Generally I don't play science fiction wargames. As a model maker, I am very tempted to get into sci-fi, because it would allow me to let my imagination loose. Often I see a piece of packaging and think to myself that it would make a great sci-fi building, or part of a space ship. The games, though, have proven a disappointment.

The two main types of sci-fi wargames I have tried are space-ships and skirmishes. The space battles were like World War Two sea battles, played in two dimensions, with fleets, admiral ships, aircraft (small fighter space craft) carriers, big guns, missiles, and even things called "torpedoes". The only things missing were weather and submarines. The skirmish games were like simplistic modern warfare games. The troops roamed about in Vietnam-like jungles, carrying guns, going about in squads and using support weapons. There were anti-tank weapons, and tank-like armoured vehicles with big guns, armour, and perhaps legs instead of tracks. The games didn't give me a sci-fi atmosphere.

If I were to come across (or write) a sci-fi wargame that were significantly different in kind from other games, demanding different tactics, then I think I would get interested. The invention of the machine gun turned warfare of open fields, cavalry, and dense formations of troops, into the trench systems of the Western Front in World War One. I want a sci-fi game where some new developments have changed the nature of war again, rather than see the modern world painted in different colours. What could these developments be? Non-lethal weapons? Battlefields without people on them at all, but intelligent robots or remote-controlled craft? Many of the possible futures for war do not lend themselves to being played out on a wargame table at all. Being vigilant for terrorist threats, making buildings and their populations resistant to biological warfare, policing the galaxy-wide-web against cyber warfare, industrial espionage, propaganda wars, massive weapons capable of destroying entire planets – none of these suits the traditional wargame, with its troop deployment and movement.

What then, would a future war be like? Well, the plain fact is that we don't know, but what might make a good game? I demand something that makes the game unlike other wargames, and demands a different understanding of tactics to win. One theme in warfare is the balance between attack and defence. In the First World War, defence was a lot easier than attack, once the enemy had dug in well. Towards the end of the Hundred Years War, castles were vulnerable to the latest cannons, and sitting behind their defences was not a winning strategy. In Frank Herbert's Dune books, shield systems have been developed that filter out fast moving objects, making bullets obsolete. Instead, infantry fight each other at close quarters with knives that they have learned to slip slowly through the shields. Lasers that hit shields cause explosions that kill both target and laser crew. Immediately with these two innovations, we have new tactics. As a sci-fi game, though, a scrum of knife-fighters still isn't appealing.

Detection is a major theme. In World War Two, troops wore camouflage, and victory usually went to the side that saw the other first and got off the first shots. One could imagine a future world in which offensive weapons are capable of flying fast to any target and destroying it, and so detection becomes paramount. This, though, takes us away from wargaming as we know it. Throwing lots of dice to see if a unit is spotted or not will probably prove tedious, and luck will probably play a large role, which is not satisfying. Also, we generally want to paint our pretty miniatures, and then see them on the table top. A game mainly concerned with concealment will require figures to be left off table, and when they are discovered, they will not last long.

The knife combat in Dune meant that the skill of the individual, his training, talent and resourcefulness became dominant over having the better weapon. One could envisage a future in which detecting the enemy by electronic means has become next to impossible because satellites all get knocked out as soon as war starts, and jamming technology has overtaken surveillance technology, meaning that the only effective way to find the enemy is to send skilful troops on scouting missions. In other words, it is possible to come up with rationales for why technology has made no difference, and warfare remains like modern warfare, but this does not give me the unique game I'm after.

One way to create tactics in a game is through combined arms. Napoleonic warfare is a good example. A bit like scissors-paper-stone, infantry square beats cavalry, but artillery beats square, but line beats artillery, but cavalry beats infantry line. The game becomes at least in part trying to get the enemy infantry to adopt the formation that suits you, by threatening him with combined arms. In Second World War games, infantry can destroy anti-tank guns fairly easily, but anti-tank guns can destroy tanks, but tanks at long range destroy infantry, and at short range are vulnerable to infantry. Infantry are the best counter to enemy infantry, so getting infantry to work in co-operation with the other arms is the key. Perhaps a future world can be envisaged in which the arms that combine, and how they combine, is different. Perhaps there are mobile shielding units that render some units near them invulnerable to certain forms of attack, but remain in turn vulnerable to other attack forms. For example, perhaps there are anti-aircraft guns that are so effective that it becomes suicide to fly near them with troop-carrying craft, but once used, the AA guns are easy to detect, and then become vulnerable to suicide robot craft that fly straight at the guns and explode if hit, taking the guns with them. This would mean that you would choose not to shoot if you suspected that such craft were in the area, which in turn could mean that an attacker could exploit this and send in troops carrying craft. On the ground, troops could have very effective detection equipment, but when used, this itself would be detected immediately and vulnerable to everything that had just been detected.

One way of tackling the problem is one used in the game Rivets: a war has been fought, and factories were made that built warbots, and these factories sent out these bots to attack the enemy factories. There are several types of warbot, and each is particularly effective against certain other types of bot, while being vulnerable to other types. The game then becomes largely about predicting what the enemy's force will consist of in a few turn's time, and then manufacturing the appropriate types of warbot to take on that army.

Rivets, though, was not a serious wargame, but a quick knock-about humorous hex-grid game. The notion that troops are being manufactured on board quickly enough to join a raging battle is a bit far-fetched, and makes for an unsatisfying situation, because it suggests two things: in war, victory will go to the side with the greatest manufacturing capacity; and the loss of a unit doesn't matter because it was only a robot, and can be replaced in a jiffy anyway.

I remember playing a space-ship fleet combat game many years ago called Star Fire, and much of that game's interest was in building the fleets in the first place, rather than playing the battles out. My brother built a fleet that had many missile launchers and many systems that allowed him to combine the fire of his many missile launchers against single targets. These targets then got overwhelmed, and he had his super-fleet, so then I had to try and build a new fleet that was good against his and on it went. If the game was unbalanced, then it might be that co-ordinated missiles were the best weapon of all, and only a fool would build any other sort of fleet, or else a balanced (bland, predictable, like all other fleets) fleet was the best, or (and this is the ideal) a balanced fleet would normally be beaten by a specialist fleet, but all specialist fleets had major weaknesses. This is the equivalent of making sure that a game doesn't have infantry square as the best thing against infantry line, cavalry and artillery (American Civil War games can be dull because infantry line beats anything else, and defence beats attack).

Full Thrust is a more recent game, and is very like Starfire in that it depicts two-dimensional space-fleet actions, and can be won at the fleet-design stage. The games themselves are largely about moving the fleets towards each other and then rolling lots of dice to see what happens next. While there is satisfaction to be had from seeing how well-designed a fleet might be, I feel that these games lack skill at the playing stage.

Another problem with sci-fi games is that people can envisage a future in which weapons are so effective that battles will be over in a couple of seconds, after one side deploys a weapon successfully. Perhaps one side detects the other, coming through the jungle on the other side of the planet Delta-9-Epsilon, and one man presses a button, and a small nuclear device detonates wiping out the intruding expedition. One way to deal with this problem is to stage a battle in which both sides are fighting over the thing they want, and therefore cannot afford to damage it. Perhaps the fight takes place inside a vital factory or power plant, for example, or in a densely populated civilian area. Neither side would want just anyone to be able to destroy the objective in order to stop it from falling into the hands of the enemy, because a hot-head could panic and make the decision at a bad moment, so any self-destruct mechanism would have to be limited in how easy it was to use, and who could use it, which in turn gives the other side a chance to stop it from being used. Another way round the problem is to make very powerful weapons suicidal to use. A gun that will punch a hole through an enemy, then through the wall behind him, then through the seals around the base that protect it from the toxic atmosphere of the planet outside, is not a lot of good; nor is one that creates an explosion that kills the target, and brings down the building that target and firer are both standing in.

A major decision that has to be made for any wargame design is what part of the decision making process is represented by the choices made by the players of the game. Is a player a supreme commander, and if so, what information does he have to base decisions on, and how much influence does he have over his forces? In historical games, troops had very limited means by which they might convey what was going on to their commanders, and so there is little excuse for allowing players to micro-manage every tiny facet of each unit's behaviour. In sci-fi, however, one could very well imagine that every soldier might have sensors on him that send back more information to HQ that the soldier himself can take in. HQ might notice things happening in the soldiers' area before the soldiers do, such as suspicious changes in temperature, a 2% change in the ratio of laser fire sounds to gauss gun sounds, or the like. If the soldiers are robots, or creatures bred specifically for combat, then they might be assumed to be absolutely reliable in following orders, and so there is now an excuse for micro-management absent in other games.

This last notion opens the possibility of a new kind of game. Many children are disappointed with their early serious wargames, because they were used to fighting their Airfix figures against each other down to the last man, whereas the serious wargame tries to emulate morale effects, and often large bodies of men run away without even fighting. In sci-fi, it is possible to rationalise how units of genetic infantrymen, or robots, or drugged-up and brain-washed people might continue with a mission down to their last breath. Weapons could be so lethal in the right circumstances, that the requirements of victory become finding and eliminating every single enemy. This might appeal to the infant mentality, and are not wargamers a load of kidults?

One possible kind of warfare is one in which the behaviour of the enemy is manipulated. All manner of alien technologies might be used to alter the behaviour of forces. Brains could be influenced by drugs, gases, hypnotism, vibrations, confusing lights, illusory projections, fiendishly subtle propaganda, perfect impersonations/simulations of commanders, and machines could be influenced by computer viruses, false broadcast information, reprogramming/hacking, signal jamming and the like.

One theme that comes up in sci-fi is the electro-magnetic pulse effect, or EMP. Most people know this as the radiation from a nuclear explosion that is said to knock out all electronic equipment. Perhaps EMP weapons could play a major role in a sci-fi game. Then again, perhaps not. One thing not so commonly known is that EMP only has its effect on Earth because the Earth has a magnetic field. The EMP effect would not occur on the Moon or Mars, because these have no magnetic fields, so a game could be set where electronics are safe. Also, not all electronics are affected by EMP. Old valve equipment is little affected by EMP. Indeed, some Soviet aircraft still carry valve electronics to protect them against EMP, and the World War Two bombers that dropped the first nuclear bombs in action were unaffected by the EMP. To us, valve technology is laughably old-fashioned, and perhaps today's circuit boards and silicon chips will soon seem absurdly primitive, and be replaced by new technology immune to EMP and things like it. Even today, battery operated equipment with no wires longer than about 30 inches are very resistant to EMP, and cars are seldom knocked out because they are insulated from the ground by their rubber tyres, and have metal bodywork that forms an effective Faraday cage around their electronics. Nuclear weapons detonated hundreds of miles above the surface of the Earth would create EMPs that affected whole continents at a time, and so a table top game involving troops and tanks would be at a scale where both sides were equally affected. Non-nuclear EMP weapons exist already, and so there is an easy rationale for more localised targeted weapons. Of course, since electronics of the future might be very different, we are free to make up such weapons and their effects.

Many years ago I played sci-fi role play games. The main one I played was called Traveller. Though I persisted with it for a while, I never found it satisfying, and I think I knew why. When role-playing in a medieval world, we the players know what is possible. In a sci-fi world, since we are not of the future, we do not know the things that people in the future would find obvious, and so we cannot role-play them convincingly. Imagine a medieval man playing in a role play game set in the modern world. He and his fellows plan to rob a bank. They concoct a good plan, but it stands no chance of success, because they reckoned without CCTV cameras and telephones. We today cannot role-play people of the future for the same reason. The referee might decide in his mind what is possible, but the players have to know too, and it has to feel right to both parties. Another problem is the unforgiving nature of modern technology. A brilliantly executed piece of espionage could be ruined by a single skin-scale being recovered by the enemy and its DNA sequenced. In a medieval world, a good adventurer can talk or fight his way out of a bad situation, using his personal skills and wit. In the far future, major societies will have so many resources for keeping order that one false move from anyone could throw everyone in prison or mortuary.

Sci-fi I have concluded doesn't really work for serious role-play games, although I have had fun playing in the familiar world of Star Wars or the comedy world of Paranoia. Warhammer 40K already panders to the fourteen-year-old-boy market. For a serious table-top game I conclude that a very different approach is necessary. Ideally I would like to have to use tactics that bear little resemblance to other periods of warfare, or at least significant differences.

I haven't written my sci-fi rules yet, but I have some ideas. One is that the people of Earth might be fighting aliens who see and hear at different wavelengths from us, and so each side has to be careful how it camouflages itself, and has to use equipment to detect in the other's wavelengths. If both sides have a mix of species, then each has to switch wavelengths to see its selected foe, and each might have machines that can blind certain wavelengths while leaving others clear (an analogy of this is the modern practice of wearing orange camouflage patterned clothing when going deer-hunting, because deer can't see a difference between green and orange, and orange clothing makes accidental shootings of people less likely – now imagine intelligent deer with access to orange clothing and goggles that allow them to see the difference…). I want to avoid teleportation and psychic attacks because they may spoil a game and I don't find either feasible. Similarly mucking about with time-travel or distortion I think I will do without. One side's having to use non-lethal force while the other is free and willing to use lethal force is another idea I think worth investigating. Some weapons that affect a large area in such a way as to degrade some troop types in it more than others could lead to new tactics. An example might be a weapon that vibrates the ground and air in a way that damages hard things or heavy things, while leaving light things, floating things, and soft things intact. My main themes will be trying to envisage weapons and defences that follow the paper-scissors-stone pattern – good against one type of foe, vulnerable to another; and combinations of combined arms that work, and the difficulties of achieving these combinations.

I would be interested to hear any other ideas.