[C3C] Combat Stats - Determining A/D Ratios

[SayHayKid]

I am in the process of fine tuning the stats of the various units. I came across this handy Combat Calculator that has me re-evaluating my initial numbers. Especially now with Flintlock the AI will actually produce and use artillery. I first started by running the naval unit numbers through the site. With my ancient naval units I originally had 1/1/3 Galley (pretty standard) and was going to add an offensive ship at 2/1/3 (Trireme). But the calculator says that the Trireme would win 80+% of the time. Much higher than I expected. So, now I am running all the units through the calculator (same era offense vs same era defensive). It got me thinking, what do does everybody else prefer for their optimal A vs D ratio?

I am definitely glad I stumbled across the calculator. It is very handy.

 

[Predator145]

Ancient age has the highest being 3-2. But the swordsman also features a 3-2 cost ratio and moves slowly. Horses are fast but only have only a 1-1 ratio. Walls are expensive in relation to units but so are barracks. The thing is swords are easily mass upgraded and can simply keep going with medieval infantry long before knights hit the scene. So I've moved the upgrade for swords much later into the tech tree. And while they're not a dead end like in vanilla, you're stuck with them for a long time after that initial punch.

Medieval times gives you 4-3 at first. Walls are cheaper in relation to units, cities grow past size 6. But you also get a fast mover with a 4-3 ratio. And fast movers change everything. So things are pretty balanced. Even muskets with 1-1 are balanced because how early you can access them and how expensive they are.

Then comes cavalry with 3-2 on a 3 mp unit vs an overpriced defender. Balance goes out the window. The game becomes hyper offensive. This is where the AI could kill you easily.

Rifles arrive much later on the tech tree than cavalry. Now you're at a healthy 1-1 with the attacker being a 3 mover. The attacker can punch through with a big stack and sack cities but will trade badly in the process because of how cheap walls are compared to the defender and cities having mostly grown past size 6. The game has a healthy ratio again, but Nationalism is far too late compared to Mil Trad and the human player would want to go for it even less. So you're often forced to peacefully bide your time with muskets, rail things up and then go on the offensive with infantry and artillery. It's very repetitive.

Once infantry hits the scene the ratio is 6-10. The AI absolutely can't touch you while you can break it with artillery. And no, the AI being able to use artillery and armies with C3X doesn't change that. I've buffed ww1 era cavalry to 8 attack. It's still a defense dominant period, but the AI with an industrialized production base will be able to make offensive progress albeit with heavy casualties.

Then come tanks. Now we're at a 10-16 ratio while the cost ratio is 9-10. It's ridiculous. Forget about advancing in the open with infantry. You'll get run over. On lower difficulties you may as well do away with building Infantry unless you waste lots of shields making tanks in some cities. Tanks will only struggle against 200% def multipliers. This is where the AI can kill you again. Higher difficulties aim at having the AI tech to tanks and then wipe you and the weaker ones out. Who ever gets there first has a huge advantage. Tanks are a way to reward teching. But IMO they're unfun and restrictive. If the AI gets to tanks first it's likely game over. If you damage the leading AI enough before it has tanks it stands no chance. So while I'm fine with tanks being offensively potent they can't be 10 shields more expensive than infantry.

Then come mech inf. It's back to before. You can easily stop the AI's tanks while modern armor is very far away. The ratio of 18-16 is brutal if the attacker can't retreat. You can attack just fine with your tanks and mech inf because of artillery while the AI can't do anything.

It seems the sweetest spot would be 4-3, even 5-4 if the main attacker is a 3 mover. But we have to factor in cost. Pikes are much cheaper than knights. Infantry likewise should be a lot cheaper than tanks.

Naval combat in stock game has no retreating. But is also heavily centered on bombarding. The human player would never take a 50-50 trade. But the AI would do so instead of bombarding. Making offensive ships have a high ratio over defensive ones helps the AI. Or you could separate bombarding ships from attacking ships. Bombarding ships should have very low or even 0 attack. 0 attack ensures the AI will bombard.

 

[Sakharov]

I faced the similar stuff also. Also I'm puzzled with cost/attack (defense) ratio.

I agree that cavalry breaks the game. So I reworked medieval techtree. Cavalry appears much later as MilTrad have more complicated prerequisites and I made musketeers cheaper but still made them appear a bit later as I swapped chemistry and gunpowder. So it looks like this (it's still in work, so I haven't added arrows):



What about cost/attack (defense) ratio? In ancient and early medieval era most units costs as they attack (defense) value multiplied by 10:

• Warrior has 1/1 A/D and costs 10 shields.
• Archers and spearmen costs 20 shields. Former is an offensive unit, latter is defensive one; they both have equal strength respectively.
• Swordsman have 3 attack value and costs 30 shields. He have 2 defense value, but it's not overpowered IMO, as they require a strategic resource.
• Horsemen have same A/D as archers, but also they are faster, so they cost 50% more. Should it work similar way with other units?
• Same thing with longbowmen and MedInf. The latter cost the same but have better defense, but they require strategic resource. Also should they be swapped?
• Same thing with Knight. I reduced their defense to 2 and cost to 60, so they follow the same logic: they cost 50% more than their slower counterpart (MedInf).

That logic breaks in industrial era. Tanks have much more attack/cost ratio, modern armor even more. I think in industrial and modern era we can multiply that ratio by 1.5 more. Tanks should have attack of 10, according to their cost. What confuses me is that ratio is different for mechanized and foot units are different. By now I reduced infantry cost to 80, increased tank to 120 and modern armor to 180, but I think there are needed more accurate calculations.

 

[Predator145]

Tank at 120 shields is a good place to start. It actually hurts the human player on higher difficulties more than the AI for the AI is likely to make tanks in 2 turns regardless. With factories and powerplants up, the human player needs to avoid wasting shields. This encourages combined arms warfare. You can't just use all your cities to make tanks. Some will train infantry and artillery. Others will train air units. Going pure tanks would result in less output due to waste and higher casualties.

180 for modern armor IMO is a bit too much. Since we've increased tank cost by 20% the same could be for modern armor at 144 shields. However, with these kinda mods I also have cheaper and weaker attack units so that my productive 60spt cities can crank them out in 2 turns late game. I've also nerfed the modern armor's defense down to 12. A 3 mover that can both attack and defend well means you don't have to make anything else.

 

[vmxa]

"A 3 mover that can both attack and defend well means you don't have to make anything else." So true and you did not even mention it has blitz and could attack three times, ouch.

 

[Sakharov]

Tank at 120 shields is a good place to start. It actually hurts the human player on higher difficulties more than the AI for the AI is likely to make tanks in 2 turns regardless. With factories and powerplants up, the human player needs to avoid wasting shields. This encourages combined arms warfare. You can't just use all your cities to make tanks. Some will train infantry and artillery. Others will train air units. Going pure tanks would result in less output due to waste and higher casualties.

180 for modern armor IMO is a bit too much. Since we've increased tank cost by 20% the same could be for modern armor at 144 shields. However, with these kinda mods I also have cheaper and weaker attack units so that my productive 60spt cities can crank them out in 2 turns late game. I've also nerfed the modern armor's defense down to 12. A 3 mover that can both attack and defend well means you don't have to make anything else.

I agree that modern armor defense should be nerfed.
But also we see that in later eras that strength/cost ratio differs for attack and defense units. Though it's justified, cause game should end somehow and it becomes more biased to attack.
In stock game modern armor have 24 attack and 120 cost. 24/120 = 0.2. Mech Inf have 18 defense and costs 110. 18/110 = about 0.163. So if we re-calculate modern armor cost 150 shield would be fair enough. Nearly what you mentioned.

What I aiming is to make foot units more reasonable to use in later eras. It's not my idea to make mechanized units wheeled, but I like it. Also, should mech infantry be wheeled? Or maybe add a separate unit, let's call it "Modern Infantry", 12/18/1 which would be successor of Infantry.
Also I think that marines should be cheaper. They cost as much as modern armor, but are twice weaker and much slower! I think that single amphibious ability doesn't worth it. 3 MP plus blitz ability is really deadly. Is it good idea that guerillas upgrade to marines?

 

[Predator145]

I have the warrior and archer line merge into marines as well. They're also much more potent in the industrial and modern versions. The modern version has 18 attack, 12 def. So you still pack a good punch with just rubber. I also have modern light infantry, a cheap grunt unit with 10 atk, 13 def you can transport around in helicopters to cover your assaulting marines. With C3X heliborne assaults and paratroopers have been heavily buffed. They can move right after dropping and paratroopers don't need airports to airdrop (if you don't want to play with the half baked airfield worker job).

With mech inf and modern armor being 20%-30% more expensive some of your cities may want to build cheaper foot units to avoid waste. And with buffed stats they can hold their own in rough terrain and cities. Being air transportable by helicopters they can also quickly re enforce overseas and far away places (rail roads are limited to 18 tiles in my mods).

 

[Sakharov]

I have the warrior and archer line merge into marines as well. They're also much more potent in the industrial and modern versions. The modern version has 18 attack, 12 def. So you still pack a good punch with just rubber. I also have modern light infantry, a cheap grunt unit with 10 atk, 13 def you can transport around in helicopters to cover your assaulting marines. With C3X heliborne assaults and paratroopers have been heavily buffed. They can move right after dropping and paratroopers don't need airports to airdrop (if you don't want to play with the half baked airfield worker job).

With mech inf and modern armor being 20%-30% more expensive some of your cities may want to build cheaper foot units to avoid waste. And with buffed stats they can hold their own in rough terrain and cities. Being air transportable by helicopters they can also quickly re enforce overseas and far away places (rail roads are limited to 18 tiles in my mods).

Have you shared your modpack?

 

[Predator145]

Have you shared your modpack?

It's work in progress. I'm making mine with a ridiculous amount of flavor units and all the excuses to cram in different unit types into an epic game while trying to maintain good gameplay balance. I'm testing a minimal mod right now with no flavor units. I guess I'll share that first.

 

[Sakharov]

BTW, if there would be modern cavalry unit in modern or late industrial era, which stats and cost they it should have? WDYT? They would be reasonable with tanks marked as wheeled. Of course, they should be weaker than tanks but strong enough to combat infantry.

 

[SayHayKid]

BTW, if there would be modern cavalry unit in modern or late industrial era, which stats and cost they it should have? WDYT? They would be reasonable with tanks marked as wheeled. Of course, they should be weaker than tanks but strong enough to combat infantry.

I'd think about putting in an early tank or an armored vehicle. I have an Early Tank 12/5/2 100 shields. Cavalry is 7/3/2 80 shields.

 

[Predator145]

BTW, if there would be modern cavalry unit in modern or late industrial era, which stats and cost they it should have? WDYT? They would be reasonable with tanks marked as wheeled. Of course, they should be weaker than tanks but strong enough to combat infantry.

In my epic game mod, I have small incremental upgrades for the cavalry line. By late industrial era, all civs basically have access to the stock game Sipahi, costing 98 shields. Then at synthetic fiber, the last upgrade of that line would be "horse cavalry" (10/5/3, 110 shields, requires rubber and horses).

 

[Sakharov]

What comes to my mind is taking early "protorype" units and scaling their stats propotionally. IMO units from the same era should have equal strength/cost ratio. It may change, but in different eras (maybe).
Let's take early medieval. Game is biased towards offense, but offensive units are more expensive. We have pikeman a defender (pikeman), light offensive (longbowman, doesn't require resources), heavy offensive (med inf, require one resource) and fast offensive (knight, requires two resources). Is it a good idea just to "scale" them into industrial era? For example, multiply their A/D by 3 and cost by 2.
So there would be infantry instead of pikeman, longbowmen and med inf would merge into marines (or maybe there should be separate "light" and "heavy" attackers?), and tanks would take place of cavalry (they gain blitz ability, but are wheeled and have 2 MP instead of 3). But there are some intermediate units between them. I have to do something with guerillas and rifleman, still deciding yet.

 

[Ozymandias]

Much combat takes place at best of odds of about 3:2 (given inherent terrain advantages for the Defender.) A quick glance at A/D stats quickly shows how these ratios are maintained throughout the game. Plus you have wondrous oddities like he Ancient Hoplite & the Medieval Pikeman have the same stats. And Bronze-Age Immortals having better stats than Roman Legions.

In Terra Fantasia I'm having each relevant, military Tech increase give the odds of, e.g., Gunpowder:Iron:Bronze:Stone each having a direct, straight-out 3:1 odds against the last iteration of Units - of which there are about 10 throughout the mod ... Yet, due to the CRT's structure, there will always be a tiny chance of in any combat.

@Sakharov - I like having the following: Heavy & Light Attackers, Fast Attackers (and Infantry/Cavalry/Light Armor merging and diverging over time) Scouts with increasing A/D over time (and sometimes "merging" with Light Attackers), "Main" Defenders & "Emergency" (fast build) Defenders ... And note that this last is often Warriors, as the AI seems to simply pick the first "D" Unit in the Unit list's entry order - yet another reason to use All Hail Civinator's CCM's "Auto-Production (and Promotion) system.

-z