PBEM v92 n03 (15Sep92) ====================================================================== Game or Simulation? Luis Sequeira ====================================================================== Five years ago I joined a PBM (not a PBEM, mind you, but a computer-moderated postal game) which is still running now, and which was constantly upgraded from month to month, never achieving its final, mature state (and it certainly never will). The game was one of those with a fantasy setting (based somewhat on TSR's World of Greyhawk, but without magic, only warfare and economics), and its rules, while never having been considered to be simple, have grown and grown to the point that its current gamemaster does not know them all (he is the fourth one trying to referee about 250 players on the same game). There wasn't even ever a rulebook! Much of this complexity was due to the fact that most players demanded more rules and more freedom of choice, thus providing them (as they proclaimed) more "reality". In fact, the reports one got were quite extensive and needed some time to read and interpret. And if you add to that the time required for all your political/diplomatic messages, one could easily spend 4 or 5 hours planning your orders for one turn. But your lost time was quite rewarding in terms of game play. In its primeval times, the game was run on an Amstrad PC, using about 1 megabyte of dBase III code, and one turn's processing look about 8 hours of computer processing time (obviously not counting the time needed to manually process each order form for all 250+ players) and about the same time for all printouts. Now it runs on a 486 clone and its 3 megabytes of Clipper code are processed much faster - but in games of this size, bugs invariably creep in. This was quite annoying for most players - but many of them actually used them to their advantage. For instance, knowing exactly what kind of orders to give to trigger some hidden bug could give that player unlimited money, or unlimited movement points for his armies or navies, or vast amounts of combat points for incredible small armies (1 man!). Some of those bugs were so deeply ingrained in the system that they soon become features (there was no other choice for the gamemaster, except to rewrite large amounts of code, which in turn would get some more bugs for the players to discover and exploit...). On the other side, new rules were created to limit certain facts, but, without careful planning done by the game creators, these rules soon became new sources for bugs. For example, to limit the number of armies, their maintenance costs were quadrupled; but the game allowed the players to slaughter cattle in order to send rations to those armies, which reduced the maintenance costs to their original prices. Cattle reproduced each month by about 3%, and consumed some grain. But players soon complained that having to raise cattle just to get it into their armies was awkward; they wanted to include the cattle in the army itself (which was somewhat logical, as in the Middle Ages this worked that way). This was accepted by the gamemaster and allowed. But he forgot the cattle-grain dependency; now cattle reproduces 3% *inside* the army, and consumes *no* grain at all! Thus, no one ever grows cattle (except for sale), nobody buys grain, and no one converts cattle into rations (except to put it into fortresses to increase the siege time - but this is more expensive than creating *new* armies to intercept incoming enemies). So, just because one bad decision was made, the game has now obsolete rules that no one ever uses, and the "reality" of the game has decreased... This happened with lots of other rules. For example, spies were necessary to obtain information on the movements of enemies and to spy upon their fortresses - but one spy could only look at one hex at the time, so great spying networks were needed. But players complained that they should be able to give *all* the spies one single order and get a map of *all* the enemies' territory, instead of getting information hex by hex. This also was allowed. However, spies were subject to counter-spies and police levels in the countries spied upon, and information wasn't always reliable, and spies *could* get captured and ransomed. Nowadays, for the price of about 10 excellent spies, you get a detailed map of the enemy country (like some kind of satellite photo), which is not subject to any countermeasures by the enemy... Thus, the game degenerated to a kind of World War I battle-and- conquest game, with medieval troops. Communication with your troops is almost instantaneous, the economy has stock markets not unlike those in the 20th century, enormous armies (80,000 men!) may be supported by rich countries, sieges are avoided because it is so much easier to "overrun" enemy fortresses by assaulting them (just 2 weeks' time instead of many months...), pirates and Vikings have fleets to rival today's superpowers, and so on. Not quite what you would expect on a medieval-fantasy setting. Soon the gamemaster abolished the idea that this was a "realistic" game. It's just a game, with its set of rules, but nothing more than that. Don't expect it to work as in the real world, they advised. What is, then, a "realistic" game? I have for some time now divided most complex games into two kinds: the simulation, and the normal game. A simulation is a special kind of game, where you may give realistic orders and get realistic feedback. How this is accomplished doesn't matter much (I'll get to details further on), but the important issue in a simulation is that you can react to it as if you would react in the real world. The other kinds of games have no resemblance to the real world; they have their rules, and you play by them, but they are abstract, not connected to reality. For instance, Chess is without doubt a kind of a wargame. But it is by no means a simulation. The knight movement rule has nothing to do with cavalry charges. This doesn't mean that all simulations must have thousands of complex rules and be terribly difficult to play! You can find simulations with few rules (but still simulations nonetheless) which are easy to play, and non-simulations like Chess, also with few rules, but quite hard to master (I just can't play Chess at all, the computer always beats me). Simulations need not depict *all* of the details of the real world. If you know Avalon Hill's Advanced Squad Leader (ASL), you understand what I mean: one *can* fight large-scale battles worrying about weapons which don't fire or about the amount of ammo *each* of one's men has. You need only to stick with a particular scale you want to simulate and create rules for it. Diplomacy, for example, has simple rules (but is quite difficult to play correctly), and only two types of units on the map - but it simulates only the diplomatic aspect of warfare, and at that, it is quite excellent. There is no need to go much more into details. That computers are excellent for simulations isn't a new discovery -- just think about the supercomputers used for weather forecasting. You can cram all levels of detail you like into them, and, in a game, just show the players what they need to know. For example, in a WWII simulation, you *can* use all the ASL rules to resolve your combats, but HQ (at which level the players sit) only receives a report like "Lost battle at Dunkirk, 52% casualties, 3 airplanes shot down". The player gives orders like "Capture Berlin" and needs not write down all details like "1st Platoon advances through the Brandenburg Arch and seizes the Parliament; all units with heavy weapons deployed here, and here, and there". Such orders would be appropriate on a tactical simulation - on another scale - but never on a strategic/diplomatic one! Of course, if you like, you *could* write zillions of lines just to get your computer-controlled platoon leaders on the field to "act like humans". Please remember to buy your Crays first. If you stay at the diplomatic/strategic level, the most complex game you need to create is something like Avalon Hill's Empires in Arms (which is not suitable for PBEMs at all, but you could make something similar). Here you get to decide which leaders get which armies, and how to join corps into armies. But you don't need to raise horses for your cavalry, or buy guns for infantry! On the other side of the simulation world, you may recreate just the Battle of Waterloo in tactical terms - but then you wouldn't have to include economic rules. You may ask why there are so few "real" simulations around. Well, for one thing they aren't that easy to create. A "normal" game is just a set of rules, and that's that. On a simulation, you must "hide" the rules - i. e. the mechanics of play - under a layer of "real" feedback. There is no point in talking about "Unit 52 stacked over level 3 city has 5 combat factors and fires twice per round, losing initiative when on the far side of river hex without type B-9 bridge and gets only 53% offensive bonus". Wargamese is out; you want to know about "Your 1st army didn't succeed in crossing the bridge and suffered many casualties" and then react to it with "Ok, get to next river crossing and try there". The rules are still there. But you offer the players some kind of "interface" where they cannot possibly know what REALLY is happening (in terms of random numbers, bonuses added and factors calculated), and which they can control and manipulate with everyday language - or "real" war jargon. While the "kernel" of the game must be able to process turns just like other games, the "realism" is experienced mostly on the "user interface" from the side of the user. However, it isn't that easy to combine those two facets. Imagine a classical wargame where one ally says to another "I'll take Berlin, while you advance on Hamburg and bomb it". For us, humans, this is very easily followed on a mapboard. But how do you get a computer to do this? Fortunately, there is no need to get your doctoral thesis in AI to write something like this. "Diplomacy" works fine, and there are surely many other simulations. I think you could also put the "simulation" label on Olympia, too, especially because of its weight on some role-playing aspects. As you see, you can get simulations in many areas, and not only with wargames. Simulations of stock markets and elections are other examples. One final note about simulations: it should be more difficult for the players to find bugs - "features" - to their advantage than in other games. If the orders you can give and the reports you receive are "real", you have an advantage if you think in human terms, not in computer terms. Thus, the emphasis is put on the way you play, not on the rules you need to learn. This is easily seen in most wargames. In a war simulation, to achieve success, you must be a good strategist. In a game, you must know the rules very well, and when to apply them - especially that Rule #2463764, which gives your B Unit those extra 5% firepower you need to smash a Z-15 tank, if you have refueled your unit last turn... About the author: The author was never good at any type of wargame, has always lost (except when playing Amoeba Wars, a 1980 boardgame, where he boasts having been third place more times than everyone else), especially in Empires in Arms where he plays Prussia until the last of his men dies (which is always almost at the beginning of play), does play Chess very badly, does not play Go, doesn't understand anything about Military History, has been writing his first PBEM since 1987 and never finished it, is proud of being able to write four-word English sentences without (many) mistakes, and still thinks XConq is the best game you can get on a Unix machine, or role-playing games on a Sunday afternoon. He can be contacted at bc@fccn.pt at almost all the time (but is quite inaccessible from most sites in the world, including from the computer on the same network next door), does not read news (proudly never installed on *our* site!), and is playing something 90% of the time and sleeping during the last 10%.