With RNG, I have a chance of winning on a longshot, even if I am in a poor position due to my own mistakes. If combat is purely deterministic, I can mathematically calculate whether or not I will be able to stop my opponent, and if I can't, I just admit defeat and start a new game, losing the fun of the gamble.
Oh sure. You claim you can "mathematically calculate" exactly how many troops the opponent will have when arrive, as well as how many and exactly what they are, all ahead of time, and then realize you've lost to the point where it isn't worth trying?
Or are you instead going to complain about knowing the attack will fail the turn you're attacking, but only if you do calculations instead of just fighting and seeing? Which angle are you going with this sillyness? Is there a 3rd one that makes sense that I'm missing? Most players can't see into the fog and know exactly what their opponent will do in terms of production, tech choices, and movement/positioning at all times.
I'm pretty sure I'm proposing deterministic combat, not omniscience. Stop arguing against omniscience when trying to argue against DC.
It's essentially a fixed exchange in which the outcome is known
Instantly wrong, unless your strategy is limited to a single turn. The outcome of a given battle can be known, but the exchange is anything but fixed right up until the fight occurs, and there's absolutely no way to know what your opponent will do.
It's more a matter of relative valuation of resources, in this case units, than risk management.
Oh, you mean like the rest of the core gameplay of civ IV in general?
You keep saying "no risk no risk no risk no risk no risk no risk no risk no risk no risk no risk". It's getting old. Your opponents actions are not known, and victory isn't guaranteed. I don't know what hocked up definition of risk you're trying to use, but it isn't an accurate one.
ou exchange some of your units health for the destruction of another unit. The result is mathematically predetermined. It's a choice of whether you feel it is worth is to exchange some of your unit's hp for the removal of an opponent's unit. Perhaps the fact that your injured unit could be destroyed in an counterattack could be construed as risk, but that is really a known tradeoff, or a result of under-preparation.
Okay. I get it now. You're insisting there is no risk based solely on the possible actions for you and you alone on a single turn. You are extrapolating "no risk on me attacking this unit this turn" as = no risk. That is fundamentally flawed, and seriously so.
There are LOTS of risks to investing in an army, declaring war, moving into enemy territory, or engaging the enemy and taking a lot of damage without complete knowledge of everything about your opponent and what choices he will make. You could discover you can't actually kill him after building the units; that's loss even if you don't attack. You could discover he was hiding fast moving units in the fog and get stack wiped. You might be counter attacked, production-spammed, or simply fail to capture enough cities fast enough to be able to react to something else across teh world. All of these are risks, and yet you have repeatedly asserted they don't exist because you know that pike will kill a knight. It's absurd.
I never said that EVERYTHING should be random, much like you don't assert that EVERYTHING should be deterministic.
Indeed. The question is where is rational to draw the line. There is no fundamental reason that combat needs to be RNG while tech rate is deterministic; that's just what the designers chose. You could achieve similar results (and indeed, similar instances of lower-probability winners) by flipping which is random.
My point is that RNG combat sticks out like a sore thumb precisely because aside from spawn balance (and apparently work has indeed been done to attain spawn balance) it is the only non-player mechanic involving RNG that can't be disabled, and that there is no reason it is necessarily required or even necessarily preferable.
poorly timed equipment malfunctions and unsolicited opportunistic 3rd party interventions constitute chance rather than strategy/tactics in my opinion.
Opinions sometimes can't be wrong, but in this case you definitely are. Equipment doesn't just "malfunction", it malfunctions for a reason. 3rd parties intervene for a reason.
The weather is an interesting case. It is deterministic in the sense that everything weather-related operates based on a cause (most of which are now known), but modeling it to the extent that it is predictable is very difficult. That's why I mentioned it was the "closest" thing to being random that has decided any known war ever.
In theory these events are deterministic, but to simulate such things accurately would require an absurd mountain of code. I'll settle for calling them extraordinary events outside the control of the commander/army, which can be roughly simulated by RNG
The reason these things are impractical to code is
the scale of the game of civilization IV itself. It isn't modeling people's lives, seasons, tactical combat, or someone carving out a wheel. We surely aren't playing a harvest moon minigame every turn to get tile yields, or a minecraft minigame. Civ IV was put on a long historic scale. Everything else from mining to farming to irrigating canals (huge project management there) to building world wonders and complex infrastructe to hiring people to discover technology is all done in a deterministic way, and yet compared to the times weather or individual error has caused inconsistencies in war, differences from expectation in such things has been MASSIVE in our history.
If we accept determinism in crop yields, there is absolutely no logical reason to reject determism in combat based on any "realism" argument. I asked that we cast preference aside until there's actually something workable for comparison and look to theory, but you're falling back an awful lot on preference.
Like I said, actual realism isn't so important to me. However, anomalous outcomes can and do happen in war.
However, they don't in civ IV right now,
with the exception of early game warfare when the cost of one loss in 
is too large a proportion of total
Show me one example of bad RNG screwing someone with 40 rifles/cannons to the point of it being the determining factor in the game. In all the years and all the games posted on this very form, it has never happened even once. However, losing 8 axes to 3 archers (a rather profound difference in
investment) is very possible, and you lose instantly even if that was your only way out. I assert that this breaks from how the game plays mechanically in later wars and from its core experience (strategy + good planning = victory), to no benefit.
RNG combat is functionally deterministic mid-late game, so the primary purpose of this exercise is to "balance" the early game. And yes, prepping a rush early game would STILL be a tremendous risk (get backstabbed and opponent manages to defend and you're still screwed)...but in this case if you actually do outplay your opponent you don't get penalized simply because you did it early rather than late.
I agree - in the game terms , Civ is just a game and there is no point to throw in reality issues but even in game - a better guy can have a bad day once in a while (RL random hehe 
- For example: Like one of the corporals of Napoleon forgot to bring a map to a battlefiled and France lost hehe) I would like to see it implemented (No RNG) for a test run but still I think that rigged-combat games are a lot less interesting than random ones and I guess yeah some people would like that and some dont 
.
), and you have to attack NOW with the best force you can slap together, sometimes leaving cities undefended for a turn or two. Yeah, a you lose a lot and then the game is over, but when you win, it's worth the pain of all those 99% battles you've lost.
