Info: Combat (GL)

[rah]

I always laugh when the settler defends the stack. (and wins)

[SlowThinker]

Sneak attack in hotseat mode

After losing a horsemen in a hotseat PBEM game and providing more tests I can claim

The sneak attack in hotseat mode don't work.
It looks in hotseat games the Civ2 engine takes other humans as AIs in every respect: F3 key invokes an AI diplomacy window, the Demographics window bug works, the sneak attack don't work.

[rah]

That makes sense.
Thanks for the clarification.

[TCO]

chieftan effect still missing...

[SlowThinker]

Can you post result of tests

[TCO]

No. Sister still has my game. Must you ask such obvious questions?

[Marquis de Sodaq]

Nice to see this thread is still active. I've been away for a long time, thanks everybody for keeping this going.

[SlowThinker]

Marquis, you should incorporate all posts to the main text

[Marquis de Sodaq]

Originally posted by SlowThinker
Marquis, you should incorporate all posts to the main text

The Sneak Attack! section of the main text has been edited. Very briefly.

[SlowThinker]

I want to point this thread out: Modifiers for Attack/Defense. It explains the practical significance of Attack factor, Defense factor, Hitpoints and Firepower.

[SlowThinker]

Combat math in practice

I will give an example how to use combat math in practice. (Read Modifiers for Attack/Defense )for details.

An example:
Artillery (10 AF, 1 DF, 2 HP, 2 FP) with vet status attacks a Mech. Inf. (6 AF, 6 DF, 3 HP, 1 FP) fortified on a rivered hills.

A.
First of all you apply modifiers:
Vet status causes the Arty attacks with AF=10*1.5=15
Rivered hills causes the Mech. Inf. defends with DF=6*1.5*(2+0.5)=22.5

B.
Then you add a 1/8 bonus to the defender and 1/8 malus to the attacker (you can omit them if you don't need to be accurate):
Arty attacks is AF=14+7/8
Mech. Inf. defense is DF=22+5/8

C.
Now the stronger stats are enhanced - it is upped by (the absolute value of) the difference between attack and defense. In our example the defense is the stronger value:
DS=DF=22+5/8 + (22+5/8 - 14+7/8) = 22+5/8 + 6+6/8 = 29+3/8
(I 'renamed' Defense Factor to Defense Strength (DS) - it is some new kind of number)

D.
Now results are multiplied by HP*FP:
AS=(14+7/8)*2*2=44+1/2=44.5
DS=(29+3/8)*3*1=58+1/8=58.125

These numbers (AS - Attack Strength and DS - Defense Strength) express a relative strength of units:
58.125 / 44.5 = cca 1.3
so the defending Mech. Inf. is about 1.3 times stronger that the attacking Artillery. In average you need 1.3 Mech. Inf. units to kill the defender. In other words the first Mech. Inf. will be killed and the second one will get a 30% damage.

---

Notes:
If you multiply HP by 2 then you double both DS and AS. The unit is as strong as two original units.

If you add a vet status (AF and DF is multiplied by 1.5) to Musketeers (AF=3, DF=3), then you double both DS and AS against Musketeers (it is different against other units): phase C. of both Attack and Defense will be
AS = DS = 4.5 + (4.5-3) = 6

So an increase of AF, DF has usually stronger effect than an increase of HP*FP. On other side HP*FP affects both AF and DF simultaneously.

---

Notice that we use all the product HP*FP. Separate HP and FP factors are not interesting if we consider strength of units.
But they affect accidental behaviour of units. We can study another quantity - a quotient HP/FP . It represents a 'combat stability' of unit. Unit with a high HP/FP (for example HP=5, FP=1, so HP/FP=5) behaves predictably, but unit with a low HP/FP (for example HP=1, FP=5, so HP/FP=0.2) has very fortuious combat results.

[duke o' york]

You mean that you'll need two artillery to kill the mech infantry don't you?

This is great work! It'd be nice to have a SG Hut Finderesque program to calculate who'll win, but that would be at its most accurate only when both units had full green. Once their health bars were reduced then it'd become far more difficult to calculate accurately.
I'm not sure that this work, admirable though it is, will have the speed that rah and the MP community will want, but it's a great basis for some trials and then once you have fought enough you'll have an idea of which units defeat which in which situations. Unless your opponent has the War Academy, then you cannot tell whether the opposing units will be vet or not, so this adds another doubt, but in MP it'd be safe to make the assumption that everyone will be operating with vet units.

[SlowThinker]

It'd be nice to have a SG Hut Finderesque program to calculate who'll win

These calculations are quite easy IMHO (if you omit point B). No program is needed.
For exact calculations you need Boco's Combat calculator (link in GL),

You mean that you'll need two artillery to kill the mech infantry don't you?

It is very likely, but it is not 100% sure.
Take Boco's calculator, double number of hitpoints for an Artillery (to simulate two units), and you will see the probability.

but that would be at its most accurate only when both units had full green. Once their health bars were reduced then it'd become far more difficult to calculate accurately.

It would not be so hard - you simply exchange HP=1 by HP=0.6 for example.

[La Fayette]

Well done, ST
This looks quite clear and simple enough.
It should help me understand what happened when I lose an unit and get angry at the supposedly cheating AI

[Bloody Monk]

This thread is a treasure. Thanks to all who contributed.

Monk