Wargaming Machines:
A Site for Computer-Assisted Historical Miniatures Wargames


Design Approaches for Computer-Assisted Games


Despite the idea being decades old, the number of computer-assisted miniatures wargames which are publically available remains fairly small. The effort and skills needed to create such games means that - unlike paper-and-dice rules - we do not see a profusion of different sets. Developing wargames rules has always been an integral part of the hobby, and most of us have probably acted as developers or playtesters in our day. When you write a paper-and-dice system, the thing you produce is a document - when you are done, you use it yourself, but you can easily publish it on the web, too, with little additional work. The same cannot be said for computer-assisted miniatures wargames: the effort involved in publishing the system goes far beyond that needed to develop a program for your own use (testing on different platforms, documentation, etc.) I am guilty of developing many systems which I have never attempted to distribute, and I suspect that there are other wargaming developers out there who do the same thing. It makes sense.

Because of this, however, we do not see as much innovation or creativity in the application of computers to miniatures wargames as we otherwise would. Gamers have developed a wide variety of paper-and-dice miniatures wargames, using various approaches, and they differ in some important ways. Consider a traditional design with a fixed I-go-you-go turn structure (A moves, B moves, A fires, B fires) and then compare it to a game with a variable sequence such as that found in Arty Conliffe's Crossfire - the two are very different. In Arty's game, one side keeps going so long as it is successful, retaining the initiative while the other side is only allowed to react. Think about Piquet, where everything is done at the whim of card decks. There are many such examples of how even fundamental aspects of wargames design can be different, and still produce satisfying games.

Sadly, the same cannot be said for computer-assisted miniatures wargames. We have too few examples of how computers can be used to help us play miniatures wargames, and the ones we do have represent only a small fraction of the possible approaches. Computers give us the possibility of designing games which would be impractical using paper-and-dice systems, but we have yet to explore a lot of their potential.

Historical wargamers tend to assume (from what I can tell) that all computer-assisted miniatures games are similar to the popular Carnage & Glory system. I have played it, and very much I enjoy it - I would recommend the system to other gamers. At the same time, many of my own wargames designs are very different. When we decide to employ a computer to run a miniatures wargame, we need to answer the simple question "Why?". One answer is that the computer will do things which paper-and-dice systems are bad at (running complex algorithms for determining results, book-keeping, simulating the fog of war, etc.) while otherwise leaving the rules systems similar to paper-and-dice ones. I would argue that this is the basic idea behind most computer-assisted miniatures wargames systems today. But if we start to think about what else might be possible, we can see that the current approach is still one that implicitly accepts the limitations of paper-and-dice systems. Because we are used to using such rules, we unconciously assume that all wargames must work in more-or-less the same way. Computers offer us the potential to change that.

Before we explore the realm of possibility, however, it is useful to have some idea of what is being done today. This article attempts to categorize and describe the different computer-assisted systems with which I am familiar for simulating tactical combat. There are several systems which offer a means of running campaigns - this is a significant aspect of computer-assisted wargaming. For our purposes here, however, I will limit discussion to those systems which moderate the tabletop conduct of battles at a tactical or grand-tactical level.

I will make a further caveat: it is not easy to classify every game as one thing or the other. Many systems are hybrids, or involve more than a single approach. While I will mention examples here, I will not try to classify every game system. Rather, I will discuss the approaches to various aspects of games design.

Automating Existing Paper-and-Dice Rulesets

This is, I think, where many of us started out when we decided to use a computer in our wargaming. It makes sense - some aspect of a rules set is annoying, or could be made easier, by using a computer to do it. The way in which the Wargames Systems games work is a good example: they took the popular WRG rules sets and made some of the thornier aspects of using those rules easier. They do this in several ways - by performing calculations for you (including doing all the look-ups for charts), by keeping track of the turn sequence for you, and in other ways. The stated goal is to speed play, while not sacrificing the realism by simplifying the system.

AWIRules is another example of this approach. Will McNally's AWI rules are an existing paper-and-dice set. The computer-assisted version simply takes them and makes them easier to use.

My first computer-assisted wargame system was developed as an automated version of the Canadian Wargames Group Great Battles of WWII, Volume I: Canadians in Europe. I liked the grand-tactical level of the rules, and the fact that they spanned multi-day battles, with the units of maneuver being battalions. I didn't like the book-keeping. I did make some changes to the way in which player initiative functioned, as the system they used didn't fit my software design, but otherwise followed that system closely. The system evolved away from CWG rules over different versions, but I still play it and enjoy it today.

In most cases, this type of computer-assisted wargame differs very little in its fundamentals from what was found in the paper-and-dice system which inspired it. They may be faster and easier to play, they may be less annoying in terms of book-keeping, but they are fundamentally the same games. This is in-and-of-itself a compelling reason to use a computer during play.

(A related type of computer program is one written to enhance the use of paper-and-dice systems. Into this category would fall such programs as the Berthier Campaign System, and also a game I have written (and may some day distribute - I'm still playtesting it) which provides a mini-campaign for The Pikeman's Lament in the context of a Vauban-type siege, which I just call Siege!. There are no doubt others. These games do not fall within the scope of this article, however.)

"Automating the Pencil" Wargames

The phrase "automating the pencil" is one that may be familiar to those who have worked as software implementors in an enterprise context. Often, when applying a computer system to a large business, you can see that there would be tremendous value to the company if they would modify their existing business processes to better leverage a particular technology. Instead - because most companies don't like having their business processes dictated to them by technologists - they implement less powerful solutions which leverage only some part of the technologies' potential, but more closely resemble existing practice. (I do not mean for this phrase to be derogatory: it makes sense in many cases, since reengineering business processes is an expensive and complicated thing, and the costs may outweigh the benefits of fully utilizing a given technology.)

The same idea applies to many computer-assisted miniatures wargames. I would put Carnage & Glory and the tactical-level systems of Computer Strategies games largely into this category. They allow you to run a computer-assisted wargame using a familiar process, even though not strictly based on any single existing rules set - in addition to being computer-assisted, they are a rules set of their own. There are many compelling reasons to use such an approach - you can leverage the benefits of a computer without significantly altering the player's experience of a wargame. The game is still run according to a familiar type of process (turns, phases, charges, firing, melee, morale checks, etc.), but the mechanics are handled by the computer. The computer-assisted miniatures systems which I use most often (my own creations) are designed like this.

Because such designs are not tied to paper-and-dice mechanics, however, they leave designers free to better-leverage those things which computers are good at. As an example, when I was creating a Marlburian system I wanted to have a charge sequence which involved sixteen different steps, many of them conditionally dependent on the outcome of preceding steps. It was a complicated process (and, I like to think, a fairly detailed model of the historical event). To use this mechanic in a paper-and-dice rules set would have been prohibitively cumbersome, to say the least. In the computerized system, you click on the Result button and it tells you what happened (if the unit charged and if not why not, who fired and at what range, how many casualites occurred in which unit, who panicked and ran, etc.), and where to move the figures on the tabletop. Behind the scenes a fairly complicated algorithm is being used, but for the player, making a charge is simple. The mechanic is more realistic than the one I would have used instead if I were designing a paper-and-dice game, since a normal paper-and-dice process would need to be made simpler, with the compromises that entails. This is still one of the few Marlburian rules set I have played in which cavalry units "thread" each other (where one cavalry unit literally charges through its enemy and out the other side) - a fairly common result historically under specific circumstances. (When my club was playtesting Under the Lily Banners - a game I quite enjoy - I suggested this as a possible outcome, and it was included as an optional rule. I do not believe it had been considered by the original designers. It just doesn't fit into paper-and-dice systems very easily.)

This design approach will probably be with us for a long time, because it is a good one. I have experimented with other types of designs (some of which I will describe below), and they are not always successful. In every case they involve making changes to the basic structure of a tabletop wargame which potentially alienate players. The "automating the pencil" approach does not run this risk.

"Environment" Wargames

Many years ago, I wrote an article about this type of wargames design, which I have republished on this site (the old one has gone extinct). Basically, the idea is that you alter the game so that instead of taking the perspective of the unit making the action (the "actor"), you force players to take the perspective of the commanders, by altering the sequence of play, forcing an evaluation of units based on their local "environment." What is seen on the tabletop is subjected to a time lag, in essence. The mechanic for doing this is to delay the resolution of combat (and execution of all orders, in fact) by changing when in the turn sequence combat resolution (or other action) is performed. This mandates some major changes to a game.

The game sequence involves going through the list of units randomly, and evaluating their actions based on their environment (orders last received, what fire they are subjected to, whether they are in cover, etc.). At this point the players can issue new orders, as part of the sequence before any resolution of combat or other actions. Orders may be received in a timely fashion or not, based on the proximity of the commanders to the unit.

I wrote several different games using this approach, and the only one which really worked was a computer-assisted game (the others were paper-and-dice games which used card-driven systems. These ended up not working well - there were too many factors for play to run smoothly.) The reason I include it here as a computer-assisted game design approach is that using computers, it becomes workable. It still violates players' expectation that the minute an order is given it will be instantly obeyed, and the outcome known, however, which alienates a lot of them. They don't like to wait.

This is a system which is essentially embracing the chaos of the battlefield. The only paper-and-dice system I've played which tried to achieve the same effect was Piquet, a game which tends to produce strong positive or negative reactions in wargamers. The designers of Piquet have my admiration in many ways - I share a lot of their opinions when it comes to why traditional wargames systems are unrealistic, failing to properly reflect the chaos of battle. But I think the challenge of designing such a system using paper-and-dice mechanics is very, very difficult to meet. Computers make this proposition much easier to manage.

Computers are very good at storing information, and then revealing it to users selectively on the basis of their programming. (If you are familiar with internet protocols, think about how much information is hidden from users when they send an e-mail or interact with a web site. If all that stuff was revealed, most people would just be profoundly confused.) This same technique can be applied to tabletop miniatures games. What we see on the tabletop in a traditional wargame ostensibly reflects the current state of the battle, within certain constraints (everything during a turn is happening simultaneously, even though we progressively move and fire each unit, etc.) Our primary record of the state of each unit is recorded on the tabletop by the models, supplemented by the use of markers of various sorts. When you take casualties, you remove figures/bases, etc. Using a computer, however, we can hide information without losing it, and only reflect it on the tabletop at the point where it becomes known. If players are willing to accept this degree of uncertainty and chaos in their games, the games become more realistic.

One of the difficulties with an environmental design is that you are showing the same tabletop to both sides, when different, opposing commanders might have different knowledge of what is occuring. The solution is twofold: either you make some compromises, thus weakening the effect you are trying to produce, or you put all of your players on the same side, and essentially have them play against the computer, assisted by a referee. If you compromise, the first commander on either side to 'know' the outcome of an action will cause it to be reflected on the tabletop, even though his opponent shouldn't know about it yet. The opponent's ability to react to it is, however, appropriately delayed. This is not ideal, but it works. The second answer - pitting all of the players against the computer - seems to work much better. I address this type of design in the next section.

This style of wargame represents something which shows the possibilities that computers allow us to consider in designing games. It is not enough, of course, just to implement a cool theory - we must do it in a way that produces games which are both historically realistic and still fun to play. Any design mechanism we invent will involve trade-offs, of course, and sometimes the cure can be worse than the disease. We can only learn through experimentation.

Pitting Players against the Computer/Event-Driven Designs

This is another approach which is not inherently computer-assisted. The first game of this type I remember seeing was a set of paper-and-dice rules for simulating the US attack up San Juan Hill (Kettle Hill?) in the Spanish-American War. Since there is a massive force imbalance, the battle doesn't fit a lot of traditional wargames designs very well. Because the Spanish mostly just sit in their fortifications and shoot at the US forces, however, the subject is a great candidate for this sort of game design - the Spanish actions can essentially be pre-programmed realistically. It is an approach which lends itself naturally to computers, however, since they are able to extend it to subjects which are perhaps broader in scope than the Battle of San Juan Hill, but still involve fairly predictable behavior on one side or the other. My first attempt to code such a game was set in Vietnam, with the players running a US column through the jungle, trying to reach an isolated LZ. As they moved, the communists would pop up and attack, melting back into the jungle when the US reacted in sufficient force.

Hand-in-hand with this approach is what I call event-driven design. Many wargames use some mechanism for random events (typically cards or a table you roll on at the beginning of the turn.) Sometimes these can be a key aspect of play, introducing a level of unpredictability into a game which is arguably very realistic but can also be chaotic (Piquet is the poster-child for this, but many games use a similar mechanism to a greater or lesser extent.) Event-driven designs take this approach to an extreme degree, where the primary activity of the game is responding to randomly-generated events. When you use a computer instead of a card deck, however, you can tune your events specifically to player actions and to other factors, which may or may not be known to players.

These games pretty much demand a referee, unless they are being played solo (and they are good for that). I have started using this design approach in a lot of my games, and I think it works quite well. As a designer, you need to create a mechanism for determining the condition of and possible strategies for the non-player side. Probably the best way to illustrate this is with an example.

Let's say we have a game where the players are commanding an Allied force in WWII in Europe, during the advance after the fighting in Normandy. The players are commanding low-level units - platoons, say, or companies. The computer will play all the German forces. At the start of play, the computer will select an overall situation regarding the Germans: the sector is held in stronger or weaker force, etc., by some specific unit types with some set of standing orders. The psychology of the German commander - aggressive, cautious, etc. can also be randomly determined. Call this a 'force profile.' From this randomly determined profile of the German forces, we can derive an intelligence summary for the Allies, which might report, for example, limited reserve forces in the area, when in fact there are two SS panzer divisions. (Say, didn't something like that happen at Arnhem?) The game could even provide competing intelligence assessments, and force players to choose which to believe. (Say, didn't something like that happen at Arnhem?)

Once play begins, the appearance of German forces is randomly generated, informed by the Allied player's actions, and the German force profile determined earlier. A rapid advance without benefit of reconnaissance would cause German troops to appear on the table closer to US forces. A deliberate advance would would mean that German forces would be detected earlier, but would also allow them to be better-prepared, and to have concentrated in greater strength. Combat, being a noisy activity, would also attract the German's attention. Allied activity in more than one area (say, from paratroops dropped behind German lines) would produce competition for resources on the German side. A cautious German commander might hold back reserves where an aggressive one would have committed them.

What the Allied players see of this is only the appearance of German troops ('events' in game terms). The profile will inform what type of German troops are encountered - are you up against a well-equipped SS panzer division, or a ragged reserve formation with little artillery and no armor? The computer hides this from the players, but will betray it as they play. The German profile can also be altered during play by higher-level random events - what happens if the sector is being reenforced? Resistance will suddenly thicken, and different troop types might appear (tanks, etc.). What if the weather turns, and the Allied air superiority can no longer be relied on? (Such factors can inform one another - reports to the Allied players of an enemy armored column headed their way won't be received if the Allied airplanes aren't in the air to spot it.) You get the idea. The referee would play the German forces which do appear, according to whatever guidelines the game designer specifies. (Obviously, SS Panzer Grenadiers will put up a more determined resistance than Eastern European conscripts, be more aggressive, etc. They are also much more likely to have Tigers!)

You might observe that such a game could be created using paper-and-dice mechanics, and you would be correct. Specifically, the use of card decks for randomization could cover a lot of the needed ground. What the computer brings to the table is the ability to hide a lot of what is going on behind-the-scenes from the players, and to allow random events to happen very specifically, according to a situation which is only exposed to the players as the game proceeds. Computers can do this seamlessly, where a human referee would need many different card decks and charts to create the same effect. Think about the mechanics needed for adjusting the random appearance of the German forces. A card would need to be drawn every time an Allied unit made an action. The possible results of "drawing a card" would number in the hundreds or thousands, based on the German force profile, what German forces had already been committed, their commander's psychological profile, the Allied activity taken, and various prior events (weather, reenforcements, etc.). Describing such conditionality is easy using a computer. With decks of cards and charts, it isn't impossible, but it quickly becomes complex and unmanageable.

"Experience of Combat" Games

This was a new type of computer-assisted miniatures game to me - the only example I've seen is Panzer Combat II. Here, the game simulates the actual experience of someone engaged in battle, rather than the experience of battle from a command-decision perspective. As such, it is a very low-level form of simulation (players control tanks directly, rather than ordering formations of them around the battlefield.) Unlike command-perspective games, this is a much less abstract thing: in Panzer Combat II a web-cam is used to "see" the battlefield (that is, the targets of fire), and the computer view gives you what is essentially a tank commander's view of the action (along with sound effects).

Such an approach is really very different from traditional computer-assisted wargaming designs. (It may be that this should be classed as a "miniatures-assisted computer game".) It focuses on using the tabletop in a different way, as a literal depiction of the action, and not as a semi-abstraction. This allows it to involve the player's senses directly (you aim a camera instead of a gun, but it is still actually, physically aiming, although the computer resolves fire). Regardless, it will be interesting to see if this type of game catches on - everyone's smart-phone has a built-in camera, which means that such an approach is more viable than ever. It would be interesting to see how it could be applied to other periods.

In the past, people have attempted similar things without benefit of a computer: you may have seen "periscope games," where the battle is conducted on a raised table, and players can only view it through a periscope positioned to simulate the view their command figure would have. I remember once running a colonials game inside a very large cardboard carton, with aperatures around the sides allowing players to look through a paper-towel tube to achieve the same thing (I think we called it a "telescope game"). As I remember, the game produced the desired effect - the British players completely over-estimated the strength of the Afghan forces, and suffered a humiliating defeat! The set-up and logistics were a nightmare, however. These kind of ideas aren't entirely new, but computer technology makes them much more practical, more effective, and more flexible.

What we learn from this type of design approach is that the question behind any wargames system is really "Whose experience of combat are you trying to simulate?". This will come as no surprise to any games designer - they have always needed an answer to this question in order to make the trade-offs necessitated by any type of miniatures wargame system. Involving players directly in combat at a personal level, as we see in Panzer Combat II, is just a different answer to this question. Computers and modern technology generally give us a better arsenal to use in designing games - one which allows us to broaden the scope of the simulations we play. The challenge is to apply that technology in fun and creative ways, while still getting to play games with our toy soldiers.


The possible number of game designs is infinite, and computers do not make a big difference in all cases. I do not believe we will ever see computer-assisted games replacing paper-and-dice games completely. Computers do help to address some of the failings of paper-and-dice systems, however, and they allow us to conceive and develop games which we practically couldn't without them. This is in no way a comprehensive review of possible computer-assisted wargames designs, but it does point out some approaches which hopefully will get designers thinking about what they can do differently to make better, more interesting games. Many of the old answers to old questions may need to be revisited - what was impractical using paper-and-dice mechanisms may be simple and elegant in a computer-assisted format.

Copyright (c) 2018 by Arofan Gregory. All rights reserved.