[R&F] “Knights are too weak!” - A tale of Combat Strength across the era

Aurelesk

King
Joined
Oct 26, 2017
Messages
602
Now that I have your attention, let’s discuss about something: the power of the units. I was wondering how the Combat Strength increases across the era. And it’s pretty simple: it’s around +17 Combat Strength for double Production. A unit that need 50 Production for 25 CS will have 42 CS if the cost is 100 Production, and then 59 CS of the cost is 200 Production, more or less.

So here a Chart:
Spoiler :

X axis : Combat Strength of the unit
Y axis: Rentability of the unit, based on this formula Rentability = 2^(CS/17-1)/Production

We can see a trend. We do see that the rentability of a unit slowly increase with era starting by the medieval era… except for ancient/classical era that are way different. As we may expected, Quadrireme and Pikeman are one of the worst unit of the game and the Slinger, Warrior, Archer, Horseman and late game units are one of the best. That’s why early warfare are that powerful.


Why +17 Combat Strength you may ask? You may observe that a Corps and a Army increase by +10 and +17 the Combat Strength of unit. Yes, +17 and not +20 for the Army. It’s because the Combat Strength isn’t linear. +10 CS is equal to +50% damage done and can endure 50% more damage (so -33% damage taken), making the unit 50% more effective. +17 is equal to +100% damage done and can endure 100% more damage (so -50% damage taken).

Here a chart:
Spoiler :

You may say that combine 3 units in 1 isn’t worth because you go from a 3× Power to 2× Power without the support/flanking bonus, but due to 1 unit per tiles and the healing factor, the Army have better survivability, have more Promotion and get better.

That’s why I am kind of dubitative about increasing by 17 the CS for double Production for units, because that’s not working that way for Corps and Army. For example. 1 Swordman is better than 2 Warrior (without Support/Flanking), because the Swordsman can focus one unit at the time, meaning he will survive. Plus, only an Army of Warrior can be as strong as a single Swordman.


The needed Production is scaled with the technology, except the Anti-Cavalry unit that are kind of misleading. I do think that the Anti-Cavalry line was suspected to be 15% cheaper to produce but suffer from -4 Combat Strength in exchange. But the Spearman cost is the same of the Heavy Chariot but suffer from the -4 CS. The Pikeman also suffer from the -4CS but cost 15% more! Somehow, it is the Knight have have the -15% Production! Only the Pike and Shot, the AT Crew and the Modern AT kind of follow this rule of less required Production but less powerful.


Strategic Resource tend to increase a little the overall Combat Strength of the unit. The Coal units have more Movement instead I think. I put all the units Combat Strength, Production / Gold cost… in a sheet right here. You are free to use it.
I tried to find a formula to balance all the units by changing the Production and Combat Strength, but I must admit that I failed. With my formula, Knights need 20 more Production but get 3 CS more. That's why I put that title!

The main problem with Knights are not their inner Strength. They are almost below average! The problem isn't the Kinghts itself but how we get to it. It is possible to get a late medieval era units by searching only 4 technologies (Animal Husbandry, Archery, Horseback Riding, and Stirrups). Plus, upgrading units are cheap in gold. For example, a Heavy Chariot is 260 Gold and Knights are 720 Gold but the upgrading is only 180 Gold, without the policy card that decrease the cost!. So you better buy the Heavy Chariot and upgrading it than buying straight away!
The other reason is there is no way to counter a Knight, having a long time window of utility. Spearman is weak: 25+10 < 48. Even the Pikeman is weak 41+10 > 48, but what is +3 CS when you have more Movement, require less Science (nor going to a leaf tech) and have an avantage against all others units even the Musketman (55 < 48+10) ?

So here my suggestions:
  • Make the early Anti-Cavalry more useful:
    • Most importantly the Pikeman that cost a lot and have a reduced CS. It should need 20/25% less Production but 7-10 more Combat Strength. Maybe change the prerequisite of Military Tactics from Mathematics to Currency and/or make Military Tactics a prerequisite of Education.
  • Add one more prerequisite to Stirrups like Iron Working
  • Increase the upgrading cost a lot:
    • It is too easy to pre-build an army with the Production card and then upgrade it for a cheap price and roll over.
  • Decrease the rentability of units accross the era instead of increase it.
  • General balance change, like increasing the Production cost of early ranged unit and/or reduce the ranged Combat Strength.

And you, what do you think about the units and Combat Strength and how it increase across the era?
 
Really interesting post. There was someone making some lengthy analysis about this about a year ago or something, probably someone will be able to link those posts. One of the main conclusions was indeed that Pikemen were too expensive (and I remember the Khevsur was even worse, although I think subsequently it got a slight, if not sufficient, improvement in a patch).

I'd like to say, I don't think upgrades should be more expensive, at least not early game, but the 50 % promotion card needs to be removed, the sooner the better. That card is incredible imbalanced and does so much bad for the game. The early game upgrade cost of 70 gold from Warrior to Swordsman seems perhaps a bit on the low side, but upgrading Archers to Crossbows for 190 gold really hurts my inventory.
 
And you, what do you think about the units and Combat Strength and how it increase across the era?

Units have no value unless the can get to where you need them. Units are no good to you dead. That's why Knights are so much more valuable at then other units of the same era. If movement scaled like attacking/defending strength knights would be a lot less valuable because they could not escape out of harm's way. Or give Knights cumlative movement every other turn to simulate the fact that maintaining a large warhorse would require feeding and rest. You get say 2 movement a turn but can save up 6 for a charge.
 
Really interesting post. There was someone making some lengthy analysis about this about a year ago or something, probably someone will be able to link those posts. One of the main conclusions was indeed that Pikemen were too expensive (and I remember the Khevsur was even worse, although I think subsequently it got a slight, if not sufficient, improvement in a patch).

I'd like to say, I don't think upgrades should be more expensive, at least not early game, but the 50 % promotion card needs to be removed, the sooner the better. That card is incredible imbalanced and does so much bad for the game. The early game upgrade cost of 70 gold from Warrior to Swordsman seems perhaps a bit on the low side, but upgrading Archers to Crossbows for 190 gold really hurts my inventory.

To me, I would love to see them completely change things around, so that in general units were ~1/2 the cost to build that they are currently, but upgrading would be ~double what it is now. Or at the very least, it should be an identical cost to build as it would be to build+upgrade.

The only other thing I'll say is that the chart is a little awkward, since you factor in combat strength on both axis. I'd suggest making the Y axis as something like lg(production)*17. Doing it that way really shows how out of line the Pikeman (and crossbow) truly is, and it's a lot easier to see how to balance, since then it's simply a matter of changing production to move them up and down, or combat strength to go left and right
Spoiler :
cs vs Prod.PNG

(I have Pikeman cost 200 from the site I pulled this from. Were they changed down to 160, or is that simply the cost that people think they should be?)
 
Really interesting post. There was someone making some lengthy analysis about this about a year ago or something, probably someone will be able to link those posts. One of the main conclusions was indeed that Pikemen were too expensive (and I remember the Khevsur was even worse, although I think subsequently it got a slight, if not sufficient, improvement in a patch).

I would like to see that. I have done a research before to see if someone have already talked about the subject but didn't find anything. I should have search longer!

I'd like to say, I don't think upgrades should be more expensive, at least not early game, but the 50 % promotion card needs to be removed.
I think exactly the contrary. It's on the early game that it's too easy to roll over someone by rushing key technology and upgrading at a cheap cost. The defender can't do nothing because he will pay the high price for the units to defend himself. So by increasing the upgrading cost, the attacking civilization will have less powerful unit and the defender will have a better chance to resist. Later in the game, that's not a issue anymore: so I'm ok by the 50% promotion card, even more if the rentability of the later units are reduced.

That's why Knights are so much more valuable at then other units of the same era. If movement scaled like attacking/defending strength knights would be a lot less valuable because they could not escape out of harm's way. Or give Knights cumlative movement every other turn to simulate the fact that maintaining a large warhorse would require feeding and rest. You get say 2 movement a turn but can save up 6 for a charge.

Well, a mobile unit should have less Combat Strength. In my failed formula, a unit lose 2 CS for each movement over 2. I like your idea, but I would rather prefer that if a unit didn't attack, take any damage for the whole turn, and didn't use all the Movement Point, then that unit start the next turn with +1 Movement. Something that will be more useful for unit with low Movement, like Melee for example.


The only other thing I'll say is that the chart is a little awkward, since you factor in combat strength on both axis. I'd suggest making the Y axis as something like lg(production)*17. Doing it that way really shows how out of line the Pikeman (and crossbow) truly is, and it's a lot easier to see how to balance, since then it's simply a matter of changing production to move them up and down, or combat strength to go left and right
I wanted to show specific things:
  • The rentability of units across the age. I should put on the X-axis the era, not the CS, to make that more obvious.
  • The higher the value, the better. I don't have to imagine a line and see where the Pikeman is. Your Pikeman is higher than the average and if you are not familiar with chart, could conclude that Pikeman was too powerful.
  • I wanted to show that the rentability of unit increase with the era from medieval to information, but ancient/classical era is a mess of over-high-rentability units.
  • I wanted to show that the Knight isn't a "good" units, and that the CS/Production do little about warfare but more specific imbalance mechanic like the Gold cost of upgrading and the Pikeman ineffectiveness. Same thing for Archer rush: pre-build Slinger and upgrade. That's also working with Swordman, but Knight roll over Swordsman, and Anti-cavalry that should roll over the Knight are useless, making Knight-rush even more powerful (and making Pikeman-rush irrelevant).

Your chart show many things useful! But not the one I wanted to show!

(I have Pikeman cost 200 from the site I pulled this from. Were they changed down to 160, or is that simply the cost that people think they should be?)
All the value are from the french civilopedia ingame. And it's really 200 Production. Just made a game just to see if the french civilopedia is, again, misleading with false value. And not, even in game the Production cost is 200. Maybe you have a mods that change the Pikeman Production cost?
 
I think I was thinking that 160 was the cost that Pikes *should* be (if you set it to 160 in my chart, they drop to be right in line with all the other units). And yes, my chart doesn't really show off the deviations from the line very well. For example, if you take a purely linear trendline to my chart, then all the early units are well below the line (ie. cheaper than their strength ratio), all the middle age units are above (ie. more expensive), and the late units actually come back below the line (apart from the Machine Gun). Your chart essentially shows the difference from the trendline, with the early and late units being above the middle, and the middle age units being below the middle.

So really, from what I can see, the main takeaway really should be that medieval/renaissance war is actually way more expensive than early or late war, based on this formula.
 
All the value are from the french civilopedia ingame. And it's really 200 Production. Just made a game just to see if the french civilopedia is, again, misleading with false value. And not, even in game the Production cost is 200. Maybe you have a mods that change the Pikeman Production cost?

I think I was thinking that 160 was the cost that Pikes *should* be (if you set it to 160 in my chart, they drop to be right in line with all the other units).

I believe GS splits the difference, and makes Pikes now cost 180. Unless I misunderstood that pending change.
 
There was someone making some lengthy analysis about this about a year ago or something, probably someone will be able to link those posts.
C'est Moi! Link here: You can see my conclusions in this thread.

And you, what do you think about the units and Combat Strength and how it increase across the era?
I think that units were designed so that a unit line would see +10 strength per era. Each class has rules dictating strength.
There are 2 systemic breaks; one is how ancient era units work and the other is how units in the modern era on work.
Cost is currently a pure function of tech tree appearance +/- a little noise. In GS they seem to be looking to mix things up a bit. It's why it's virtually impossible to find a good regression on the unit's other stats; pikes, xbows and knights were made to fight each other, so why does it matter that one comes a tech ahead of another?

Production scaling in this game is 10:1, which is a real pain.

They've also only ever touched the strength of a handful of units post release: swords (not their UUs) and military tactics melee unit UUs. I think the latter was preceded by enormous fan discussion/hating on how bad khevsurs were.
Pikes I suspect were touched pre release in direct response to how civ5 BNW's meta turned out after the last patch in ~2014-2015.


The State of Combat Units (Post Spring 2018 Patch)
Then there's another post i made on the second page detailing that costs are just a disaster.

Costs in civ6 scale 1:10 from turn 1 to future tech/civic, and linearly as you go through the tech tree. That means if you start at 1 unit, 11% through the tree you're at 2 units - double the cost. 22% gets to 3 units; only a 50% rise over the last era. Etc. So it's a weird ramp. It makes the last upgrade super powerful because you get that era and a half power bump for a pretty small cost increase.
Districts and buildings get hit the other way; since building yields don't grow the same way, late game buildings stink.
 
but the 50 % promotion card needs to be removed, the sooner the better. That card is incredible imbalanced and does so much bad for the game.
This does more to promote the "war is optimal" and "chop til the ice caps drop" meta than any other feature. It's turbo cancer. If they want to give players a card to cut down strategic resource usage in GS sure, but get rid of the gold discount. Instead, they are rolling both effects into one card. I cry.
 
In GS the Knight will cost 220, and the pikeman 180. Plus maintenance 4 vs 2. And 20 iron for making a knight.
 
In GS the Knight will cost 220, and the pikeman 180. Plus maintenance 4 vs 2. And 20 iron for making a knight.

Is 20 iron a lot? I hope it takes much sacrifice and effort to build an army of 8 or 9 Knights, because that amount of Knights combined with a battering ram on turn 85 wrecks everything in its path with ease.
 
Is 20 iron a lot? I hope it takes much sacrifice and effort to build an army of 8 or 9 Knights, because that amount of Knights combined with a battering ram on turn 85 wrecks everything in its path with ease.
One iron deposit provides 2 iron per turn. You can store something like 20-25 iron base, +10 per encampment/encampment building.
Further, there is a card (equestrian orders) that gives +1 resource/turn to iron and horses; the professional army card also now drops upgrade resource requirements by 50%. So while it's 10 turns per knight per source of iron base, that drops down to 6.7 turns w/ equestrian, 5 turns with pro army from chariots, and 3.3 with both. 2.5 if you're england and gain another +1 from iron and coal sources.
 
Speaking of four techs... I sometimes get to Renaissance era researching only a few techs using the naval line at the top.

It's hilarious, but kinda immersion-breaking.
 
Speaking of four techs... I sometimes get to Renaissance era researching only a few techs using the naval line at the top.

It's hilarious, but kinda immersion-breaking.

Thankfully that has been fixed a bit. There is one more tech there, so you don't fully skip the medieval era.
 
Upgrading cost for each type should scale downwards with progress in the tech or civics tree. Cost * (#tech before prerequisite tech/#techs researched). The nice thing about that system is it's really severe early when rushing is a bigger problem because players have fewer units to lose. I think there would be some push back because lots of people like tech rushing units.
 
One iron deposit provides 2 iron per turn. You can store something like 20-25 iron base, +10 per encampment/encampment building.
Further, there is a card (equestrian orders) that gives +1 resource/turn to iron and horses; the professional army card also now drops upgrade resource requirements by 50%. So while it's 10 turns per knight per source of iron base, that drops down to 6.7 turns w/ equestrian, 5 turns with pro army from chariots, and 3.3 with both. 2.5 if you're england and gain another +1 from iron and coal sources.

Professional Army was already too powerful. Letting it discount resource requirements will make it even more broken, and undercut any limitations resource requirements created. Hopefully this gets rethought.
 
Professional Army was already too powerful. Letting it discount resource requirements will make it even more broken, and undercut any limitations resource requirements created. Hopefully this gets rethought.
Just wait until the AI won't use the card and humans will, leading to an even worse gap in fielding modern armies. One hopes this was considered in testing and design.

Plus maintenance 4 vs 2.
The difference in production isn't really very meaningful (220/180=22%) as much as the combat difference (15% favor to pikes) and this maintenance. 4gpt is enough to really hurt you. Before, both cost 3.
You only pay the extra production once and the gap is pretty small; the issue is much magnified with tanks and AT crews. Tanks cost 20% more and have 80 str, while AT crews have 70+10 vs mounted, and have 2 less move. I'm not really sure you can call that a counter unless GS really restricts your oil.

In relevance to the thread, tanks are a more extreme version of knights. Knights have +3str over the trend, pikes have -4.
Tanks get +5 while AT and infantry get -5 in this one era for no known reason. This leads to: Infantry whoop AT, Tanks beat Infantry, tanks match AT crews but can just drive around them anyways. And MGs dont come until atomic; there's no ranged unit to save you from tank onslaughts like xbows can ping at knights. Oh well. Easy tweaks in a community balance patch after the 3rd expansion :D
 
Last edited by a moderator:
Top Bottom