Once and for all: who is the dominant offensive late ancient age UU?

[MiloMilo]

Legion gets my vote. Equal defense to MWs means even chances when attacked BUT: only in open ground and unfortified. If Legions have a geographical advantage they become clearly superior.

Also, the retreat can actually work against mounted units. If they retreat they don't use that last hit point to attack, so we're effectively talking about (for vets, say) a 3-hp MW against a 4-hp Legion. Which mean even in open ground, even unfortified, the defending Legion has better odds half the time.

Plus, you have to look at other factors. My legions will have a couple of horseman slowly advancing with them to pick off stray MWs. Will the Iroquios drag spearmen along with their MWs to defend them? Then they lose the mobility advantage.

Finally, the upgradeability is a non-factor. When I research chivalry and my Legions are finishing up their various campaigns, I will simply build knight armies fresh. Even in the middle ages, I will rather have 20 knights and 10 legions than 25 knights (or whatever it evens out to taking upgrade costs into account).

I know Aeson is the king of strategy 'round these parts, but I have to disagree here.

[I-iZ-1337]

Mount33d warr33orz r00lz d00dz! S0r33 t00 all j00 naj sajerz but th33 2 m00ves juzt d00 it for m33. Thinkz... th33 r33tr33t m33nz j00 iz gonna l00ze much lez than j00 wi11 l00ze 33ven with mortalz. If j00 iz fightin mortalz j00 can wear them downz, attackin and retr33tin and watchinz them all diez. 33ven if th33 mortalz ar3 on mountainz j00 wi11 sti11 winz mozt oftenz. L33ginz winz on mountainz mozt oftenz so they iz tougherz, but th33y diez s0 oftenz againzt 33ven a cit33 with wallz and a couplez sp33rz. S00 th3yz wi11 l00ze a f33w cominz in t00 attackz and l00ze a f33w attackinz th33 citi33z. T00 top itz all off, they can b33 upgradedz and lazt right upz to the early induztrialz. They iz 1337z!!! But I iz the 1337ezt.

H00p3 j00 takez note of my tipz and l33rn fromz my mad 1337 skillz! I Iz th33 1337ezt.

[King of Rasslin]

I iz m00r l33t th3n u d00d!!!

[I-iZ-1337]

Wa33t ti11 j00 playz MP againzt m33, I will r0xx0rz j00. J00 don't havez the skillz d00dz, j00 iz not jet 1337.

[vulture]

Originally posted by MiloMilo
Legion gets my vote. Equal defense to MWs means even chances when attacked BUT: only in open ground and unfortified. If Legions have a geographical advantage they become clearly superior.

Also, the retreat can actually work against mounted units. If they retreat they don't use that last hit point to attack, so we're effectively talking about (for vets, say) a 3-hp MW against a 4-hp Legion. Which mean even in open ground, even unfortified, the defending Legion has better odds half the time.

For a veteran 3 attack mounted warrior against a veteran legion unfortified on open ground (3.3 defence), the legion wins a whole 27.9% of battles, with the MW winning 40.5% and retreating 31.6% of the time. If the retreat is turned off (or we are dealing with a vet swordsman attack) the the legion wins 44.8% and loses 55.2%.

If the legion is able to counterattack the wounded MW (assume we're dealing with a 1 on 1 fight) in open territory, then those 31.6% retreats become 30.4% wins for the legion and 1.2% for the MW (total 58.3% wins for the legion, 41.7% for the MW). In the real civ3 world, things are more complicated, since both sides will usually have more than one unit involved in the fray.

Plus, you have to look at other factors. My legions will have a couple of horseman slowly advancing with them to pick off stray MWs. Will the Iroquios drag spearmen along with their MWs to defend them? Then they lose the mobility advantage.

MWs do well attacking from defended positions (cities, stacked with defensive units) or when they can use roads to attack and retreat out of range in the same turn (comments about horsemen stacked with legions noted, but the counterattacking horsemen are themselves guaranteed prey in this case since they can't get back to the stack in the same turn) if necessary. In rough terrain their movement advantage is negated and the retreat actively works against them with the whole couterattack problem (by essentially making them fight their last hp with a combat strength of 1 rather than 3 if they manage to retreat).

Surprise, different units are more useful in different situations.

[MiloMilo]

Vulture, how did you come up with those numbers? 3 attack MW against 3.3 defense Legion and the Legion only wins 44.8% of the time (given no retreat)? Makes no sense... Anyway, 2 points:

I count a retreat as a win for the Legion - that's why the horsy retreated in the first place. If you're lucky enough to get it to your city to heal well, you better hope you stay lucky...

And yeah, my Horseman taking out your wounded MW will in all likelihood mean one quickly dead horseman. But that's a trade I'll make. If we trade 1-for-1 Horsemen for MWs all game, guess who comes out on top?

Anyway, yes, of course the poll is kind of silly. Different units for different situations.

[alexman]

Originally posted by vulture
If the retreat is turned off (or we are dealing with a vet swordsman attack) the the legion wins 44.8% and loses 55.2%.

I think your calculator is a bit off. The highest modified strength (the legion's 3.3 in this case) should win more often than it loses, no?

[Arrian]

[SIZE=1] I count a retreat as a win for the Legion - that's why the horsy retreated in the first place. If you're lucky enough to get it to your city to heal well, you better hope you stay lucky...

That's just it. Retreating isn't the same as losing. If a legion loses a battle, it dies. No more legion. If a Mounted Warrior is forced to retreat, you can heal up and fight again (anyone who consistently loses beat up mounted troops exposed in the field doesn't know what they're doing).

Furthermore, if a MW attacks a legion and is forced to retreat, the Legion does not benifit as it would from a victory (killing another unit). By this I mean promotion, or rather the chance of promotion or leader generation.

So I would not count a retreat as a win for the legion.

-Arrian

[vulture]

Originally posted by MiloMilo
Vulture, how did you come up with those numbers? 3 attack MW against 3.3 defense Legion and the Legion only wins 44.8% of the time (given no retreat)? Makes no sense..

Mea culpa. I transposed the numbers - legion wins 55.2%, MW 44.8% if there is no retreat. The other numbers are correct.

[King of Rasslin]

The legion will become more powerful when they (hopefully) add in a middle age advanced swordsman and make commercial a better trait. But the Iroquios deserve the MW because they are stuck with expansionist.

In the actual game, sending horsemen with legions is good. It just isn't fast. MWs have the speed advantage. They are better at crushing AIs. As for multiplayer, MWs will still be better because of retreating.

In fact, most people would choose horsemen over legions. It's all about the mobility. In fact, MWs don't go obsolete quickly, and the golden age comes at a better time than the Romans. Iroquios are a religious civ. An early golden age will hurt the Romans bad.

If they attack early, they lose a good GA. If they attack late, the Iroquios will have enough MWs to kill the attack.

The main point here is that it is very easy to defend and very hard to attack. But when I want to attack, I choose MWs. I like legions more than immortals but not as much as MWs.

[jshelr]

It's interesting that I would rather see an AI civ that I'm attacking have MW first, immortals second, and least of all the legionary. They are the toughest to attack and the AI does not use stacked MW's very effectively in my experience.

But for my own civ, exactly the reverse is true: MW's are the best, immortals second, and legionary third. Against another human player I'd reverse legionary with immortals. But attacking the AI, the 4 rating counts big.

Basically, I like the sound the MW makes as it races across the screen.