Combat Odds oddity

[tokala]

Either I have forgotten how the CIV combat mechanics are working, or something is wrong with the calculations if you attack a hill giant (See attached combat log).

First combat is hill giant attacking warrior fortified in city with 23.5% odds. If the warrior gets hit, it takes 17 points of damange, the hill giant takes 23 points of damages. Sounds about right.

Second combat warrior attacks the damaged hill giant (2.8/3.9) with 35.8% odds. Warrior is taking 26 points of damage each turn. With better odds than the first combat, shouldn't it be rather less damage than 23, not more?

Even weirder was a combat of a scout against a badly damaged hill giant with 82% odds, the scout losing without inflicting any damage, getting hit for 29 points each combat round.

And from my playing experience, it always felt extraordinarily hard to inflict any damage to defending hill giants. Do those have any hidden defense bonus?

 

[odalrick]

The hit strength isn't based on combat odds, it's the other way around.

First combat both are healthy, which roughly means combat odds depends solely on hit strength.

Second combat, hill giant is damaged, which means not only does it hit weaker, it can take less damage too. It hits for 26 damage so it must hit four times to win, but is has fewer chances to hit than in the previous combat, because it takes more damage per hit and has less HP.

Scout vs hill giant: A badly damaged hill giant will be defeated if hit once, so it must hit three times in a row. Since your Scout had 82% chance to win, we can deduce that the chance to win three times in a row for it was 18%.

Or maybe it could survive one hit, the point is that hit strength is only part of what determines combat odds and everything else is only equal when both combatants have the same HP. Ignoring first strikes, withdrawal and anything else.

 

[Calavente]

oldarick :
what seems strange here is that :
as the hill giant gets less strength, he hits harder.
normally he should :
-hit less often,
-get hit more often,
-give less damage per hit,
-take more damage per hit.

Here, it seems that two things are broken :
-the wounded hill giant is hit less (0/4 instead of 3/8)
-the wounded hill giant, with less strength, hits a lot harder than with full strength. (EDIT : 23%/hit, then 26%/hit, then 29%/hit !!)

 

[odalrick]

Perfectly ordinary, the defending warrior had enough bonus to make him 6.45 vs hill giant 5.
Then the giant is wounded, but the warrior has no attack bonuses for 3 (warrior) vs 3.92 (giant).

He should hit comparatively weaker units harder.


In the scout example, the giant was injured but has such a high base strength and maybe some defence bonuses as well that it still had high strength compared to the scouts 2. The scout just had so much more health that he had a good chance to hit that one time he needed to win.

 

[Calavente]

ok.
I missed that different strength values..
you are right.

 

[tokala]

Okay, so the high base strength of a hill giant is somehow averaged with the actual strength in the wounded state for the purpose of calculating damage per hit. A defending wounded hill giant will hit rarely, but if it hits, it hits hard. One the other hand, it will take only one or two hits to finish it of, but if it survives the combat, it will have a comparatively high chance of not sustaining further damage. Right?

 

[odalrick]

I don't think strength is averaged to calculate damage per hit. It's just (HP/100)*base_strength compared to the same for the enemy. A wounded hill giant will hit as often as any other unit with the same adjusted strength. However, since it is damaged it will be able suffer less damage, and thus lose more often than an undamaged unit with the same adjusted strength.

And, given that the unit survives the combat, a damaged unit will generally have suffered less damage, yes. For no other reason than that it has less health to lose. A one HP unit will never suffer any damage from a combat it survived.

 

[tokala]

I don't think strength is averaged to calculate damage per hit. It's just (HP/100)*base_strength compared to the same for the enemy. A wounded hill giant will hit as often as any other unit with the same adjusted strength. However, since it is damaged it will be able suffer less damage, and thus lose more often than an undamaged unit with the same adjusted strength.

After a bit of experimentation, it seems that there is some weighted strength being used for the calculation of damage per hit. The effect seems to be that combat odds of a damaged unit with remaining strength X versus an undamaged unit with base strength X is kept pretty close to 50%. A 4 Def wounded hill giant definately hits harder and takes less damage per hit than a 4 Str axeman. But in any case you convinced me that the computer is probably not cheating on me with the displayed combat odds

 

[dunedainjedi]

Personally I've always felt the combat odds to be mostly accurate. Yeah, occasionally you lose a 90%+ match and occasionally that summoned skeleton actually kills something despite the less than 10% odds, but it pretty much evens out in the long run.

Did you lose with 82% odds? It happens around 18 out of 100 times. Other than the fact that a hero with 9.99% chance of victory loses half of the time, the game seems pretty good with the combat odds. Anyway that's my two cents.

 

[Calavente]

Tokala :
the different strength we spoke about was :
1st combat was Giant (5) vs warrior (6.45) giant deals 23% damage per hit
then second combat : wounded giant (3.92) vs warrior (3) : (Giant:strength)/(warrior) advantages more the giant than in the previous combat : he deals more damages per hit.
On the other hand, he is wounded : he can only take 3 hits before dying (instead of 5before): thus the odds or better for the second warrior than the first warrior.

But IIRC, it is possible that a 8strength unit wounded to 5 hits harder than a 5 unit. There were some threads on it in the vanilla forums.. I read them 3 years ago but forgot most of it