Does terrain affect attack strength?

[geofelt]

In some posts, it was suggested that the terrain from which an attack is launched may affect the attacker's strength. As far as I know, no documentation or strategy guides suggest that this is so. Yet, in a current game, barbarian archer on a mountain attacked a city protected by the great wall and some fortified phalanx. The archer won, then won again, even in the yellow. Even at raging hordes which is supposed to increase attack strength to 150%, I don't think an archer should have won. If the barbs had a vet archer(4), increased another 50% to 6, it should have lost to a fortified phalanx behind walls(7).
Because of this, I suspect that there may be something there. If so, my playing strategy could change. Has anybody done some experiments on this?

[Krushala]

I have noticed this a couple of times as well. But it may be just a coincidence. As the battles are random and sometimes my units perform exceptionally well. But I think I'll do an experiment on this.

A few times it seemed like my archers on the mountains were having better success. But after trying this about 10 times. It really appears to be random. I tried with archers on a mountain and archers on a plains river square on the same city.
[This message has been edited by Krushala (edited December 23, 1999).]

[Smash]

I would have to say yes.Definitely.I've fought enough wars to notice and I will always attack from forests, hills, mountains etc(when possible).I would be quite surprised if someone could show otherwise.

I don't have much faith in the Archer for offensive wars in this game.I've seen them beat legions and lose to horsemen.Considering the important role the longbow played they should have been given more punch.

Same type of inconsistant results when the archer is defending.

[Aurelius]

Good question, Geo. I'll often build a fortress next to a city before I attack it. But in this case, I haven't noticed much advantage in attack or more importantly, defense. Of course, the fortress defends the stack and keeps the democracy intact if I take the city

I've often wondered if units attack better from mountains or whatever due to added defense.

Aurelius.

[Xin Yu]

I MP I sometimes used a warrior to attack a horse on a hill and won. I guess attackers were given a better chance to win than before.

[Genghis Al]

I have noticed that dragoons attacking down a mountian tend to win. But it still seems pretty random.

------------------
Ah man! Ghandi! New game.

-Genghis Al

[Theben]

Well, I don't think it's true, but there is another phenomenon I've noticed. It seems that results are somewhat pre-determined. I see this most often with triremes. Say you've saved the game and move to attack. Each time you try you lose. Eventually you move to a DIFFERENT square to conduct the attack from. This time you win, and if you load and attack again from that square the odds are good that you'll win. I think that is what happened in those "mountain attacks". The most obvious way I've noticed this is when I make a mistake moving my trireme and cannot get back to land. It seems that if I try to get back to land I'm guaranteed of losing my ship (before seafaring). So I always head further out to sea. Not perfect, but I'd say that I survive 90% of the time.

[VetLegion]

Hehe. You just remembered me of myself when I was playing Civ1. I hadn't known anything about 50/50 chance of getting a trireme lost at sea. So I thought that some areas of ocean were "safer" then others. If my empty trireme made it to a continent 4,5,6 tiles away, I used to send loaded triremes exactly the same way. A few did get lost, but I had a lot of succes with this tactic.

Or is it that the luck rides with the brave (ignorant)?

[Aurelius]

Sounds like you both, Theban and VetLegion, need Lighthouse

Going back to units attacking from mountains or hills or defensively 'enhanced' terrains: it's been awhile since I played some of those hexagonal mapped wargames--like blitzkreig or Africa Corp or Battle of the Bulge, but if memory serves me, units lose their 'enhanced' bonuses when attacking. I've also assumed, possibly wrongly, that units don't get their defensive bonus' unless they are fortified (well, I just have been operating with this assumption--not convinced.)

Another assumption, is that fortified units regenerate faster after a battle. This I'm more certain of.

I think it's coincidental that a unit from the mountain does so well. I would hate to attack it though. Would love to hear the definitive answer! Someone should provide us with a percentage breakdown of the defensive conditions as they provided percentage breakdown concerning happy factor as it pertains to critical city limit under certain governments ...Oww!

Aurelius

[VetLegion]

I'll post this on behalf of my friend who doesnt have a modem (but will soon).

This is what he did. Before a year or two he passed from civ1 to civ2 (I stayed with Civ1 as I had 386 then ). As we both weren't sure how this new combat was being calculated, he did a test.

He calculated odds on a paper, and then attacked. Wrote down result. Then he loaded, and did it again....

He did this a number of times ( I laughed at him when I heard ) as he tried to find a good formula to describe combat. Anyway, when I bought a modem, I read about how combat is done and he compared it to his "experiments". He found that they don't match! He is a good matematician.

What do you think? Perhaps it is time for another test? If anyone does it please post it so we all know.

[Theben]

There are a couple of threads buried in this forum that include tests of various things; suffice to say what Microprose said does not always mesh with reality (FE, x2 vs. horse is really x1.5, and only works vs. 1hp units).

[VetLegion]

What?? Really??
Where are those threads??

[Theben]

VetLegion,
I'm not sure. They're way back there somewhere, they may even be archived. I don't remember what they were called. The gist of what I remember is this:

The x2 horse described above;
x2 vs. air is really x3 vs. planes/x5 vs. missiles;
The rules say, FE, "x2" and "+100%" interchangeably, but in reality everything is multiplied one after the other. So

2 DEF * 1.5 (fortify) * 2 (on hill) * 1.5 (veteran)= 9 DEF, you don't add the bonuses together 1st and then multiply. Does anyone else remember any other tests?

------------------
Theben
Co-Moderator of the Civ3 Forums

[wonderdog4]

Yes. In my opinion, definitely yes. I haven't carried out any experiments to prove it yet, but there is no doubt in my mind anymore.

The terrain from which you are attacking definitely has an impact on the outcome.

Knowing this, has helped my early game tremendously -- especially with raging hordes. I will back up back up back up until the attcking unit is forced to attack from plains.

It isn't foolproof, but this is a game of percentages and even one warrior surviving a horse attack early can be the difference between finding a free city on not finding one.

[Aurelius]

Time to do some curve fitting with all your data.

I'll say this, my elephants are always getting killed in the mountains--they don't even get to attack (which is the discussion.)

To answer the discussion question, you have to consider the defensive nature of both units as well. And then you have to figure out the little idyiosyncrasies the makers put into the formula.

In the end, you you'd have to make a chart of every unit attacking every unit in all it's potential defensive situations. DAUNTING mathematics!

But all things being equal, I like to guestimate too.

Aurelius.

[Flavor Dave]

Here's what I don't get--why couldn't Microprose just make the game mesh with the manual?

It's nice to know now about the lesser power of the pikeman. I don't know how many times I've expected a fortified pikeman in a forest or on a river defeat an elephant (should happen, at X2, even if the elephant is a vet.)

Also, we had a big discussion long ago, and the consensus from guys who tested it out was that bonuses aren't cumulative.

A rifleman, fortified on a river, and a vet., should be

4 + 2 (fortified) +2 (River, forest, jungle, swamp) +2 (vet) =10.

NOT

4*1.5*1.5*2=18. Think about it--in your experience, how would this hypothetical rifleman do against cavalry? In the first example, the rifleman has a slight edge against non-vet, slight disadvantage against non-vet. And that's what I see.

In what theben says, the rifleman would almost never lose even against a vet, and would stay at least "yellow" against a non-vet. That's not what I see.

[Seeker]

I thought the fortified bonus was only 25%, or +1 to a rifleman.

[Theben]

Flav Dave,
There's an error in your calculations. It should be 4 * 1.5 * 1.5 * 1.5 (not 2)= 13.5, not 18. That's a big difference. A vet attacking cavalry has a 12, which is close to the defender. So mostly it'll lose, but occasionally it'll win. Which fits in with my experience. Non-vet cavs almost always lose which also fits the formula.

[Flavor Dave]

whoops. I'm tempted to delete my post, destroy the evidence. I know what a vet does, don't know why I made this mistake. Got it right the first one.

Still, I stand by my point, that in the old thread, the consensus was that the hypothetical rifleman would have defense of 10.

I saw the other, bumped thread, which indicates the opposite. I'm not sold.

[geofelt]

I think civ2 uses the same approach for all augmented calculations. shield production, trade, and also defensive strength. The confusion is that the vet status is intrinsic to the unit. A vet phalanx is 3, a vet rifleman is 6. You get this by multiplying by 1.5, or adding 50%--either works the same. After that, the bonuses for terrain, fortification, and walls are additive, based on the base value either vet or non-vet. I suspect that the AI may use the same calculation on attack, adding 200% when attacking from a mountain. I do hope someone can test this.

[Sieve Too]

I did some experiments on defensive bonuses way back in the last millenium . I gave each unit 100 hps to reduce random number bias. Bottom line: defensive bonuses multiply together.

Although the original post seems to have vanished, a summation can be found here:

http://apolyton.net/forums/Forum1/HTML/000629.html

All my tests were done by varying the defender and his terrain. Attackers were always on grass so I don't have any data on whether attacking terrain makes any difference.