Some Thoughts on Design

Greetings, True Believers!

Design is very much like being a sculptor; a designer's job is to remove all elements which are unnecessary. Cool ideas like rolling dice against charts, or having hard counters based on colours (or whatever), and so forth are all older than Methuselah. The trick is to home in on your vision and then just remove as much cruft until you have the core gameplay.

Painting isn't hard. Anyone with access to PBS will have no trouble watching enough Bob Ross re-runs to become technically competent. The trick, according to an artist friend of mine who is quite good at his craft, is not knowing how to paint: it is knowing when to stop painting.

Consider Street Fighter or Samurai Showdown. These are indeed the architypical examples - the very blueprints - of the fighting game genre. Gameplay is fluid, ergonomic, and accessible. Differing styles of personal choice are facilitated via multiple avatars, and yet all avatars comparable to such a degree that in any 1-on-1 match no avatar is at any perceptible disadvantage. There are simply no extraneous elements, and there are no elements lacking. All functionality required to whup some ass is present, and any additional functionality would actually hurt the product. Another example of elegant design might be western Chess.

As an exercise let's look at some sort of combat game. We could go in the ever popular tradition of Dune 2 (successor to Battletech 2, and precursor Command & Conquer) where we buy units and buildings with finite mutually accessible resources in an attempt to defeat the OpFor. To mix things up there are restrictions such as certain units may only be purchased after specific buildings have been purchased. This is not the only approach. I recently kibitzed a game of Advanced War 2 and was happily surprised by the elegance of the design.

There are no extraneous units (often the unit list of a game is simply an insulting exercise in hyperbole) because each unit serves a specific purpose and all units are necessary. Gameplay is simple, thus liberating the belligerents to focus on their battle plans rather than forcing them to navigate esoteric icons and continuously peruse arcane manuals for unit capabilities. Advanced War 2 is turn-based. This refreshing blast from the past not only permits to one a moment of contemplation, but goes so far as to require it. One cannot modify the play area in any way; there are no accessible means of destroying cities or planting down new factory complexes. The resources are infinite but bound and are designed around the concept of the constant- or zero-sum game; capturing an opponent's city deprives him of revenue while enriching your coffers at the same time.

These RTT games are most usually misnamed as an RTS - misnamed because according to Carl von Clausewitz (father of the modern field of Military Science) the play involved is almost entirely within the Tactics side of warfare. So let us, for sake of argument, ponder the design of our own wargame, and let us assert we are not interested in making a clone of Dune 2 or Advanced War 2. Where to begin?

Warfare is a political activity. From December of 1941 through September of 1945 it was the foreign policy of the United States of America to be at war with the Empire of Japan. The every day sense of diplomacy, talking about aims and grievances, would not manifest the goals of the United States and the Empire of Japan made no attempt to do so (indeed, relations were already strained by our repeated protests against their Manchurian campaign). So perhaps one might brush off their Kissinger books and perhaps let the players command nations instead of battalions. We see that countries typically go to war only after some pretext has been established (even Hitler made excuses to justify many of his conquests), and we need look no further than Chris Crawford's excellent Balance of Power. The object of that game was to avoid a nuclear showdown with the USSR (or a nuclear showdown with the USA if playing as the USSR) because such an event ends the game (an interesting exercise in a negative Victory Condition!) and everyone loses because the Earth has been annihilated. If MAD is proscribed (or at least not made automatic), then one might permit conventional war in addition to cold war; escalating the DefCon through diplomatic crises would send the army rolling along to decide with bullets what the diplomats could not conclude with words! If MAD is desired, perhaps after a certain intensity of conflict could unlock the nuclear options (thus leaving the original Victory Condition intact, but permitting greater freedom with one's foreign policy).

But wait! We could drill down a level here. Dönitz was most preoccupied with destroying as much Alliance shipping as possible for the obvious reasons: weapons never constructed are never deployed. This then brings up an idea for industries. Since the Axis powers were making preparations long before they struck, and the United States only lately entered into a wartime economy, one might even attempt to simulate the salient aspects of modern economic thought. After all, a nation cannot indefinitely maintain, to the satisfaction of her populace, a wartime economy during peacetime, and peacetime economies behave very differently than wartime economies (even among socialist economies), one could add considerable depth. Perhaps one could list out a score or two of commodities, then create production functions to transmute some into others (input-output production functions), and perhaps even more to include the notion of intermediate commodities (consumers don't directly buy ball bearings, but they buy plenty when they purchase an automobile engine)…

Perhaps now we see our enthusiasm getting the better of us. This is not so much a game, as it is performing some min-max equations in a spreadsheet. The idea of my economy being hijacked because you annexed the major sources of rubber and gum might sound exciting, and just as I might be excited at the prospect of seeking to pressure you into abandoning the junta you recently installed in Nicaragua by wrecking your economy via currency dumping, doesn't mean these things are fun. Torpedoing elements of a convoy might be fun, but look at what we have here. We ought to assume, for example, that as leaders of our nations we have assembled economists and industrialists to make the optimal production schedules on our behalf because (to put it simply) none of our players are that smart. Players want to influence their productions (they may want to maximize aircraft production, or they might opt for a land war, for example) but they by no means want to micro-manage every single stinking thing. You might think you can automate much of this with a spreadsheet macro (or a small subroutine if one is making a computer game) since the Simplex method is well known. If that is the case, then why include all that rubbish in the first place?

Consider Axis & Allies. Each territory has a certain amount of value, and players are free to purchase whatever they wish with their tax revenues. Ahh, but we might still want some of those juicy strategic aspects such as oceanic commerce raiding. This can be accomplished if one required colonial areas to actually transport their revenues via ship. Blockades and raiders are now permissible, and we haven't turned a good idea into a terrible spreadsheet. This is an example of knowing when to stop painting. So we seek political prestige via foreign policies, and at our disposal are the various proxy wars, diplomatic crises, direct conflict, and various economic machinations (embargoes, blockades, commerce raids, annexation, etc) and we haven't required of our players an advanced mathematics degree. The rules (if using BoP as the starting point) are easy to learn and yet difficult to master. All features of the game enrich the game, rather than detract from the core play. There is both a positive Victory Condition (have the greatest prestige score after so many turns) and a negative Victory Condition (do not initiate a nuclear war) so play is guaranteed to end, yet there are a multitude of strategies available.

All of the previous being as it may, what if you really want to make some sort of economics game? If the warfare is minimized, and play centers on logistics (moving commodities from one location to another) and production (harvesting or manufacturing commodities), then you know who you are, and you also know who your players are. You're going to need to study up on linear programming for the various production functions required to harvest and manufacture commodities, logistics to actually transport your commodities from production centers to consumption centers, and the mathematics of game theory (John von Neumann and Oskar Morgenstern) for your production schedules and end-user consumption. I warn you: the first thing you'll trip over is the labour function. I would suggest reading Mancur Olson's three Earth-shattering works. The second thing you'll trip over is your economy probably won't support Stagflation. If you somehow proscribe that, then you're really missing out of the beauty of falling consumption concurrent with rising price levels. That's the sort of thing which would make your hypothetical Sim Industrialist Tycoon get off the ground. Hint: your simulation will (usually) stagnate unless and until you get a financial market up and running, because currency is a commodity just like any other. Bonus points: Let the business cycle be monthly, but let the banking system operate quarterly, update populations with a variant of The Growth Equation, and run the game for 200 years. It's a hoot!

In conclusion we see much of design theory focuses on identifying core gameplay. Once this has been identified, the game is not complete unless and until there is nothing left to remove! There is rarely anything to add, but usually oodles and oodles of distraction and chaff which must be diked out and chiseled off because extraneous elements make the game worse rather than better. Adding something is not the same as improving something. If something extra is desired, it ought not to overlap with an existing feature. One example would include how the units of Advanced War 2 do not overlap in function. Another would be our improvement to BoP, where we added a new avenue of play but did not demean the existing modes.

I cannot stress this enough. Your MUD does not need two healers. The Martians did not need two tripods. Your fighting game does not need twenty eight different buttons labeled punch. I swear to God: the next time I ask for a unit list and receive an insulting exercise in hyperbole, I will break the fucking fingers of the brain-dead designer who cursed this Earth with their rubbish.

Well, God has arrived. I met him on the 17:15 train.