Crossbows came first. They open the way for easy killing power to the common peasant. Suddenly it was easy to train a soldier to kill a knight. Such was the fear of the crossbow by the church and nobles that it was banned many times. The common peasant could stand up and kill a knight in full armor with one shot. Whereas bows required great upper body strength, skill and training to operate effectively the crossbow could be trained in weeks if not days and used by a common peasant to kill a knight.
The trend didn't stop there. Firearms quickly became popular for the same exact reason. It is relatively easy to pick up a gun and use it to kill. If it's so easy to use, what's the value of skill and training? Aside from issues like weapon cleaning and upkeep, there has to be some value when it comes to combat, what is it? This of course is a rhetorical question. It is clear that skill in firearm use is a must. While shooting someone at short range can be easy, doing so at longer ranges can get complicated. The real question is: How to represent this in a realistic way? By realistic I mean a way that means something substantial in the game at the right moment.
This has concerned me ever since I started writing up the combat rules for the game. Linear probability distributions don't work to well with this. Every + 1 is the same probability increase no matter what. It's all 5%. Change to d100, each plus is still the same 1% increase. A - 5 for range can be countered by a + 5 for skill. This + 5 is a good bonus. But someone with a + 2 almost cut that benefit in half. How long did it take to train to have a + 5 and how long did it take to train a + 2? Is it fair? Probably the shooter with the + 2 just trained a few weeks and the one with + 5 months of not years. Now lets consider point blank range, + 0 modifier. Can we honestly say the expert shooter with the + 5 bonus is safer than the one with the + 2? Are the odds of missing at point blank so off that those bonuses would actually matter? Honestly I don't think so. They're both in a load of trouble at point blank range.
So to make skill matter what I'm using now is a bell shaped distribution. One in which the end point values add little benefit to the odds for each skill bonus. Range and weapon quality put you some point along the line. Point blank places you at the top most values where each + 1 from training counts very little towards increasing your odds. Facing off an untrained shooter will not be that much easier than a trained one. There is a very low probability of missing and you're in a load of trouble no matter how good you are.
Things become interesting when you begin to put some distance between you and your target. The range places you on the part of the curve with greater slope. Now skill begins to matter. Each + 1 means a much bigger increase in the odds of hitting than at point blank range. Suddenly the + 5 means a whole lot more than + 2 and the better shooter has higher odds of hitting its target, by a long shot, than the lesser trained one.
Finally there is long and extreme ranges. Targets so far that shots are impossible unless specialized equipment (rifles and scopes) are used to take the shot. Unskilled users will have practically no possibility of scoring a hit. Skilled users, those trained in the use of the rifle and scope, will be able to benefit from the bonuses these pieces of equipment give. Yet these bonuses are just enough to put the shot at the beginning of the distribution curve. Just as it begins to slope. Good skills will count, but not enough to ensure a high hit rate. Excellent skills, those acquired after months of training and years of experience are needed to get those high odds to hit required for a one shot one kill.