Now that my expanded editor is almost finished, and Sandris packs are almost done for the ancient era, it is time to start working again on my mod.

I'll probably restart" from scratch", but with the new editor it should be fast.

I'd like to post here my thought about the unit stats.

I will focus here on the land forces.

Scale of the progession

First, I have 24 subperiods in my game. Not every civ will have units available for all of them, but still the progression of the unit stats should theoretically allow that.

If I use a base "strength" of 1 for the first era, and then each following era increases the strength by 50%, I would end with a strength of 11233, clearly beyond the limit of the game...
To remain within an acceptable limit, I need to limit the increase to 30%, the maximum would then be 418 for the last era.

However, I don't want to start at 1, but at 3, so I can have some variation from civ to civ which are no to drastic. In that case, the increase between era should be 25%, for a maximum of 500 in era 24.

This was just to quickly estimate a progression scale.

It means if a unit has a strength of 10, it would upgrade to a unit with a strength of 12.

So the advanced unit would have 12/22 = 55% chance to win. It's still an advantage, but not to big, so you have some incentive to improve your army, but it's not a game breaker.

If you really lag behind, like a WWII unit vs a Napoleonic unit, this would give 80% to win.

However, such a regular progression is not very reallistic.

Is a Frankish spearman stronger than a Gallic spearman? There weapon and armor are roughly similar... So instead of a regular progression, I have decided to keep the 25% just as a guideline for major steps, and then to give a strength for major changes.

In between, it's fine tuning using different parameters: training/skill of the unit, equipment (armor? Shield? Helmet?), sometimes playing with the HP to differentiate between the units, etc.

Just for gameplay, I've tried to make upgraded units always better than the previous one, although in some rare occasion it is not the case (some kind of barbarian invasions. However, it happens that the upgrade in A/D is very small, or virtually non existant, but then the cost can be lower, or there could be a HP bonus.

Unit role
I don't really like the way the units are handle in CivIII.
Let's take cavalry, with 6/3... Two cavalry who fight in open terrain should roughly be the same strength and have a 50% to win.

But here, the attacking unit has a 66% chance to win...

With a pikeman it's even worse... Two pikemen units who fight on open field should be of similar strength. But as they are 1/3, the attacker has only 25% to win.

I don't really this, so I've decided to break the unit in something slightly different from the usual attack/defense lines.

My lines are:

- Main line: spearmen, pikemen, musketeer... They form the bulk of the army, and have a A roughly equal to D, but with a slight advantage to D.
Therefore a pikeman could be 3/4, no 1/3.

- Assault line: swordsmen, foot knight, grenadier. They are expensive, and have less HP, as they are small size unit. They have a better A than D.

- Range line: Javelineer, Archer, Crossbowmen. They have a range attack and can thus bombard, and usually have a ZOC. Javelineer have a defensive bombardment, an average attack and a low defense. Archer have a 1 range bombardment, ROF of 2. Crossbowmen have a 1 range bombardment, ROF of 1, but a higher bombard strength.
These units don't have a very strength bombard strength, but can protect themselves if attacked, although their defense is not very strong (so they are vulnerable and most be protected by main line units).

- Artillery : catapult and guns. They have a one range bombard attack, much stronger than the range line, but have a D and A of 0 and must not be left alone.

- Cavalry: the cavalry main advantage is speed. Its divided in 3 lines:
-- Light: low D, average A, good speed. Can have ZOC, and sometime blitz: so they can harrass, but are weak. Good for scouting, not so good for battles.
-- Heavy : good D, strong A: mainly used to crack ennemy lines, but expensive, and less HP (small size unit). Usually have blitz, but are wheeled.
-- Range: with a small bombardment strength, similar use as range infantry, but faster.

Example
This is just a base work, as it will vary from civ to civ. This is what a base unit line could be for an European civ.

http://stephane.f.david.free.fr/CivMod/Images/UnitLine.jpg

http://stephane.f.david.free.fr/CivMod/Images/UnitLineCavalry.jpg

oooh I dunno. I like your unit line ideas but the stat progression scares the willies out of me. Usually, the AI is leaps and bound ahead of me in tech but I can hold on in there by concentrating my forces and using my superior brain-power ;) I doubt I could achieve this (despite the aforementioned huge brain) if the AI was punting out 30/25 Grenadiers against my 8/9 Pikes and I'd have no chance of taking any of his cities. I suppose if you were to balance up other areas of the game such as tech trading then this might address the above.

Shouldn't grenadiers from 1700 easily destroy pikemen of year 1000 anyway?

And even then, 30 vs 9 = "only" 75% chance of victory.

Until widespread use of gunpowder, I think the difference is not so important: medieval army vs roman legion would have some advantage, but perhaps not that much.

Then Napoleonic infantry vs 1750 musketeer, the difference is also relatively small in term of equipment.

But Napoleonic infantry vs pikes?

I really like the idea of bigger increases, it should go a long way in avoiding the famed unfair spearman/tank encounters. I just hope the AI will not do miserably at upgrading them.

The attack line though, seems to have a pretty rough time of it with the -1hp penalty.

Using the combat calculator, a 5 hp, period 7 pikeman on grassland defending against a 4 hp swordsman of the same period will have a 53% chance of winning if not fortified, 64% if fortified.

If their roles are reversed with the swordsman defending the pikeman will win 61% of the time if the swordsman is not fortified, 51% if he is.

For comparison, the three medieval offensive units used in regular civ would all (all 3 had A of 4) have a 60% chance of beating a pikeman on grassland (given everyone was veteran) if the pikeman was unfortified and 49% if he was fortified.

A period 14 fusilier, unfortified on grassland, 6 hp would beat its offensive counterpart (period 14, 4 hp, grenadier) 68% of the time unless fortified in which case it would win 78%.

Needless to say "offense is the best defense" is going to be a little more challenging but then maybe that's what you're after. IMO, and without seeing your other lines--specifically the heavy cavalry, this greatly increases the importance of artillery which is something the AI stinks at especially when they (the artillery) have no A and thus no attack strategy.

You never actually gave the base hp and increases, all my calculations above will apply to a base 3, base 4, or base 5 hp system with 1 hp upgrades.

- Artillery : catapult and guns. They have a one range bombard attack, much stronger than the range line, but have a D and A of 0 and must not be left alone.

I you keep this big Civ 3 exploit in favour of human players (as artillery has a non-working AI-tactics), you don´t have to think too much about the stats of other units. They always will be spoiled (in my eyes).

I'm taking a similar approach in my mod. But I only have 8 ages (2 per science era) so the total increase in stats is not as large. The start unit is 2/2, with each upgrade increasing by about 50% strength (this starts to break down in the last few ages as new lines/roles form). The strongest land unit is the MBT at 160/100.
I also reworked the Offense/Defense roles and strength ratios to a lesser extent. My lines have roughly the following A/D ratios with respect to each other:
Balanced infantry: 1/1 (spearman, pikeman, musketman...)
Attack infantry: 1.5/1 (swordsman, man at arms, grenadier...)
Light infantry: .75(1.3*)/1.3, -1HP (javelineer, archer, line infantry...) *defensive bombard only
Light Cavalry: 1.3/.75 -2 HP
Heavy Cavalry: 1.6/.8, -1 HP
Base HP is 5-6-7-8. I'm still mostly sticking with the use of A/D strength to determine role, but I've mixed up the HP and defensive bombard to make some units better at specific roles.

Some interesting remarks here. When looking at the regular unit, you'll notice that with riflemen and infantry, they are 4/6 and 6/10, making them virtually useless in an offensive role, and I strongly dislike it as infantry should remain an important part of armies, even in attack role (WWI?).

That's why I try to have something more balanced.

I've updated the OP, with some improved stats for the swordsman / grenadier, with the idea they are elite forces and so should be quite stronger than main line.

That's also why I put a -1 HP, they are supposed to be small size. For the same reason, range line units also have a -1 HP.

However, I'm thinking of something else to make a difference.
1- The main line units will be draftable, and only them.
2- What if the main line units require no maintenance, but all the other do?

I have also included in the OP the stats for the cavalry.

Also, keep in mind this is just a guideline. There will be variation from civ to civ and unit to unit depending on the weapon/armor in the available animation, and some historical consideration as well. The goal is to make playing with each civ a unique experience in some way, even if the tech tree remains the same.

Example:

http://stephane.f.david.free.fr/CivMod/UnitIcones/Ancient_Infantry/Ge_GermanLaterFootKnight.gif
German late medieval foot knight, heavy armor, two handed sword.
Instead of being 16/12, he may end 18/13.

http://stephane.f.david.free.fr/CivMod/UnitIcones/Ancient_Infantry/It_ItalianSwordsman.gif
Italian foot knight of the same period. One handed sword + shield, but has only a chain mail, makes him normal a 16/12.

http://stephane.f.david.free.fr/CivMod/UnitIcones/Ancient_Infantry/Eg_MamlukSwordsman.gif
Mameluk, chain mail, small shield, lighter weapon, he will be only 15/11, but will be less expensive.

You have very good ideas and this has to be the most sophisticated and thought-out stat concept I've seen and I'll certainly try this when it's ready, despite usually not being a epic-gamer-player. I especially like how you adjust stats according to the ecquipment.

However, I'm not sure about the suggestion regarding maintenance. How does the AI handle maintenance free units? If this could lead to line-only AI armies it would be a pity not to see all those unit lines in action. Plus it would be very annoying to fight endless stacks of free defenders in every city the enemy possesses...

Hence the question ;). I'll have to try it a bit.

Also, my mod will come with different scenarios, thanks to the awesome new expanded editor one of the smartest poster here made. :mischief:

The idea is to make a big epic scenario, with 20 civs and 24 subperiods.

But at first, only the first 8 periods will be available (the first era, name undecided, including bronze age, iron age, middle ages), and for 8 civs.
Then I will expand it to more civ (Arab), and later Asian.

Then I can make another period, for instance modern only...

But from the main epic game, I will generate "sub scenarios", focused on some civ and some era.

Like only the 1st era for 10 civ.

This will make complete subscenarios, fun to play for all the included civ, without the need to discover everything.

So someone who wants to play directly with aircraft can select the modern scenario only.

As the generation of the subscenario from the main one will be mostly automated, it means I can work and upgrade the main scenario only, and then regenerate the sub scenarios, so it should be fairly easy to keep all of them up to date :).

Personally, I like the Stronger Stats and have found that you cannot gain the Kind of Superior Strength desired for a unit by what would seem to be Fantastic STATS because random play factors disregard the settings.

Defense and Hit Points are the Strongest settings to help insure Unit Battle Strength.

IF a Unit is set as requiring No Support...the AI will build extreme numbers of them. Other Units will be built but not as many as the ones that Require No Support.

I really like the way you're planning to set out your units. I have been similarly very unhappy with the simplistic way civ threats units stats.

regarding this,
"- Range line: Javelineer, Archer, Crossbowmen. They have a range attack and can thus bombard, and usually have a ZOC. Javelineer have a defensive bombardment, an average attack and a low defense. Archer have a 1 range bombardment, ROF of 2. Crossbowmen have a 1 range bombardment, ROF of 1, but a higher bombard strength.
These units don't have a very strength bombard strength, but can protect themselves if attacked, although their defense is not very strong (so they are vulnerable and most be protected by main line units)."

I like this and have been thinking of doing something similar, though never implemented it. I never implemented it because I figured the AI would never use them correctly or bother to build them. Have you got ideas on how to get the AI to have more variation in the units it builds and/or uses? Or is your unit variation purely for the benefit of the human player who can understand and enjoy such variation.

The only way I have ever managed to add variation to the units the AI uses is to have different units, that the AI would not normally build, auto-produced by buildings. Are there any other ways to add variation to the AI's units and the way it uses them?

Again I like your ideas a lot!

I don't think the AI will use the unit very well, but that's what multiplayer is for ;).

True. Good ideas anyway. I'm all for trying to build in more strategy/realism to the warfare in Civ 3. I look forward to your mod!

Again I like your ideas a lot!
He has good ideas :)

The only way I have ever managed to add variation to the units the AI uses is to have different units, that the AI would not normally build, auto-produced by buildings. Are there any other ways to add variation to the AI's units and the way it uses them?

HERE WE GO AGAIN (http://forums.civfanatics.com/archive/index.php/t-62138.html) ;)

... Now, back OT:

Hey Steph - I've been playing around with unit stats for my own one-day-when-Hades-freezes-over to be finished mod, and I have a suggestion, which you yourself point to when you rhetorically ask about gunpowder vs. pikemen.

Breaking the unit sequence into 24 stages is both commendable and interesting. I haven't gone to that degree of granularity (I'm using 16) BUT I draw lines between each group of 4 (which are delineated by a great technological leap like gunpowder) and then, with each new group of 4, increase the A/D factors dramatically.

I also try to alternate the advantage from defender to attacker and then back again.

So part of my Era 1 European unit line looks like:

BEGINNING UNITS:

D: Spearmen = 1/2/1
A: Chainmail Knight = 4/3/2

"PHASE TWO"

D: Pikemen 2/3/1 (with tech "Close Order Drill)
-then-
A: Platemail Knight 6/4/2 (with tech "Steel" - IIRC ca. 1200 CE historically)

- Anyway, the strongest unit A/D (there are others which take their own places in the line, e.g. bowman > crossbowman OR the English UU longbowmen) is the 6/6 crusader.

The first gunpowder unit - "Musket & Pike" - is a 10/8 which (even though we're still playing with small numbers here) gives the M&P ~85% of defeating a crusader in the open (I give flat terrain a 100%, not 110%, defensive bonus) and a crusader attacking a M&P ca. 40% chance of victory.

So tactics also change as different weapons systems become available.

- And so forth :king:


Best,

Oz

I haven't used the Combat Calculator here so the odds are a bit off, but generally...

Looks pretty good. I'd be inclined to give a bit stronger defence to the WWI-era trench infantry, reflecting the difficulty of any troops really overcoming the machine guns and barbed wire of the era effectively. The Gallipoli expedition shows this most dramatically, but even in France you have huge expenditures of manpower for miniscule gains, and very often with very high casualties. A 10% defence bonus for regular ground troops seems a bit small, and a Stormtrooper having a 15 vs. 11 ratio in favor of defeating a defensive unit seems rather high. Most odd is the ranged unit having double the offence of defence, when in most eras it is about the same offence and defence. If MG stands for Machine Gun, this really doesn't make sense - the machine gun is a fundamentally defensive weapon, and standard machine guns are usually too heavy to be used effectively in regular combat. Light machine guns were developed near the end of WWI (ex. BAR 1918), but especially for the WWI era, the defensive value of this unit should be much higher than the offensive value. Really it should probably have a higher defensive value than a regular Trench Infantry, at least equal, countered by either fewer hitpoints to reflect the smaller size of an all machine-gun squadron, or a higher cost in shields. This doesn't quite fall in line with most of the ranged units, having somewhat poor defence on their own (as they should), but most machine gun squadrons would be fairly good on defence provided they were prepared to defend the area.

A late-era Cavalry will still stand a 10 vs. 14 ratio of winning on an attack against a WWII tank - that's almost identical to regular Civ3. They'll also have a 4 vs. 1 in the tank's favor on defence, actually better than in regular Civ3. This seems somewhat necessary to keep infantry values in line, but seems a bit of an anamoly. Late 19th-century Cavalry is presumably heavy saber cavalry, which probably wouldn't fare too well in such a situation. Mounted infantry that used horses primarily for faster transport would have a more realistic chance of success, and could perhaps be used by civilizations that lacked Oil to still have a fast unit in the first part of the 20th century (provided they did have horses). This isn't any great gamebreaker, but Cavalry seemed a bit stronger against Tanks than they probably should have been in regular Civ3, and this doesn't actually correct that at all for late Cavalry. Again, I'm not sure how big an issue this will be - if there aren't very many of the latest Cavalry built before WWI Tanks come along, this issue will sort itself out pretty well. But if someone upgrades 100 Cavalry to the last horse cavalry, this could play very much like regular Civ3 in that regard.

All things considered though, it looks excellent. From the spreadsheet, it looks like it'll play quite well. Looking forward to seeing how this ends up.

This makes me wonder, though - is the game going to be significantly longer with all these eras? Seems like with 24 subperiods of about 20 turns each, research would have to go lickity-split to fit in several different technologies for each era. Although I suppose this leads into the next thread that you've posted - good anticipation of what questions this thread would lead to!

Looks pretty good. I'd be inclined to give a bit stronger defence to the WWI-era trench infantry, reflecting the difficulty of any troops really overcoming the machine guns and barbed wire of the era effectively.
A 10% defence bonus for regular ground troops seems a bit small
Yes, but don't forget you are yourself giving example of fortified position. So the WWI-era trench infantry has only a slight defensive bonus, but if you factor bonus for defensive position + bonus for fortified units + bonus for fortification in the tile (that would represent correctly the situation), then you'll see that the WWI infantry would suffer a lot of casualties when attacking a defensive WWI infantry.

, and a Stormtrooper having a 15 vs. 11 ratio in favor of defeating a defensive unit seems rather high.
The stormstrooper were relatively effective in their assault, and with the above remark you'll see that the end result to attack entrenched units is not as good as it seems.

If MG stands for Machine Gun, this really doesn't make sense - the machine gun is a fundamentally defensive weapon, and standard machine guns are usually too heavy to be used effectively in regular combat. Light machine guns were developed near the end of WWI (ex. BAR 1918), but especially for the WWI era, the defensive value of this unit should be much higher than the offensive value. Really it should probably have a higher defensive value than a regular Trench Infantry, at least equal, countered by either fewer hitpoints to reflect the smaller size of an all machine-gun squadron, or a higher cost in shields.
I forgot to include a little something... MG should ahve a stealth attack vs "soft" units (infantry) while AT will have a stealth attack vs armored units. Hence the relatively good attack. Beside, the MG has a bombardment factor, (I forgot to put a high ROF), and that's how it will be used for defensive purpose. If attacked by infantry, the infantry needs first to come adjacent, then the MG can bombard to wear it down, and then the infantry attacks the next turn.
If attacked by fast moving unit (ie Tanks), the MG may not have time to fire.
Result: MG are good to defend against infantry, not against tank. It achieves its historical goal, but with a different approach.


A late-era Cavalry will still stand a 10 vs. 14 ratio of winning on an attack against a WWII tank - that's almost identical to regular Civ3. They'll also have a 4 vs. 1 in the tank's favor on defence, actually better than in regular Civ3. This seems somewhat necessary to keep infantry values in line, but seems a bit of an anamoly. Late 19th-century Cavalry is presumably heavy saber cavalry, which probably wouldn't fare too well in such a situation.
Actually, late 19th cavalry is equipped with rifles and probably grenade. These stats are just general guidelines, and will be adapted on a case to case basis to reflect the actual available units.
I agree it looks odd, but the only way to make it reallistic would be to make tanks much better than infantry in offense AND defense, but then the AI would build nothing else. If we did not have the building limitation, I would have used auto producing buildings for tanks, aircrafts and ships, in order to limit their number, but as it is now it's not very easy to do.

I want to test with main line infantry costing no maintenance, and tanks costing maintenance, to see if it put the infantry/took in good proportion, and have the infantry cost population to keep the numbers somewhat reasonnable, but I'm not sure it will work well.

Anyway, I'll focus first on the ancient/medieval period, where this problem is not so important.


This makes me wonder, though - is the game going to be significantly longer with all these eras? Seems like with 24 subperiods of about 20 turns each, research would have to go lickity-split to fit in several different technologies for each era. Although I suppose this leads into the next thread that you've posted - good anticipation of what questions this thread would lead to!
Indeed, good anticipation :)