Be the attacker!

[carn]

In old civs units had a different strength score for attacking and defending, so it was always important to manage to choose, whether one is attacking or defending with a specific unit.

I thought first, that that changed with civ4, so its no longer important whether one attacks or defends, except for getting specific effects, like terrain bonus, fortify bonus or collateral damage.

But inherently built in is a small bonus for the attack value of most units, since only combat promotions modify an attacker's str value, while a defender's str value can be modified by additional bonuses and bonuses are added and not multiplied. Therefore if those additional bonuses are identical whether a unit is attacker or defender, the benefit of combat promotions is smaller for the defender as they are added on other bonueses.



E.g. Maceman with combat 1 against Praetorian with no promotions:

Praetorian attacker -> Preatorian str 8 and Maceman str 8 (5*(1+0.5+0.1)) -> 50% chance for maceman

Maceman attacker -> Maceman str is 5.5(5*(1+0.1)) and Praetorian str is 5.33(8/(1+abs(-0.5)))
-> 64% chance for maceman

In the example the effect is of course very big due to jumppoint at equal str, but the effect is also there away from jumppoint:

E.g. Maceman with combat 1 against Praetorian with combat 1:

Praetorian attacker -> Preatorian str 8.8(8*(1 + 0.1)) and Maceman str 8 (5*(1+0.5+0.1)) -> 32.0% chance for maceman

Maceman attacker -> Maceman str is 5.5(5*(1+0.1)) and Praetorian str is 5.71(8/(1+abs(-0.5+0.1)))
-> 35.4% chance for maceman

E.g. Maceman without promotions against Praetorian with combat 1:

Praetorian attacker -> Preatorian str 8.8(8*(1 + 0.1)) and Maceman str 7.5 (5*(1+0.5)) -> 28.4% chance for maceman

Maceman attacker -> Maceman str is 5(5*(1)) and Praetorian str is 5.71(8/(1+abs(-0.5+0.1)))
-> 29.9% chance for maceman


This effeect is always present as soon as at least one side has combat promotions and other bonuses apply, but no bonuses for any defending side(e.g. fortification, terrain) is there.

So as a rule of thump, be the attacker, unless one side can get defense only bonuses.



I do not know whether this effect is intended, but it should be easy to mod or patch away by changing combat promotions from general bonus to bonus against all unit types. It would still always help but would be calculated into the bonus sum and therefore be equally important for attacker and defender.

Carn

 

[s.c.dude]

E.g. Maceman with combat 1 against Praetorian with no promotions:

Praetorian attacker -> Preatorian str 8 and Maceman str 8 (5*(1+0.5+0.1)) -> 50% chance for maceman

Maceman attacker -> Maceman str is 5.5(5*(1+0.1)) and Praetorian str is 5.33(8/(1+abs(-0.5)))
-> 64% chance for maceman

In the example the effect is of course very big due to jumppoint at equal str, but the effect is also there away from jumppoint:

E.g. Maceman with combat 1 against Praetorian with combat 1:

Praetorian attacker -> Preatorian str 8.8(8*(1 + 0.1)) and Maceman str 8 (5*(1+0.5+0.1)) -> 32.0% chance for maceman

Maceman attacker -> Maceman str is 5.5(5*(1+0.1)) and Praetorian str is 5.71(8/(1+abs(-0.5+0.1)))
-> 35.4% chance for maceman

E.g. Maceman without promotions against Praetorian with combat 1:

Praetorian attacker -> Preatorian str 8.8(8*(1 + 0.1)) and Maceman str 7.5 (5*(1+0.5)) -> 28.4% chance for maceman

Maceman attacker -> Maceman str is 5(5*(1)) and Praetorian str is 5.71(8/(1+abs(-0.5+0.1)))
-> 29.9% chance for maceman
yeah but macemen come much later in the game than praetoians try again this time with swordsman an axes

 

[DaveMcW]

Just replace "maceman" with "axeman" and read the post again.

 

[Paul666]

I must disagree with this analysis because of the assumptions used. There are almost always hill, woods, or jungle very near your units. Effective use of terrain (I don't spel gud) is much more important than small calculation changes. Also, you must factor in where you stand--are you defending your land?; are you trying to take cities?; are you trying to pillage valuable resources, or even save them? These factors (in my not very humble opinion) must be considered.

 

[greenpeace]

I think that the defenders position/fortify/promotions will usually makes up for the attacking bonus. However, if this is an attack on open terrain then yes, you should attack. I think it would be more benficial to do an analysis on city attacking because that is the most common attack. However, I have to say thank you so much for spending your time trying to help us!

 

[Draconian]

E.g. Maceman with combat 1 against Praetorian with no promotions:

Praetorian attacker -> Preatorian str 8 and Maceman str 8 (5*(1+0.5+0.1)) -> 50% chance for maceman

Maceman attacker -> Maceman str is 5.5(5*(1+0.1)) and Praetorian str is 5.33(8/(1+abs(-0.5)))
-> 64% chance for maceman

I assume you are talking about Praetorian vs Axeman not Maceman there, otherwise those numbers don't fit.
And afaik (unless this was changed in 1.61) those +x% vs unit-type bonuses are never added to the strength of the unit with the bonus, they are allways substracted from the strength of the other unit.

Axeman attacks Praetorian:
5 vs 8*(1/(1+0,5)) = 5 vs 5,33
Praetorian attacks Axeman:
8*(1/(1+0,5)) vs 5 = 5,33 vs 5

So it doesn't matter if you defend or attack.

 

[Zombie69]

And afaik (unless this was changed in 1.61) those +x% vs unit-type bonuses are never added to the strength of the unit with the bonus, they are allways substracted from the strength of the other unit.

No, they're always added to the defending unit, no matter if it's the defender or the attacker that has them (except for combat that is always applied to the unit that has it). It's always been this way, ever since the game was released.

 

[Draconian]

I just tested it and you're right. Looks like my memory is playing tricks on me.

 

[malekithe]

I just tested it and you're right. Looks like my memory is playing tricks on me.

Try this thread, http://forums.civfanatics.com/showthread.php?t=137615.

It'll save you many future experiments and incorrect assertions.

 

[Calavente]

That's fun ...
if there is no combat modifier form the terrain and no +x% from fortification, why would one be the defending unit?
Even if you had not this tweak of math that favors the attacker (I thank you for that, I hadn't noticied it myself).
'coz you know: if you attack, the experience your unit gains is higher or equal to the one you may gain by defending.
so even with equals combat chances in attack and defense, it would be best to attack.

For me one defends only if : leaves a city under-protected, attacking let the ennemy pillage one's ressource, unit is trying to heal, unit is at 3rd or more turn of fortification, attack and winning means going for a poor defensive tile in range from a 2nd ennemy unit, unit has strong defensive promotions (city defense/ defense), ennemy has strong defense promotions and on defensiv terrain.

So : in defensive position (terrain or strategic): defend; else: attack


EDIT : oh! beware of rivers... defending means +25 for you, attacking means -25 for you : swapping from defense to attack means roughly a -50% bonus for you math people may give you the real impact but not me