In practice it’s pushing some of the generals are moving towards heavy mathematical analysis to determine the strictly optimal moves, while others would rather have a general understanding and enjoy other aspects of the game.

The debate between players who like the crunchy aspects of game and enjoy deep mathematical analysis and players who are drawing on other elements and find that getting bogged down in calculations ruins that has been going on forever and generally board game designers have arrived at a simple solution: Different games for different folks. I don’t think that there is a right or wrong way for Empire to handle its military rules, its a big game that caters to lots of different tastes, ideally the military system would handle things one way that’s enjoyable to a certain types of player and those players will gravitate towards that part of the game while others preferentially move into different arenas.

The thing that surprised me was to hear the argument “The people running the game don’t have a choice, whatever system they use will eventually be reverse engineered by the players, so they have to make the details of the system public which in turn promotes the number crunchy version of the game regardless of the designers wishes.” Apparently a lot of folks involved in writing the game have taken this line, though I’ve not seen it directly – but as a game designer I’m not sure it’s true.

There are plenty of ways that players could run a series of inputs (their actions) and outputs (military reports) through a system and not be able to derive the underlying system. On the most basic level an element of randomness would require a lot of tests to reach statistical significance and as there are only four games a year that alone might be enough to mask a system for the lifetime of the game – but lots of people don’t like randomness (for various good reasons) so what else can we do?

An extra set of hidden inputs seems like the obvious move. It’s hard to reverse engineer a system when you don’t know if your data is complete, but you could still get the intuitive feeling that “big armies are better” and “those guys are super dangerous”. For instance each enemy army could have a rating associated with the quality of its commander that affects the formula somehow. The players can never access this number directly though could get hints (perhaps the rating is 0-100 and scouting/divination might reveal a “poor” or “excellent” comamnder). That’d open up new areas in the game, rituals to affect commanders, skirmishes to take them out, genuine hostage negotiations in which handing the enemy hostage back has a long term meaningful impact (which it rarely does at present).

Generally making any information vaguer makes it hard or impossible to determine an underlying system. Simply making all reports (victory points, casualties taken and inflicted) accurate to within 5% would go a long way to making it hard to crack the underlying system – it also discourages gaming the system over making in character decisions. No real general goes “Well in this campaign we’ll lose exactly enough men that the army doesn’t quite break so charge!”, knowing that the figures for your number of remaining troops might be out by a few percent encourages making realistic decisions over those based on mathematical thresholds even if the system is cracked.

There are plenty of other options, many of which could add something to other areas of the game as well as making the military system slightly opaque , but let’s move away from Empire and apply this line of thinking to board game design. In a referee’d LARP it’s relatively easy to hide parts of the system from the players, how could it be possible in a board game in which the players need to read, understand and enforce the rules?

It almost goes without saying that Risk Legacy’s hidden rule cards that are slowly stickered over the book changing the rules with repeated plays achieve this. I’ve gushed about how awesome this approach is before though and am very much looking forwards to the upcoming games using similar systems. For now, let’s pretend that we’re trying to design a game that can survive a infinite number of repeated plays.

Mathematically I’m not certain that it’s technically possible without some sort of computer assistance, because while one way cyptosystems are a thing the necessity of involving the players on both sides of any information fence means that they wind up understanding the whole thing. However I can imagine a game in which the players understand how to generate the rules and how to interpret them, without knowing exactly what they are.

Consider a game in which players move armies around a map and fight each other. Each unit type has an associated bag, when two units fight one cube is drawn from each bag and the colours of these cubes determine which unit wins, the cubes are then returned to their original bags. Rules could exist allowing the players to seed the bags at the start of the game (Stick ten random cubes in without looking) and to let them understand the outcome (A yellow cube beats a red cube) but the actual rules they’d generated for the game would be hidden. By looking in the bags you could determine things like “Infantry vs infantry battles are mutual annihilation 80% of the time” or “Infantry lose to Cavalry more often than not” but the players don’t know that – they’d have to engage carefully with scouts and try to determine through experience which actions will work out in this game.

It’s not perfect because it creates a solvable second order problem. You could determine “Given what we’ve seen so far the optimal assumption is…” but the level of number crunching required could brought out of the reach of mental computation depending on the complexity of the interacting rules. That gives a possible route to implement hidden rules in a board game and I’ve already mentioned another!

Given the increasing ubiquity of these things, “Not possible without a computer” is no longer synonymous with “Not possible.” It’s still a risky move from a commercial point of view, since you limit your market, but from a design point of view there are ways to enhance games with digital devices that we’re only just beginning to tap. It’s a really exciting field and sooner or later (I predict sooner) we’re going to see the first truly great game that fully takes advantage of this technology.

If some elements of the game are accessed by inputting numbers into a digital device that runs some program to compute outcomes you could include all kinda of neat tricks – hell it’s probably within scope to apply weather modifiers based on what the weather is like wherever the game is being played. There’s seriously wasted potential in how digital board and card games are currently being handled (I touched on this in the randomisation timing post a few weeks back). Hiding elements of system from the players is just one example, but it’d work well. There are a hundred ways to fudge inputs and outputs to make sure players never become explicitly aware of the system, while still being able to master it through experience and intuition. That’s how people get good at a lot of things in real life after all.

Now I’d end on a note of caution: Just because we can do something doesn’t mean we should. A realistic simulation often isn’t more fun than a stylised one, besides if games were 100% like reality there’d be no point in them. The decision obfuscate parts of a system from the players should be a deliberate one taken with some particular goal in mind about how the game should play or feel to play. Some of these ideas are neat and I’m looking forward to playing with them, but it’s got the potential just to be frustrating or to promote lazy design if not handled with care.

Happy designing 🙂