Unit Strengths by Era

[exsanguination]

Originally posted by pcasey
I said it in another thread of the same title, but I'll repeat it here I suppose.

IMHO cavalry are overpowered. You can get them a good 6 techs before nationalism brings riflemen into the game which is a minimum of 24 turns for you to exploit your cavalry vs an enemy which has, at best, musketmen as its defenders.

A six attack, 3 move attacker in an era when the best defender is defense 4 is overpowering.

I'd make them 5/2/2 instead of 6/3/3.

I was browsing this thread and noticed this. I'm not sure whether anyone else has brought this up, so if you have just ignore this.

6 techs? In most of my games its -1-3 techs! How? If I am Scientific, I go into the indust. age BEFORE researchign Military Tradition (which I generally find not worth researching unless you can get it in 4 turns with a huge cash buildup), so I get Nationalism for free, BEFORE MT.

Otherwise, Its all depends on how I play it. I rarely ever research MT, as said above, and usually steal it. The latest I usually get it is after Medicine, which is generally 2 techs. Cavalry is definetely NOT an overpowering unit. Remember, if you use it against your enemy, they will use it against you. I like Cavalry - they are good for border patrol and for defending cities as the attacker.

I think 6/3/3 is perfect.

Also, I do not think Cavalry should upgrade to tanks. Its very realistic - Calvary were used for a long time before tanks. IMHO, I do not think any human/living unit should upgrade to a machine - so I changes it in my mod (infantry does not upgrade to mech inf., etc). I'd like to see the day I'm turned into a Mechanize Infantry!!!

[punkbass2000]

"6 techs? In most of my games its -1-3 techs! How? If I am Scientific, I go into the indust. age BEFORE researchign Military Tradition (which I generally find not worth researching unless you can get it in 4 turns with a huge cash buildup), so I get Nationalism for free, BEFORE MT. "

If you beeline, it's quite possible. Look at it this way:

Beeline for Military Tradition
Engineering
Feudalism
Invention
Gunpowder
Chemistry
Metallurgy
Military Tradition
Total techs = 7
Requisite techs=6

Now, all the requisite techs not in that list:
Monotheism
Theology
Education
Astronomy
Navigation
Physics
Magnetism
Theory of Gravity
Total techs=8
6+8=14
14-7=7
Therefore, if you researched at the same speed, you should be able to research Military Tradition 7 techs before the Industrial Age + Nationalism. If they're Scientific, subtract Monotheism and Nationalism, and you get 6 techs.

[King of Rasslin]

Cavalry shouldn't have less defence than knights. But having 3 movement is a little steep. I think they should be more expensive. That would solve the problem.

I think infantry are much more powerful than cavalry. Their high defence makes artillery more valuable, but I think having 10 defence is a lot. But when you consider the time period between tanks and mech infantry, having 10 defence is perfect for infantry.

I think making marines 10/5 and available earlier would make a difference. Marines also need an attack bonus when making an amphibious assault. The amph. assualt is usually against fortified enemies, so the marines need more power.

The privateer should be have 3 attack. Yes, 2 is good, but not for 60 shields. That is so expensive for a 2/1 unit.

The musketman is too expensive. He should be 10 shields cheaper.

Paratroopers need a serious overhaul.

I think modern armor have way too much defence. They are too powerful compared to mech infantry. They deserve 12 defence, but not 16.

Frigates should be 4/2/5. They require both iron and saltpeter. Because they require 2 resources, they should be stronger. They use the same cannons as ironclads, so their attack should be 4. They don't have all that iron weighing them down, so they should have 5 movement. The Man-O-War should be 4/4/5. It goes obsolete so quickly, this increase is deserved.

And... Thats all I can think of for now.

[Theseus]

I restate my original premise: Immortals are hands down the most powerful unit in the game, considering relative strength and longevity of application.

Using the Civilization Calculator (can;t believe I just found this!!), a vet Immortal has the following likelihood of destroying a reg defender, fortified in a town on grassland or plains:

Warrior: 96%
Spearman: 82%
Pikeman: 65%
Musketman: 51%

So, from the discovery of Iron Working through GUNPOWDER, this bad boy has a better than even shot of winning.

BTW, elite vs. Musketman is 63%, which would probably be the case.

Yes, I agree, retreat makes a difference. The most comparable unit I can think of is the MW. Same approach:

Warrior: 92%
Spearman: 70%
Pikeman: 51%
Musketman: 37%

Even with retreat, the Immortal's longevity is the trump. Mobility? They're industrious. Promotions? Forget militaristic, you when most every early fight.

'Nuff said.

[King of Rasslin]

The calculator is probably wrong. A fortified spearman has 2.5 defence, which rounds up to 3. It is the immortal's 4 attack versus 3 defence. Yes, the immortal has 1 more hp. But I see it as the immortal winning 60% vs. a fortified spearman, not 82%.

As for musketmen, a fortified musketman has 5 defence vs. an immortals 4 attack. Here, the veteran immortal will probably lose because he has such an attack disadvantage.

[Theseus]

The calculator is not wrong. It uses a factorial approach to considering the iterations of combat, which is where experience level shows up. Check it out.

Quote from a thread by Valant:

P(x) = (n over x factorial) * p^x * q^(n-x)

Where
n = number of trials
p = probability of success
q = 1- p
x = number of successes in n trials
n-x = number of failures in n trials

So
n = attackers HP + defenders HP - 1
p = attackers attack power / (attackers attack power + defenders defensive power)
* note do not forget to add in the defenders defensive bonuses it its attack power before plugging it into p*
q = 1 - p
x = is a variable that ranges from a min of the defenders HP value to a max of n
n-x = is a variable that ranges from a min of zero to a max of (attackers HP - 1)

The program then goes through and then determines that out of n trials (max number of tuns in a battle before a unit dies), the attacker will win at least the defenders HP number of turns.

[King of Rasslin]

I guess being a veteran makes a big difference in combat, but I see its value is less in the actual game. There is usually a tendancy for units to get a "streak." Like the spearman taking 1 damage at first, but then getting 4 hits in a row without being hurt. Or the immortal taking 2 damage and then finishing the spearman without damage.

I think it can get a little weird out there, but in a perfect world the formula should do just fine. However, this is Civ 3, not Alpha Centauri

[Theseus]

True.

When I was checking around for a really good approach to combat odds, including the impact of experience, I also found some threads about random number generation, which drives the actual results. Evidently it's a little flawed, which could account for streaks.

I really like that Civ Calculator though. And I stand by Immortals!

[punkbass2000]

Well, there is no such thing as *truly* random number.

[Theseus]

42

[King of Rasslin]

It's because today's computers are too weak and slow to make a perfect random number. In 5 years we can get a real RNG.

[publius]

i have modded my marines to treat all terrain as roads, like alpine troops in civ2. this makes them a little more useful. in the modern age with air support they can take out armored units, and they're much more effective for pillaging, which makes sense. i think it adds an element to modern warfare/post ww2

[Lawrence of Arabia]

I think that the persians on a whole are unbalanced. they always have a big civ and their immortals are just that: immortal
Also, Ironclads should be restriced to coastal waters. That way frigates will have a longer shelf life.

[King of Rasslin]

No, ironclads should just be really slow. Frigates should be faster. Immortals are only invincible to a dumb AI. They wouldn't be that effective in muliplayer because they are slow. Mounted Warriors are much better.

[kring]

Ironclads are not seaworthy. The Monitor sank off the coast of North Carolina. While I realize there was a hurricane issue, the point is still the same. Ironclads, whether of the Monitor or Virginia classes were not designed to be seaworthy; they were designed to be used along coastlines and on rivers and other calmer venues. Their gun ports were close to the water line and not designed to be watertight.