Military

Discussion on game mechanics, balancing etc.
Forum rules
Posted relevant content can be used under GPL or GFDL (http://www.gnu.org/) for the project. Thanks!
Haxaco
Posts: 48
Joined: Sat Mar 17, 2012 1:07 pm

Re: Military

Post by Haxaco »

Is it possible to use a unit names file as used by Victoria, Hearts of Iron?

Basic template of the unitnames.csv is like this:

Country tag:Unit type:Country specific unit name

So you could have

France:Infantry:1ère Regiment d'Infanterie
....
Prussia:Infantry:4. Garde-Regiment zu Fuß
Prussia:Infantry:Garde-Grenadier-Regiment Nr. 5
....
and so on and so on.

etc. etc. this could also be used for ship names.
User avatar
Trilarion
Founder
Posts: 845
Joined: Thu Jan 12, 2012 9:27 pm
Location: Central germany

Land units and land battles

Post by Trilarion »

Types
  • Militia: stationary in provinces, max. limit per province, mostly good for defense, probably self-creating, low upkeep
  • Infantry: cheap, short ranged, slow, weaker than cavalry, good in defense and in city attack,
  • Cavalry/tanks: expense, short/medium ranged, fast, weak in cities, effective against artillery and infantry
  • Artillery: long ranged, slow, weak defense, effective against infantry and cities
  • Officers: only one per army, increase moral and attack values of surrounding units, only self-defense values
Properties
  • Attack strength: integer number 0-20, no separate defense strength
  • Unit strength: percentage 0-100% (rounded to multiples of 10%)
  • Attacking range: integer number 1-5 (responsive fire only for attack range 1)
  • Movement: integer number 1-10
  • (Morale: percentage 0-100%, do we need that?)
  • Training/creation costs: $
  • Training time: turns
  • Upkeep: $
Battle
- Field = Square map of staggered layout tiles, with 15-25 tiles length. City (1-3 tiles?) in the middle.
- Terrain is autogenerated from underlaying province terrain (can be edited in editor if wished)
- Some terrain (mountains, forests, swamp) slows down movement (-1) and others gives defense bonuses (+25% on defense on hills, ...)
- Each unit moves and attacks. It doesn't have to move, Attacking always ends turn.
- Each unit can maximal two times per turn respond to an attack with firing back.
- Maximal number of battle turns per battle - next game turn the province can be attacked again

Damage model
- Unit 1 attacks Unit 2 with attack values a1 and a2
- Unit strengths are s1 and s2 respectively
- Damage on unit 2 is d
- Division formula: d = c * s1 * a1 / a2 with fine tuning constant c
- or Difference formula: d = c * (a1 * s1 - a2 * s2) * 2 / (a1 + a2) with fine tuning constant c
- maybe add chance element to attack values before (gaussian random modifier)
- Strength is reduced by damage immediately, then other unit strikes back
- If s2 <= 0, unit 2 is killed and taken away
User avatar
Trilarion
Founder
Posts: 845
Joined: Thu Jan 12, 2012 9:27 pm
Location: Central germany

Post by Trilarion »

I just presented here the land unit and combat model as I think it could be done. Main difference to original Imperialism are:

Shorter ranges
I hope this gives a more distributed action where position of the units towards each other is more important

Limited city space
Not all units should sit inside the city walls. Artillery can sit behind.

City walls not impenetrable for infantry
City walls should help for sure but in the original they played an extremely important role. I want to weaken this a bit in order to promote other strategies.

Together with the different terrain I hope for a more diverse and a bit like Panzer General-like experience.

Hexagonal tiles are not needed though. I checked the staggered layout and it is completely equivalent to a rotated hexagonal layout, i.e. the geometry is the same.
User avatar
Trilarion
Founder
Posts: 845
Joined: Thu Jan 12, 2012 9:27 pm
Location: Central germany

Call for opinions?

Post by Trilarion »

So what's your opinion about this land unit/battle model?
Xylander
Posts: 37
Joined: Wed Mar 06, 2013 5:32 pm
Location: Germany
Contact:

Re: Call for opinions?

Post by Xylander »

Trilarion wrote:So what's your opinion about this land unit/battle model?
Sorry, but I never wanted a different battle model for Imperialism. In this kind of game I usually skip battles and I am satisfied with automatic resolution. When I want to play battles I either play Panzer General (or the like) or Total War.

Fighting battles is time consuming and I like to concentrate on economy, diplomacy and grand strategy (not tactics and unit movement).
Don't repeat yourself. (DRY)
Keep it simple, stupid. (KISS)
You ain't gonna need it. (YAGNI)
http://www.clean-code-developer.de/
User avatar
Trilarion
Founder
Posts: 845
Joined: Thu Jan 12, 2012 9:27 pm
Location: Central germany

Post by Trilarion »

No problem. Some people do the battles, others don't. I did the battles in the original Imperialism and always felt that city walls are too decisive. Quite often the attack was aborted completely or the AI was acting suboptimal by losing units without gaining anything. In Civilization for example they have economy, diplomacy and battles nicely connected - everything is fun there. I think military and battles should remain integral part of the game play, but of course with the option to automatize battles. And for the battle rules itself I want to change what I feel should be changed. The main difference would be that the range of the units is shorter, so position and terrain is more important. Also the city wall should be more penetrable and not hold enough space for all units, additionally giving room for more field encounters. Yes this will change the battles - I hope improve.

Separate from all this we need a damage model. Even the original Imperialism had one which we just don't know.

Anyway there will always be an option to automatize battles. The player arranges the army, selects a general attitude (aggressive vs. cautious) and watches the battles or moves fast forward to the results.
Xylander
Posts: 37
Joined: Wed Mar 06, 2013 5:32 pm
Location: Germany
Contact:

Re: VII. Military

Post by Xylander »

I don't believe that we'll be able to get an AI doing the battles you think of. So tactical battle might be an option only for player vs player battles. As most people play the game against AI this part of the game will be - compared to the effort to realize it - unimportant. Sorry for saying that :-)

Even if we stick with very simple battles like in Imperialism our AI will probably a very weak opponent to a human player. Allowing human players to fight tactical battles destroys the game's balance. Fighting tactical battles against AI oppenents in Imperialism is (like in Total War) an exploit. It allows players to win battles they should not win, giving them the advantage to have more survivors from battles while spending less money on military. That makes human players overly strong (compared to AI players).

Many games compensate for this giving AI players advantages. Players either face opponents with much bigger resources (most strategy games) or watch the AI 'cheating'. I dislike both approaches.

Maybe we can design our game so that fighting tactical battles is a game on its own? We could define how to export settings for a battle and how to import results and start a project on its own that focuses on tactical battles.
Don't repeat yourself. (DRY)
Keep it simple, stupid. (KISS)
You ain't gonna need it. (YAGNI)
http://www.clean-code-developer.de/
User avatar
Trilarion
Founder
Posts: 845
Joined: Thu Jan 12, 2012 9:27 pm
Location: Central germany

Post by Trilarion »

I would like to keep tactical battles in the game. Just having two armies crashing into another and then an algorithm to compute the output is not interesting for me since it's in a way like comparing two numbers and the larger one wins.. If I make the effort and build an empire to support an army I also want tactical battles and exert some direct influence on the result.

If the average AI has a worse performance than surely players who play with tactical battles have an advantage and will win more often, making the game easier. Without giving up tactical battles I see no way around it?! One should not in any case compare game progressions for players with tactical battles and players without.

Now wanting a strong AI and not spending too much time in the battle part are big arguments for simple battles. But then one shouldn't forget that Imperialism is also a war game. Building up an industry and making treaties is nice but in the end you have to conquer one or two countries at least. For me that is part of the fun of the game and an important part of winning too. So battles should not be too simple.

Now the difference between my and the original battle system is basically the range, so positioning of the units matters more. However it already mattered in the original to some extent. I don't really see one of them being simple and the other one impossible to realize. But it will be interesting to investigate how such battles can be simple and challenging and fun and doable by an AI at the same time. One element might be favor small numbers of units or to carefully adjust the range (fraction units/tiles on map). We could explore it in it's own project if enough interest is for that.
Veneteaou
Posts: 280
Joined: Sat Aug 25, 2012 4:23 am

Re: VII. Military

Post by Veneteaou »

Separate from all this we need a damage model. Even the original Imperialism had one which we just don't know.
There are the unit stats from the manual/quick-reference cards available on GOG.com that we could use as a starting point. I'm attaching a spreadsheet with the information. From there, we have a couple of things we'd need to test to know for sure:

- Total HP of units (IIRC this is a fixed number for all of them, somewhere in the 50/100/150 range)
- How the experience multiplier works (Based on gameplay I'd say that full 4-medal units deal double the damage)
- How morale works (this is getting complicated)
- What if any random number ranges are applied into the calculations.

Honestly, I've always considered battle systems to be a pain in the ass to build without testing extensively. Instead, I recommend we build the mechanical functionalities we want to see (morale, experience, etc.) and then test the combat system until we are pleased with it. The easiest way to do that would be to create a simple 2-player battle system where each team has the same units on a bland terrain so they can just go at. Another great alternative is to build a quick battle system like Star Control had where players could pick units and then fight each other.
Attachments
ICS.csv
(6.19 KiB) Downloaded 474 times
Xylander
Posts: 37
Joined: Wed Mar 06, 2013 5:32 pm
Location: Germany
Contact:

Re: VII. Military

Post by Xylander »

Making a working tactical battle system is equally costly as making the rest of the game. Furthermore it is independent of the rest - it's a game in the game. Making war is not just moving units around in a tactical batlle. You have to decide which units to build, where to move them, how many to use for an attack or defense, where a frontline is, where you have to push and where to fall back. This is making war, at least for me.

When time comes to combat resolution and one side has twice the numbers and equal tech level I expect this side to win (even if it's my oppenent). I don't like the fact that I can trick the AI (and I haven't fought against an AI I couldn't trick yet). It's not that I could win every battle against an AI opponent. But I usually can cause severly more damage than I take and I'm much better in keeping important (read: expensive, experienced) units alive. It makes winning a war easy. And games (including Imperialism) compensate this with 'total war'. Usually all AI oppenents will declare war on you at some point. A country should be dead when this happens but instead it forces you to do battles manually or to use save-and-reload (e.g. to find out where the next attack will happen).

Of course, you can play on easy or medium difficulty but then the game is not challenging. And if you play on hard or harder it's unfair and you start to exploit weaknesses in game mechanic or AI tactics. That is not the gaming experience I am looking for. Even great games like Total War or Victoria 2 fail completely with tactical AI and allow players to cause 10 times the damage they take and beat super powers with little nations...

Now if you say tactical battle will be integral part of the game - how will you handle the problems I described?
These are the problems:
1. Fighting a war against a single AI oppenent is easy going even if the oppenent is much stronger than you. You simply have to chose the battles you'll fight, win them and cause big losses on the other side. If you can't do so, retreat (which will prevent losses on your side). You build up an experienced core army the AI can resist even less. Now you destroy the enemy's ability to fight and then conquer his land completely.
2. AI players will have to ally against the player in the end ignoring historical relations completely. Diplomacy feels awkward when you have alliances and really good relations to other great powers who fight each other and then they all start to declare war on you. This is how it works in games like Imperialism or Total War. If the AIs didn't do you would easily win the game and ask for a challenge.

The problem may arise without tactical combat, too. AI isn't much smarter on large scale (where to attack, how to hold a line etc.) but with tactical battles it get's much worse because people like us won't build a sufficient AI in ten years! I understand that you like this feature and find it important! If we integrate it we'll end with the well known All-AI-vs-player thing that feels so unfair if you lose and so unrealistic if you win ('hey I've beaten an alliance of GB, France, Russia and Germany playing Italy'). This is what I believe!
Don't repeat yourself. (DRY)
Keep it simple, stupid. (KISS)
You ain't gonna need it. (YAGNI)
http://www.clean-code-developer.de/
Post Reply