It's tricky to get a decisive result.
I designed a skirmish system, to allow me to play with my wargaming figures as individuals. It came from playing RuneQuest scenarios that involved too many characters. The games took ages. I played a few games of Advanced Heroquest and was inspired to come up with a system better than it, that would be realistic in that it accorded with my beliefs in how ancient weaponry behaved, and that would suit wargames where I covered my table in terrain and set my miniature warriors loose.
My first attempt at the new system had one big flaw, which was that each figure had a number of "wound points" that he could sustain before he died, and this meant bookkeeping, and it meant that the game took too long. A later draft of the rules got rid of this complication, and instead had combat results of: unaffected, driven back a bit, wounded, and out of the fight. All wounded figures suffered the same penalties, and could be marked as wounded with a plumber's washer.
After many games, the system was working well, and I could simulate fights between men armed with a variety of weapons, on various terrain types, and things worked well in that I could get satisfactory results in fights between a few figures, but the games overall were usually not great. I would give each player what I deemed to be a reasonable force, well enough matched with his foe's, and then start the game. If both sides were controlled by competent players, I found that the amount of luck each had in the many little combats evened out, and each side would be whittled down at about the same rate. The games never were decisive enough to be interesting, but were just exercises in micro-management and attrition.
It is my feeling that skirmishes in real life are if anything more decisive than large battles. Men involved in them can see a high proportion of what is going on, and the loss of a few men might collapse one side's morale. Men are often within shouting distance of each other, and events are likely to happen quickly. Once one side has gained the advantage, the other is likely to run for it.
I can remember games I played with my toy Airfix soldiers when I was a small boy. One side would nearly annihilate the other, and sometimes it would come down to a fist-fight between the last two men standing. Though this seemed fun at the time, this isn't realistic.
One way to solve the problem is with rules for the overall morale of one entire force. At some point, the morale of one side cracks and its fighters flee. This is realistic, but it doesn't give you a good game.
You could define a point at which a side will flee. For example, when one side has lost half its men, it flees and the game ends, with the other side as the victor. The trouble with this is that I found that when one side had lost half its force, the other side had almost always lost almost as much, so the result still wasn't really decisive in feel.
You could come up with a die-roll system that requires one side suffering some adverse result to make a die roll, and if failing it, run away and cede victory. Again this doesn't give you a good game, because in small skirmish games, a force is small, and a few casualties and a bit of bad luck could mean that a player could lose through no fault of his own.
My first victory system required the fight to continue until one side was doing significantly better than the other. This didn't work either, as the game would normally have to go on for hours and the casualties pile up before this situation was reached.
The first practical solution seems to be good scenario design. A skirmish scenario should not require both sides to try and destroy the other, but should instead define some definite end-point at which victory can be resolved. For example, a scenario might require one side to escort a person safely across the table. If the VIP makes it, the game is won. If not, it is lost. Another example might be to define a moment when overwhelming reinforcements arrive for one side, chasing the other side off, and victory goes to the side that had done better up to that point.
The next solution is an initiative system. With one of these, players may not move every man they have as far as he can go every turn. When players are able to do this, they quickly see developing threats and counter them, leading to stalemate. If one side is able by some means to make several actions before the other can react, then a decisive game is possible. When one side has the initiative, it must choose wisely what it will do with its opportunities. Assessing priorities becomes very important, and this means that players have to make difficult decisions, and this leads to a good game.
I have now devised three initiative systems, all of which have at some time given me a good game. One involves cards. When the card of a particular set of troops comes up, those may act. This is quite good for multi-player public participation games, because each player seldom has to wait long before getting the opportunity to do something. It also can produce a fun game for a small number of players because when the action gets to the crucial parts, the tension of waiting to see whether a card that will allow X to happen before the card for Y comes up can get quite intense. There are problems with it though. One that is easy to describe is that it is not much a matter of skill that determines which unit acts next, since a player will almost always take an opportunity to do something when he can.
The next system I came up with was inspired by the system posted on the internet called "Rencounter". This involved a player's rolling to see how many activation opportunities he has in his turn, and then deciding what to do with them. Certain troops are very easy to activate, while others are not. This requires good decision making on the part of players, and is much better for games of skill. It encounters problems with multi-player games, since if the three players having a turn before you roll high numbers for activations, then it may be a while before you get to do anything, and if you then roll low, and fail to activate anything, this can be somewhat frustrating. A temptation, commonly yielded to with this system, is that people ignore their low-quality and difficult-to-activate troops, and rush ahead with their good troops, who then inevitably get themselves killed. You might say that stupid players deserve to be punished, but I have seen games ruined by many players falling into this trap.
The most recent system I have concocted is based on no other, but is instead, though I say so myself, rather original. It works very well for two-player games, but I am yet to test it with more players than this. The other two systems both suffer from the complication of mixed units. If a gang of men involves a rabble of peasants and two doughty and valiant knights, and the card system brings up the card for the peasants, do the knights get to go with the peasants, or do they stay where they are and watch the peasants make off on their own? I have come up with solutions to these problems, but I have found that players tend not to understand them, and so end up grouping their figures together by type, and the game ceases to be a skirmish game as I understand it, and becomes more like a battle game, with units of like-typed troops moving around in blocks.
My new system gets round this by allowing a player to do whatever he wants with any troops he wants. After each action, his opponent has a chance to snatch the initiative off him, affected by various tactical factors. It is still not a perfect system. Low-quality troops are almost as easy to move as high-quality ones. Perhaps I should combine it with the other two systems, but then might it not become unwieldy?
Good scenario design is the key. Initiative systems help, as do morale systems, but there is no substitute for a good story, and a scenario is like a story with a few possible outcomes and sequences.
A while ago, I played a few pirate skirmish games, and found them rather boring. In 25mm scale, a pirate ship is very big. There was not much room for manoeuvre on the table, and it was difficult to see why a pirate ship would want to attack another pirate ship. By and large, real pirates attacked merchant vessels, not each other as in the movies. Furthermore, if the two ships were evenly matched, then the only thing to do was to grapple the other and pour as many men as possible onto the opposition's ship. Each side therefore adopted identical tactics, and ended up alongside each other. Then the players rolled an awful lot of dice to fight each figure against its opponent. There were no difficult decisions to be made. Players just kept rolling and hoping for luck. You can see that this would be a dull game. If one ship out-matched the other in boarding party strength, then the other would want to escape. If it did succeed, then it would sail off the table and there would be no boarding action and a rather short dull game. If it failed, it would then be boarded and almost certainly lose. Again: dull.
My attempt at a solution has never yet been play-tested. It was a system for resolving a fight such as the boarding action mentioned above, not by fighting every man against his foe, but by coming up with two grand figures: one that represents the chance of a resolution to the fight in a given turn, and another that represents the chance of victory for one side. You keep rolling every turn to see if the fight has been resolved, then if the roll says that the fight has indeed resolved, you roll to see who won, and then consult a chart that allows you to calculate the numbers of dead, wounded, captured and fled.
The idea is to use this as a means to keep scenarios fast moving and fun. Imagine, then, a game where pirate ships do indeed clash. Each turn, we see if any boarding action has resolved. Each turn, those not involved in the boarding actions get to do their bit. So, a little jolly boat slips away from the fray and is rowed to the shore, and a few men get out with their treasure map, and go looking for the buried booty. Will they find it and bring it back before the fight ends? What will they do if their ship is taken while they are away? Will the rival shore party find them before they can cross the little island to another friendly ship? The interesting actions take place involving the few characters away from the action, and the main fights are just decorative backdrops. If I ever get this system to work, I'll tell you.