In the spirit of constructive criticism, I'll elaborate on what I think needs to be done.
First, the battle system randomness needs to be toned down a little bit. Not much, mind you. I can't say I agree with the 40HP idea - it would take all randomness out of the game. This game needs some randomness. If it doesn't, I don't think it would correctly represent the idea of a war, and it would lead to a lack of excitement due to already knowing exactly how the battle is going to end up. However, it currently has too much randomness. I would like to see it where, on occasion, a cavalry unit withstands charges from an enemy cavalry unit of twice the size. It just shouldn't happen this often.
Speaking with numbers, I think it might solve the problem if you changed the HP values to this:
Conscript: 5
Regular: 6
Veteran: 7
Elite: 8
That would make the battles less random, yet still have an intrinsic amount of uncertainty built in, especially on those close battles, like Medieval Infantry vs. Swiss Pikemen on grasslands, unfortified.
Secondly, and this is not something that could possibly be implimented until Civ. 4, having a tech lead should make a much bigger difference. When you look at it in real life, like I said, a unit of A1-Abrams is worth more than twenty four (24) units of stone age warriors. Personally, it should be worth about seventy two (72) units of the same stone age warriors. And this should apply in smaller degrees when considering smaller technology jumps. Technology has always made a huge difference on the battlefield. Hitler lost the Battle of Britain because Britain developed radar. Knights disappeared very fast when pikemen appeared. Widespread use of crossbows made long bows practically obsolete, and led to the downfall of platemail. The fact is, if you can develop a new battle technology before your opponent and successfully implement it, you should dominate your opponent.
However, this exaggerates another problem already in existance in Civ. 3 - that once you're ahead, you stay ahead. This problem has been mentioned elsewhere in these forums. My solution is simple: that your enemy can have your tech's "leaked" to them without much trouble. Think about it. If your army of French Knights goes up against an army of English Longbowmen and gets obliterated, once the battle has been lost, wouldn't you think that you could easily find out how to organize an army of Longbowmen by simply examining their longbows and how they fought? It doesn't take years of research to emulate something that has already been done. Only in the most severe cases of secrets is this so. In fact, the only example of this that I can find is very recent, and that is nuclear power. This was a heavily guarded secret of all the nations who knew it, so alot of nations still don't know how to make nukes.
All of this, though, adds up to four things which needs to be added in order to successfully emulate this in a game of Civ. Firstly, all techs must be separated into two categories - techs which the general populace would know (Type 1), and techs which the general populace wouldn't (Type 2). They would know of iron working and bronze working and map making (shipbuilding) - many common professions were based on these techs. They wouldn't know of literature and philosophy - those were reserved for the upper class. Secondly, if you have a border with another country, if you jave a certain Type 1 tech and that country is currently researching that particular tech, there is a significant chance that your knowledge would diffuse across the border and give them a bonus to researching it - up to three quartes of the tech, depending on how big the leak was. Thirdly, you could reduce the chance and severity of a leak if you highly garrison your border cities with military police, the more, the better. Type 2 techs would have to be researched separately, or stolen. Fourthly and finally, stealing techs needs to be made cheaper, and espionage should be made available long before nationalism. Sun Tzu had spies in 2000 BC for crying out loud. The entire thirteenth chapter of the Art of War dealt with using spies.
The final component of making it harder to maintain a lead deals with revolutions, but that is not something I'm done thinking with, and is much more complex.
All of this boils down to these four things:
1. Battles need to be made a bit less random, but still somewhat uncertain.
2. Having a tech lead should net far greater success on the battlefield.
3. It should be much harder to keep a tech lead.
4. It should be possible to keep a tech lead through use of severe measures, if you were determined enough.
WOW, that was a long post.
about that last part, of course.
