Military Advisor relative strength assessment definition

[Qitai]

Anyone knows the exact formula for this? The one I am refering to is in F3 when the advisor says
"Compared to XXX, we are strong"
"Compare to YYY, we are average"
"Compare to ZZZ, we are weak"

Just curious as to the exact formulation.

Thanks

 

[Yndy]

All that I know is that it was originally based on the sheer number of units (workers included?) but later Firaxis changed it in a patch so that strenght of the unit is considered (not clear if only attack and defence or also movement points).

I noticed that strenght rating changes most often following a mass upgrade, which is normal given that you get a lot of extra attack/defence.

 

[The Last Conformist]

Originally, it was dumb unit count. Then they excluded workers and settlers (and scouts/explorers?). Finally, they moved to a system where units are weighted to their combat strengths.

I don't actually know how this weighting is done, but I read somewhere that it simply adds up the defensive values of all your units and compares to the AI's defensive points total. If this is true, it'll have some weird effects - Pikemen would be worth the same as Cavalry in the assassment!

 

[lz14]

I think they weigh by total units' shields now. So Sipahi is as good as tank.
If speed is 1.5, count the ADM would be a good system. That's how I weigh things.

 

[Oystein]

I did some testing on vanilla v1.29f

Compering armies of only warriors
24 vs 19 strong 24/19 = 1.26
25 vs 20 average 25/20 = 1.25
16 vs 20 average 16/20 = 0.80
15 vs 19 week 15/19 = 0.79

I conclude, if the strongest army is more then 25% stronger, it is strong, the enemy week. Else they are both average.

If I have not made any mistakethe strength of a unit is not messured by shield or weighted average of ADM. Maybe a combination.

It does not look like catapults and ICBMs counts.

 

[Qitai]

Thanks for the test. So, we know now that 25% is the cutoff. I assume you test with single type of troop so that is the quantity part. Can you test the quality part?

 

[Oystein]

My best guess so far is 2*A+D, movement and shield cost are ignored. It is not too far away in the limited number of tests I have made.

 

[Oystein]

I have done some more testing.

I have mostly tested how many catapults I need to have a strong/average army compred to a single regular attack/defence unit.

The unit stregth is (3A+2D)HP+B
A - attack strength
D - defence strength
HP - helth point
B - bombard

Then, the total armystrength is sum of unitstrengt + 4
The +4 is probably an easy way to get rid of divison by 0 for civs without any units.

My superspeed chariot 1.1.100 was no better than a warrior.
I have just test HP a few times, needs more.
I have not tested bombard rate/range, combination of bombardstregth and attack/defence.
4 catapults is just sligtly better than 1 regular warrior.

 

[anarres]

Wow, good stuff! Please keep it coming Oystein!

I am amazed to see that it doesn't just take the easy option: counting shields of military units...

 

[Qitai]

Great Research Thanks.

This will help me greatly to assess my military compared with AI before IA.

 

[blobglob]

Wow, so the Military Advisor is more clever than we thought. This is some great research - I second anarres. Hm, in fact, I second anarres in everything he posted. ^^'

I just love this Civ community...always coming up with more and more interesting questions you would love to know the answer to and then researching and discussing, coming up with more and more interesting stuff that my knowledge-hungry head can eat up. ;D

Strategic and tactical knowledge has to be one of the most satisfying things possible for me. (Y'what? Just a game?! ;p) Thanks everyone.

 

[SolarFlare]

Originally posted by The Last Conformist
Then they excluded workers and settlers (and scouts/explorers?).

Someone should test to see if explorers count. I certainly use them in military operations.

 

[wilbill]

Great work, Oystein! Thanks for the info.

 

[Oystein]

Originally posted by SolarFlare

Someone should test to see if explorers count. I certainly use them in military operations.

Scouts, explorers, tactical nukes, ICBM, leader, army(without units) have no effect. At least my catapult was strong compared 25 of them. It is of course no reason to fear a civ with 25 ICBM when we can defend our self with a catapult.

 

[Arathorn]

Great stuff, Oystein. Truly marvelous.

Maybe you could add some good insight into the AI unit build tendencies? There's a thread ( http://forums.civfanatics.com/showthread.php?s=&threadid=62138 ) trying to look at that issue, but I'm not sure how intelligent our testing is....

Finding the (3A+2B)HP+B factor is very cool, especially if it carries over to build priority...

Keep the numbers flowing!
Arathorn

 

[Pook]

C3C 1.15, Emperor level, normal land distro. I found the Celts on a well-watered mid-size island, late in my Industrial Age. They had built six cities, then went no further, were at least a full age behind in tech, and were still in despotism (their preferred government). Going for a domination victory, and feeling compassion for their state, I chose to bring the glories of Sumerian civilization to their land. I ended up refighting the Boer War, with my infantry, artillery, and cavalry taking on their hordes.

I have a question concerning my Military advisor advice- Corporal Punishment kept telling me that the Celtic forces outnumbered me and that my military was weak compared to theirs. I've read this whole thread including Oystein's formula, and I certainly appreciate the effort he did in putting this together, but something is just not adding up. I had 26 cities filled with riflemen and infantry, they had 6 cities with pikemen, and I'm weak compared to them? This experience makes me wonder if the Advisor still just takes a raw unit count, rather than factoring in unit quality.

 

[Grille]

Pook,
C3C (even 1.15) has some flaws that had been present in early versions of civ3 and were later fixed by some patch (IIRC, the military advisor issue is mentioned in the patch fixes/changes/additons-history-readme of civ3).
I don't know why some 'old' flaws showed up again in C3C, but I now *suspect* this is also the case for the military advisor comparison (just guessing here!). At least I've become a bit precautious when generally adapting vanilla rules to C3C.
Furthermore, if my assumption with the old, 'returning' advisor bug was wrong, then C3C might even use another, vanilla-different rule to calculate and compare the strengths of 2 civs.

 

[Pook]

Thanq, Grille. I wrongly assumed that the latest patch would have the most sophisticated formula.

 

[kb2tvl]

There are 3 places I look at when fighting. The first is the power graph. This seems to give a good relative representation of your forces(and allies) compared to your enemies. Next I look at the military service number. If I am first on this list, I have zero cash left for research.... generally. The last place I look is the military advisor. I think this guy is inexact and too vague to be of much use. He, apparently, compares you to the biggest civ.

 

[Rirse]

System for the military advisor is still a little off. In my previous game, he kept saying Egypt army was strong compare to our. Yet I had a spy in their country, and all he had was a bunch of infantry and only three calary. When I think of their army, I tend to think of the attackers, not their defender in the cities.