Today, a couple of friends and I got out the pack of 100 army men we bought at the toy store a couple of weeks back, cleared some space on my floor, and tried playing a little battle. We got about three moves into the game and had to go, so it's still covered in army men [grin]

The first step was to figure out what we had to work with. While the pack was of 100 figures in dark and light grey plastic, the numbers were not precisely equal between the colours, and likewise the different types of unit for each color. So, once they'd been sorted by colour and then by figure type, we discarded units (or dipped into the second bagful) until the two armies were equal.

We decided up-front the sort of structure the game was going to take. Movement would be done in 'marks,' which were measured by a long string of bits of paper taped together and marked out at regular intervals (roughly half an inch apart). Each player's turn would consist of a movement phase which gives each unit a chance to be moved, and a shooting phase which gives each unit a chance to shoot. (I'm still pondering a different approach to this that doesn't seperate out the two stages, and allows a unit to move, shoot, and then move again - letting units pop out from behind cover, fire a shot, and get back behind cover again. My fellow players weren't so keen, but maybe we'll try it some other time). A unit could move as many marks as desired up to that unit's limit of movement points for the turn. Shooting would be done based on the range; we'd define accuracy percentages for different distance ranges. To fire a shot, the distance to target would be measured, and a D100 would be rolled - if the value was less than or equal the percentage for that distance it counted as a hit, and then another dice would be rolled to determine the number of damage points dealt.

Once sorted, we had about six classes of unit (probably more like eight, but we merged a few): Machine gunner, Rifleman, Officer, Sniper, Machine gunner with grenade, and Rifleman with bayonnette. We fired up Excel and started defining some stats for each type: movement points, accuracy at <5 marks, 5-10 marks, 10-20 marks, and 20+ marks, the dice to roll to determine damange, the number of total damage points a unit could take before dying, and any notes. The riflemen, of whom there were the greatest number, had 4 movement points, 60% accuracy below 10 marks (30% at 10-20mks and 10% anywhere higher), rolled a D6 for damage and had 5 health points. By contrast, the snipers had only three movement points, and their accuracy increased as the distance increased (10% at <5mks, 20% @ 5-10, 50% @ 10-20, and 80% @ 20+), rolled a D10, and had only 4 points - but they couldn't both move and shoot in the same turn.

We played about three turns each before stopping for dinner. I commanded the light grey team, my friend Tom commanded the dark grey team, and George sat at the computer spectating and feeding us stats as and when they were necessary. Each turn took quite a while - a good five minutes - and much knocking over of pieces with the tape or people's heads as they checked line-of-sight occured.

Despite the fact that we both set up right off at the opposite ends of the field, after Tom had moved his troops, he realised that a few of his riflemen actually had clear shots at mine through all the terrain. So he took some potshots and actually managed to wound a couple of them [grin] I, of course, would not stand for this, and after moving my troops a little closer to cover I took a few potshots back and managed to kill one of his riflemen with a single shot [grin] [grin]

We each tried moving snipers up the ramps we'd set up on top of the obstacles where they'd be able to pin down any units that weren't behind cover. However, moving the snipers into positions where they had clear shots at the other side meant the other side had clear shots at them - and while the accuracy may only be 10% for riflemen at that range, when you've got 10 of them all taking shots at one poor sniper... well, he ain't gonna last very long. Clearly, for snipers to be effective in this kind of way, their movement and operation needs to be kept secret from the other players. I'm not sure how to do this; I considered giving each player a number of pieces of paper, marked on the underneath with a tick or a cross, and letting players move these in place of the sniper pieces - so the other player doesn't know which ones are actually snipers and which ones are not - but I couldn't figure out how to make that work when it came to shots being fired. It seems like it's the kind of thing that needs a third-party enforcer, to track where the units actually are, to calculate and perform shots, and to give the other player the appropriate hints as to where the snipers are so that they stand a chance of taking them out. The computer's suited to this.

The guys with machine guns and grenades are also probably too slow - only 2mks per turn - though they can throw their grenades up to 10mks (+/- D10 -5). These guys again most likely need to be kept secret from the other player to maximize their effect, but then again it may be sufficient just to speed them up and to ensure that they can throw their grenades over cover.

I need to start thinking about formalized rule structures for this. What I'd like to do is develop a context-free grammar that can express the game rules, all the variations I've thought of, and a decent number of other play styles.