It's usually too deadly.
It seems to be the case that before gunpowder became common and firearms effective, shooting seldom decided any battle. Even famous encounters like Agincourt, where archers seem to have won the day, were decided in the end with melee. Henry V's archers certainly caused many wounds to the French, broke up their charges, and bunched them up ready for the slaughter or for drowning in shallow mud, but the French still did advance, and even Henry's archers had to get stuck in with swords, hammers, and daggers to finish the job.
In a skirmish game, archers do not shoot arrows a very long way, to drop them on to large formations of enemy troops. Instead, men shoot with flat trajectories to pick off individuals at short range. The chance of hitting a competent individual foe at long range is so slight that long range indirect shooting is best ignored in skirmish games. The most likely effect of such archery would be to get someone angry. To create the rain of arrows that can have an effect on a large formation of troops requires large numbers of archers, and is something for battle games, not skirmish games. In battle games, an individual archer is pretty much insignificant. In skirmish games, mass archery should be relatively ineffective.
I have done a lot of re-enactment and live-action role playing. One thing this has taught me is to use ranged attacks against people who are not paying attention to me. If there are thirty men on the opposition's side I can see, and I have a bow, then I would be a fool to shoot at the guy who is eyeing me suspiciously, clearly waiting for me to shoot, and who will almost certainly side-step my arrow or take it on his big shield. Instead, I am far better off shooting at the guy who is looking at someone else, perhaps watching his unit's flank, or getting orders from a commander. He won't see the arrow coming, and I am therefore far more likely to hit him. One archer can do this. Fifty cannot.
Imagine that a group of twelve soldiers is walking through a village. No one is shooting at it. The men walk through the village with ease. Now imagine that one archer is sniping at them as they go. All of them will now have to be careful, perhaps dodging from one piece of cover to the next. The sniping archer will be looking out for any man not in cover and stationary long enough for him to get a shot in. The difference between not being shot at, and being shot at by one man is therefore huge. Now imagine that a second archer is sniping from the same place. Will the behaviour of the twelve soldiers change much? Will they dodge twice as hard? No, and the number of them likely to be hit is likely to be little more than with one archer. In short, two archers shooting from one place are nowhere near twice as effective as one.
Imagine a line of twenty swordsmen. Opposite this line are twenty archers in another line. The swordsmen charge their foes. Every swordsman will be very keenly aware of the arrows coming at him, and none will be caught napping. If an opportunity presented itself, such as a swordsman's stumbling over for a moment on the way, then probably only a few of the archers would be able to exploit it, and even if they all did simultaneously, that would be twenty arrows shot at one enemy man, giving a greatest possible result of one enemy casualty. The archers will not all be men of the same calibre. Many studies have shown that in war, a small minority of men does most of the killing. Most men in World War Two did not shoot to kill, but instead if they shot at all, they shot to be seen to be doing their bit, and to frighten or suppress the enemy. In a skirmish, one should therefore imagine that of those twenty archers, only a few are really aiming to kill, and the rest are just hoping that the charge will come to a halt. Some might be hoping that the shooting done by their fellows will do the job, sparing them having to do any shooting, or perhaps even covering their retreat. Again, this points towards the same conclusion: twenty archers are nowhere near twenty times more deadly than one. If one man charges one archer, the archer will have to shoot to kill to defend himself. Perhaps the charge has been well-timed, and he won't get an arrow off in time, or perhaps it was badly timed and the charging man is very vulnerable. We write wargame rules to reflect this, and we give an archer a decent chance of hitting his man.
Unfortunately, we tend to leave it there. I have played many skirmish systems, and so far they all have a mechanism for working out the chance of one arrow's hitting and damaging an opponent, based on things like range, skill of archer, armour of target, amount of cover, and so forth, but then there is no mechanism for applying this chance to mass shooting.
Imagine that a figure is in a farmyard in a wargame, and for reasons we can only guess at, 10,000 archers are trying to shoot him. Now imagine that after the first 5,000 have shot, the man remains unscathed. Does the second 5,000 stand the same chance as the first 5,000 of hitting the target? I would say no, and here's why: if so many men shoot and all of them miss, then we establish that the target cannot be hit for some reason. Perhaps he is hiding in a well. Perhaps he isn't there at all, but is hiding around the corner, and no one saw him go there. Perhaps he is under an up-turned bath. If you were the commander of the archers, and saw that 5,000 arrows missed, would you not think that there must be something wrong? It's like tossing a coin 5,000 times and getting tails every time. Wouldn't you suspect that there is no heads-side on the coin?
With pretty much every task, there is a law of diminishing returns. In melee, one man can put another in danger. Two men put him in a bit more danger, but not twice as much. Only about four men can effectively fight against one other, and most skirmish rules have little trouble modelling this diminishment of returns in melee. Shooting is a different matter, though. Any number of archers might choose to shoot at a single target, effectively dooming that target if no law of diminishing returns is modelled into the rules.
One thing I have come across many times is the inverse square law. This comes up again and again in science. Breeding rates of animals, intensity of sound, effect of gravity, radiation of light, all these things seem to have laws, or good rules of thumb, involving square roots. To give one example, two cellos playing side by side are not twice as loud as one cello. Four cellos are twice as loud, and nine cellos are three times as loud. This is why full orchestras don't make people's ears bleed, and why soloists can be heard above the rest. It strikes me as reasonable that mass shooting in skirmishes should be about as effective as the square root of the number of men shooting, times the effectiveness of one man.
So, four archers should be about twice as effective as one, and nine archers should cause on average as many casualties as three singleton archers, and to be four times more effective should need about sixteen archers. How a particular rules system does this will vary, depending on the mechanisms used, but in my fast play medieval rules, I found that one very simple rule got amazingly close to the required result. If the first archer of a group shooting at a target shoots at full chance, and each subsequent archer needs to roll one higher (on 1d10) to get a hit (11 and all higher numbers are counted as 10), then the effectiveness of archery is almost exactly the square root of the number of men shooting. I tested the rule at long and short ranges, and with 4, 9, 16, and 20 archers, and the results were so close to the exact square root, that I was satisfied. This mechanism would not work for very large numbers of archers, but who ever heard of fifty archers shooting a single volley in a skirmish wargame?
I have played using this rule, and it seems to work pretty well. Archers and other missile users are still very useful. They can harass the enemy, pick off men in ones and twos, and force an enemy to act rather than stand and take a long series of missile attacks.
It should be borne in mind that the effectiveness of archery is greatly altered by the amount of terrain on the table, the size of the table, and the initiative system used. I had been using sets of rules without any rule of diminishing returns in skirmishes for ages, and had not noticed that archery was too effective. However, I always played with small groups of archers and lots and lots of terrain. Archers therefore seldom caught sight of the enemy for long enough to shoot more than a couple of arrows, and the enemy was seldom in sight of many archers at once, and so the problem was not obvious.
Recently, I played a very large six-player game with open terrain. The forces set up with the front ranks just within bow-shot of each other. Archery was very effective in this situation. I don't like to balance the rules by forcing players to pick forces with very few archers. This seems artificial and to be dodging the real fault of the rules: that mass shooting is too effective.
The most simple initiative system is one in which a player has a turn in which he uses all his forces, and then his opponent has a turn in which he uses all of his. With such a system, it is possible to predict exactly how many arrows an archer will be able to shoot against a foe charging him. In my fast-play rules, an archer has a range of 30", and a typical swordsman charging him would move 6" per turn. If he started 31" away, then he would get shot at when he was 25" away, and if he survived, again at 19", 13", 7", and 1", which is five times. In 25mm scale, an inch is about six feet, or two yards, which means that an archer got off five arrows as a man charged over sixty yards. If the charging man was moving at half the speed of a modern sprinter, that's still one properly-aimed arrow every 2.4 seconds, which is pretty good going under pressure. That's 25 arrows a minute.
The basic questions being asked of all the die-rolling are: What is the effect of the archery against the charging men? Does the charge go to contact, or is it brought to a halt or broken up? Do the charging men arrive in a fit state to beat the archers in melee? It strikes me that the overall effect of five volleys is going to be too predictable. In reality, I'd guess that either the charge would slow down and break up, allowing the archers to shoot more and more arrows into the enemy until they broke and ran away, or else the charge would be swift and terrible, and the archers would only get off one or two good volleys and the rest would be panicky and ragged shooting. The easiest way I have thought of so far to simulate this, is to use an initiative system that means that the chargers might get to move more than once between shots at them, or that the archers might get to shoot more than once between moves towards them.
The exact mechanisms for initiative are beyond the scope of this essay, but the big point of relevance here is that initiative systems will generally lower the effectiveness of archery, and sometimes by quite a bit. People will charge when they have the initiative, and consider the time to be right, and defending archers will therefore seldom get to shoot after every move by their enemies. So, if you, as I do, play games sometimes with a very basic initiative system, and sometimes with a more colourful one, bear in mind that you may have to tweak the effects of shooting up or down accordingly.
If you have a rules set for skirmish games that makes archery too effective, you might be tempted to lower the chances for missiles to hit. Having done this, you might find that lone archers or even small groups of them are too ineffective. What you need is some way of toning down mass shooting without affecting single shots too much. That's what I think.