Combat is too random

irie-mon

Chieftain
Joined
Dec 7, 2005
Messages
14
Location
Vancouver BC
It seems like combat is more random than the "Odds" display would have you believe. The game has a tendency to let the underdog rout their opponent who should have a decisive advantage. This is especially apparent in the early game. I've lost battles where I had a 6 vs. 3 odds advantage where the enemy took no damage. I've been experiencing this a lot in my latest game, especially with Axemen vs. other melee units. Anyone else noticed this?
 
Combat is strictly based on the stated odds and the current HP of the units involved (and certain other promotions, e.g first strikes, but mainly strength and HP). A unit at 5/10 strength is at a huge disadvantage against a unit at 5/5 - but the displayed odds don't reflect that. If this isn't understood, it's easy for results to occur that seem highly unlikely but are really exactly as expected.

The basic upshot is that the combat is really a lot less 'random' than the displayed odds suggest (strength 4 attacking strength 6 has odds much worse than 40%), but there are other (nonrandom) factors that greatly affect the outcome as well.
 
One cannot judge the randomness of the game... not without testing at least 100 times, listing the mean and variance, and show that the Random Number Generator is indeed generating uniform random number with the correct mean. blah blah blah....

U said that u lost battles where u had a 2 to 1 odds, how many battles that u have lost exactly? And how many that u have won exactly ?

Use the world editor and put a stack of 20 Axemen and a stack of [whatever u like], then carry out the experiment and see for urself the 'randomness'.

An easy scientific approach , better than guessing that if the RNG is broken..
 
it's either that or I'm extremely unlucky and just happened to get the incredibly rare scenario that my players would get slaughtered without doing any damage to the disadvantaged enemy, as well as having 4 losses in a row even when I had the distinct advantage. all of these units were at full health before starting battle.
 
I have found that a small advantage, does not mean a win. Small is less than a 30% positive prediction. 8 to 5....anything less is a crap shot. I have had level 5 swordsmen with cityraiderIII loose to simple archers defending a city that had an extra first strike. Such is battle.
 
I feel that it is almost the opposite of what the first post states. My riflemen fought a lot of knights last night at about 14 to 10 strength. I had not one loss in about 30 fights as far as I can rembember.
 
joasoze said:
I feel that it is almost the opposite of what the first post states. My riflemen fought a lot of knights last night at about 14 to 10 strength. I had not one loss in about 30 fights as far as I can rembember.

Riflemen get a 25% ( or is it 50? ) bonus against mounted units. So your 14 vs 10 odds are actually 14 to 7.5. There is a detailed thread out there that you should read about the odds of combat. I forget where its at.

Anytime a unit gets an inate bonus against other units, the bonus is applied against the other unit. So a str 10 knight becomes 7.5 vs riflemen
 
Bezurn said:
Riflemen get a 25% ( or is it 50? ) bonus against mounted units. So your 14 vs 10 odds are actually 14 to 7.5. There is a detailed thread out there that you should read about the odds of combat. I forget where its at.

Anytime a unit gets an inate bonus against other units, the bonus is applied against the other unit. So a str 10 knight becomes 7.5 vs riflemen

Yes, but they also usually get a defensive bonus so that the odds stated before battle is something like 14 to 10. It varies a bit with promotions, terrain +++, but still its hardly too random when the strongest almost always wins.

Edit: In other words, I am the one attacking them and thereby give them defensive bonuses
 
Those aren't the actual odds. They only compare strength ratings, without considering other factors - especially HP.
 
If a unit is attacking without being at full hit points, (fully healed) they do suffer greatly, and have a large chance of being defeated since they are not fighting at full strength. The odds displayed due take into account all the bonuses that each unit has, but DOES not factor in the health of the units, or as you say HP.
 
Why do you say that the odds displayed in the game doesn't factor in the health of the unit?

AFAIK HP's are not a separate value from the unit's strength. A unit has just the one Strength value, combining Strength and HP's - and it seems to always be shown correctly in the displayed odds in my game, severely cutting the advantage if one unit is wounded.
 
Combat does have a random element. Theres a dice roll, then the units strenght is added along with all the other combat modifiers and penlties and stuff.

So even if you have a advantage theres still that chance you may lose.
 
Beamup said:
Combat is strictly based on the stated odds and the current HP of the units involved (and certain other promotions, e.g first strikes, but mainly strength and HP). A unit at 5/10 strength is at a huge disadvantage against a unit at 5/5 - but the displayed odds don't reflect that. If this isn't understood, it's easy for results to occur that seem highly unlikely but are really exactly as expected.
A 5/10 unit vs. a 5/5 unit has a 50% chance of winning. I think that saying the 5/10 unit "is at a huge disadvantage" is misleading.

Beamup said:
Those aren't the actual odds. They only compare strength ratings, without considering other factors - especially HP.
They consider everything. It shows you all of the relavant bonus of both the attacker and defender, and it gives you the total effective strength of each.


In my experience the odds are consistent. Sometimes weaker units can win. That's just how probability works. 80% change of winning means exactly that. Don't make the mistake of thinking it means you will definitely win. 1/5 times you will lose. But don't take anecdotal evidence, do some careful trials yourself and take notes as you do it. I think you'll find that the combat system is just as it claims to be.
 
karadoc said:
A 5/10 unit vs. a 5/5 unit has a 50% chance of winning. I think that saying the 5/10 unit "is at a huge disadvantage" is misleading.
No, they don't. The 5/10 unit will have a 50% chance of winning each round of combat, and will do the same 20 points of damage per hit the 5/5 will. BUT it will be destroyed by only 3 hits, while the 5/5 will require 5. This means the 5/10 only has a 23% chance of winning. For a more dramatic example, consider a 4/10 vs. a 4/4. The 4/10 has only a 10% chance of victory despite having identical strength.

karadoc said:
They consider everything. It shows you all of the relavant bonus of both the attacker and defender, and it gives you the total effective strength of each.
And if the total effective strengths were the only factor, you would be correct. They aren't. If you want all the details, see Arathorn's Combat Explained thread.

karadoc said:
In my experience the odds are consistent. Sometimes weaker units can win. That's just how probability works. 80% change of winning means exactly that. Don't make the mistake of thinking it means you will definitely win. 1/5 times you will lose. But don't take anecdotal evidence, do some careful trials yourself and take notes as you do it. I think you'll find that the combat system is just as it claims to be.
I find it amusing that you advise this when you apparently haven't done it yourself. The effects of losing HP (beyond simple strength reduction) are quite strong, and easily observable.

Keep in mind that first strikes are ALSO not shown in the odds display. And can make a dramatic difference in the outcome.
 
I find it amusing that you advise this when you apparently haven't done it yourself. The effects of losing HP (beyond simple strength reduction) are quite strong, and easily observable.
I must admit that in my own experience I didn't pay a lot of attention to damaged units fighting against undamaged units. But in all seriousness, I'm not convinced that you are right. I haven't looked at the thread yet, but what you are saying seems like a silly way to do things. I'll look into it.

Keep in mind that first strikes are ALSO not shown in the odds display. And can make a dramatic difference in the outcome.
Are you sure first strike isn't taken into account? Civ4 isn't on this computer, so I can't check right now; but my memory tells me that it is _definately_ listed, and _probably_ taken into account in the numbers.
 
First strikes are listed but don't affect the numbers, because the numbers displayed are the modified STRENGTHS of the units, which first strikes don't affect, NOT the odds of winning the combat.
 
Back
Top Bottom