FreeOrion

This is a separate semi-official thread from the 'building the design doc' because it's a little more abstract.

In everything we've done so far, we've looked at a lot of previous products in our on-going quest not to re-invent the wheel and to perfect what's there rather than try to outsmart people who have been doing this for their whole lives.

Along those lines, I have spent entirely too much of the last two weeks playing an enormous game of MOO3. I hadn't touched it in a couple years, and I went back and applied all the user fixes make it more playable, but nothing that would drastically change how it was intended to be by the developers. I kicked myself out of the micromanaging mode that I sort of get in by default when I think of MOO and thought 'I shall only consider problems from the highest level, use what ships my colonies give me, and make decisions only on the macro plane' -- excepting two planets in my home system for the emergency cases of 'OMG i need a carrier fleet now.'

What I found was that the turn-based strategy portion of the game was actually quite fulfilling, even if the the couple spots where it was obvious they'd intended a very elaborate solution only to have the plug pulled at the last minute were readily apparent. Even late-game (300 turns in) I didn't have to spend any time administering colonies; I tweaked a few high-level settings and got, more or less, what I wanted.

The real problems came when I would get involved in space combat. The tactical combat engine of MOO3 is its greatest weakness. I started thinking about the sorts of things I like worrying about in combat engines - things that appear in tactical ones like Total War or Homeworld as well as strategic ones like Hearts of Iron.

And then it occured to me - other than Homeworld, I couldn't think of a single tactical space combat engine that actually involved any strategy whatsoever, and Homeworld's is limited more to what you build rather than how you apply it (e.g. it's great and impressive, but it's difficult enough to control the 3D plane that you end up picking a formation and hoping for the best).

I thought about the elements that make up a good war simulation. In Total War, there's unit quality, experience, morale, general's ability and traits, line of sight, weather, and terrain.

In MOO3 there's ... unit quality.

MOO2 has experience and 'leaders', although their effect is sort of invisible. But this got me thinking: before we blindly go reproducing any sort of MOO elements in the form a tactical engine, what, really, was fun about MOO combat? I think the only thing any MOO game got right combat-wise was that it let you take something you had painstakingly constructed in the strategic mode (ala Hearts of Iron or EU2) and apply it on a battlefield -- it wasn't that the things you applied actually had any strategic difference from each other. MOO3's attempt to accomplish this through the use of task forces and different size ships failed; even the AI knows, at this point in my game, that if it builds a lot of Indirect Fire and Carrier ships, it will beat any Long Range or Short Range armada I throw at it, so I am now routinely losing all LR and SR groups to inferior tech races because they have nothing but missiles and fighters. And in case my LR and SR armadas live long enough to get close, they get toasted (or do the toasting) not with the real guns put on the ships, but by the lightning field generator point defense weapons. To say nothing of the fact that all of the effort put into the art assets of the ships for MOO3 was a wash because you can't see anything.

So answer this question: What is the most fun about a tactical combat game engine in any genre? What about space genres specifically? Should we even be trying for real strategy? Has anybody before us accomplished this, and if so, what were the key ingredients of that accomplishment?

_________________
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!

(for those who wish to skip to the answer, please thread to the bottom)

A great example could be Starfleet Command. Your ships were durable. Big ships (like a Klingon Battlecruiser) had lots of armor. Small ships (like a Starfleet Frigate) were much harder to hit and had better speed. You had tactics because it wasn't a popcorn fest - ships weren't exploding left, right and centre. I pulled quite a lot of hats out of the rabbit by doing a lot of hit-and-run tactics destined slowly disable huge battlecruisers. Sometimes when severely outnumbered (4 battlecruisers versus 2 frigates) I'd self-destruct one in the middle of the BC fleet, which (if I was lucky) would cripple my enemy for a short while. I would try to capture and sometimes succeeded. Othertimes, I'd try to pladce mines and have my shuttlecraft harass. It really helped if I had an Elite crew as well.


- - - -


Another game that I think does things well is C&C Zero Hour. You have units that capture other enemy units, disable enemy units, can kill enemy units instantly, are ineffective versus other units, etc. It is a lot more fast-paced than Starcraft, and really gives mean to attacking by the flanks (or the rear) because your enemy often can't react fast enough if they are fighing from the front. It does suffer from the "what you build" scenario though, because the paper-stone-scissors aspect of that game is very strong.


- - - -

That said, I think what works to the benefit of tactics is...

1. Staying power of ships to regular weapons
2. Vulnerability of ships to very special / unique weapons (capture, terrorist attack, disable, etc).
3. Ability to create special / unique weapons that WORK (capture pods, Torpedo artillery, Cloaked suicide ships, etc)
3. Vulnerabilities due to ship facing, and the ability to exploit this vulnerability. (IE: It's no use to have beamships with weak aft shields, but you can't ever get behind a beamship because they pick you off at long-range).
4. Use of terrain. (I know we don't have much, it's in space, but maybe things like missiles and fighters don't fire well if you are facing the sun, nebulas severly reduce fog of war, and asteroids / planeraty rings / comets give you excellent cover to repair)
5. A weak paper-stone-scissors-gattling gun-broccoli approach. (That gives some bonuses to certain damage types, but is not the all and end all of combat.

Oh, and last but not least....

6. Close ups. There's something to be said of watching those behemoth beamships in Freespace 2 duke it out. In that game, they used the "less is more" technique.


Otherwise, it becomes a popcorn fest of mass-producing two types of ships to counter everything.

_________________
There are three kinds of people in this world - those who can count, and those who can't.

One of the factors that make a game more strategic are the abilities or weapons that your ship can use in battle. In starcraft and C&C series abilities and unique units were what set them aside from other RTS games. Unique abilities can be applied to combat whether it is RTS, Turn based or Phased Time.


This was certainly great to have in Moo2. Those Antarens could easily move behind your ships.


Objects like missiles, fighters and mines actually could be counted as terrain because you don't want to get near them.


I hate games that are purely RPS. Usually I just make a random mix of units and then mass attack, so there really is no strategy for me.





Overall I think I would base a combat engine on Moo2s combat engine. There would be improvements such as changing it to phased time, so that all sides have an equaly oppurtunity to attack first. I liked the close in feel of Moo2 and there are certainly good macro controls that could be implemented, eg group attack (like in homeworld where you hold down ctrl and select a group of ships to attack).

First, we shouldn't confuse strategy and tactics. There's not really much significant strategy in a tactical combat engine in which you can't build anything new. You can have general plans like "stay back and let them come to me" or "use my X to protect/kill my/their Y so my/their Z won't/will be vulnerable", but even these are closer to tactics than strategy. Alternatively, you could set yourself more specific goals that are important to the larger picture but might be counterintuitive if you only considered the battle, like wanting to take out a particular ship or installation and not caring if your fleet is destroyed and the rest of theirs is not. Essentially strategy = long term, which isn't really possible in a fixed-composition five-minute battle, other than in an applied sense.

Anyhoo, the ideal of tactical combat is Advance Wars. The really fun part is arranging the different unit types to best complement eachother as a combined arms force. You need a front line of big tanks to protect the rockets which in turn protect the tanks by covering the area in front of them with return fire should anything try to breach the tank-wall, and it's good to have some anti-air nearby to get any bombers that come by, and also some recon on fog-of-war maps to see what's going on. Then if you want to go on the offensive, you need some bombers of your own to knock out their front line of tanks, some fighters to protect the bombers from their fighters, maybe some helicopters to knock out their anti-air missiles without being too expensive to lose.

Another important thing is the rather significant CO powers, which can be exploited to knock holes in someone's defensive line in a sudden push. These really effectively prevent things from stagnating in a situation where it'd be too expensive to be the first person to attack, since both armies have a well-set-up defensive wall that's hard to breach without crippling losses. They also make the use of these powers even more interesting by having the "power meter" for the CO powers charge up whenever your units are damaged (and to a lesser degree when you do damage). This results in situations, at least against the AI, you avoid doing damage or killing units just so you can exploit the way the AI picks when to use its CO powers (at the start or end of any turn if the meter is full essentially). Even more interesting, the turn after you use your CO power, you don't gain any charge on your CO power meter, so it's best to do a lot of damage immediately after someone uses their CO power.

There is no unit experience in AW. There easily could be, but it's not really necessary since there's enough going on to keep you occupied without it, and the units don't persist between maps anyway.

The COs are perhaps quite a bit like leaders, with each having his/her own CO powers and different unit strengths / weaknesses. Some COs are better on certain types of maps than others, as a consequence of their powers and advantages / disadvantages. If FO has combat-relevant leaders, they should definitely have advantages and disadvantages, perhaps with certain kinds of ships.

It might also be good if every fleet has (or works best if it has) some of all kinds of ships, rather than just long range, short range, fighter, missiles, damage tanks, scanners, etc, so that there can be the interesting combined arms stuff going on.

Yeah, exellent game.

I don't want the combat to be "fast paced" or "real-time" or any other of those corporate buzz-words. I rather enjoyed Moo2 system. It did not enforce positioning issue, however, especially towards late game.

I would also prefer LESS, yes you heard right LESS ships that are more durable overall but with destroyable components. In moo2 large battles were rather a chore than enjoyment.

_________________
Staying awake and aware is perhaps the hardest thing to do.

By 'strategic elements' I am referring to things that still occur within the 'tactical engine' -- for some of them (like morale and experience), 'tactical elements' is just as much a misnomer, but yes, I am talking about stuff that matters inside the tactical engine.

I agree that Starfleet Command handles this well, but you only ever have to deal with a handful of ships. We're again at the 'tactical v. strategic' distinction, which is really just another way of talking about scale; when the game is about empires and space opera, then 100 ships facing 100 ships is tactical next to the Empire-level strategy you were dealing with before the combat started. But when you're dealing with 100 ships on each side, the capability and flexibility of one group of 10 ships is a tactical choice, while your formation/Order of Battle is a strategic one.



I don't see that there is any choice other than some form of RPS, because RPS imitates life. Are you suggesting that in a game like Total War, which is still based on RPS, you just make any old combination of any old unit and hope it wins? That's just making a choice by not choosing. I think that if you don't accept that you need to inherently weight some types of units against others, then there is really only one unit type and it's just a question of who can cram more guns onto that one unit type. This is what MOO2 does, and I don't think regurgitating MOO2's combat engine is really a valid choice unless we accept that it's impossible to create a real strategic sim in this manner and that the mission of the combat engine is simply to create a gigantic mud pit to sling everything you've got at the other guy.

_________________
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!

Really, what I said about RPS varies from situation to situation, game to game. Some games RPS works well, eg StarCraft. In StarCraft it has been designed for some units to beat others, etc so StarCraft really has a complex RPS system. Age of Empires might have three RPS elements for each level of strength (didn't play the game for long, so could be wrong), basically along the lines of sword vs spear vs bow. The thing I don't like in this case is that it is based on luck. Sword, Spear or Bow. Which one? Hmmm, Bow. My Bow men get annihilated by the other guys Sword men. But in Age Of Empires I can scout there base and see what they are building, same in StarCraft. If we are to have RPS, we need to be able to scout in FreeOrion.


Even though Moo2 was really a case of cramming more guns onto the biggest ship, there was still an element of RPS I guess. You could have pd weapons on your ships, to deal with missiles and fighters. Fighters and missiles on your ships. Beam weapons on your ships. So, Missiles/Fighters vs PD weapons vs beam weapons. Though it was a faint RPS system.


In any case, I am sure we will have some kind of RPS system in FreeOrion, but I don't want it to be fake, but natural. For example, fake is when I call my ship a carrier, so now it is weak against beam weapons (basically artificially calculated RPS system). Natural is when my ships has some fighters and no beam weapons, so it effectively is weak against beam weapons because those fighters will take a while to get to the enemy, compared to beam weapons which could kill the enemy now.

What is the most fun about a tactical combat game in any Genre?
-The ability to make decisions that influence the outcome, specifically expressed by defeating a more powerful foe through your use of the units.

Space games specifically?
-See above, add in large amounts of explosions and lasers.

Should we even be trying for real strategy?
-Of Course! It will have to be relatively simple, both to assist the AI programmers and to cut down on time required per combat, but it should definately be included.

Has anyone before us ever accomplished this, and if so, how?
-Well, you mentioned Total War yourself. But if you are specifically going for space combat...
Homeworld 2 pulled off some very interesting strategic elements, much of it internal balance (all ships in a fleet had a purpose and could be used, unlike many RTS where many units are just dumb to use except in extremely small amounts), speeds, sensors and stealth, hyperspace, and the third plane. The problem with these elements is they require
A: The ability to build
B: Long battles
Games in the same genre as this... Closest you get is probably MoO 1 or 2, which, as stated already, isn't good enough

The unspoken question: How do we include strategy on the tactical combat map that hasn't already been decided?
Here are the things I think should be a consideration:
Ship doctrines
Limited use systems
Range
Differing sets of attacks/Defenses


Ship doctrines mean this: Ships can be set for a given power/crew distribution. A ship may throw extra power to shields at the cost of weapons, for example. Some devices may add new doctrines, like Cloaking Devices or early versions of the Lightning Field. Ships may change their doctrines, but the change takes time. This adds the additional dimension of "How do I want my ships to act." For those who play Battlefleet Gothic, its basically special orders.

Limited use systems are things that cannot be used constantly over the course of a battle. This would include things like Missiles and Strike craft, but might also include things like 1-use ECM systems (chaff). In addition to things being limited by numbers, they can also be limited by time. Torpedoes in MoO 1/2 are limited examples of this, and other weapons could certainly be used that are similar. However, this is better applied to abilities, say a limited use Sub-Space teleporter or Warp defense field that can only be raised for 2 rounds out of every 8. Basically numbers based systems, where the numbers regenerate slower than you use it. The requirement for this is it cannot be always better to use a Limited use system the instant it is ready. This can also be tied to doctrines, as power might be devoted to quickly powering up special weapons, or crew might be used to fire those missiles even faster off their racks.

Range is a key factor that is often overlooked. Many people consider the maximum range of weapons, but what is frequently ignored is that not everything within maximum range is equal. If I have Tachyon cannon, I should be able to fire at unlimited range. Of course, the odds of me actually hitting anything is about equivalent to the odds of me currently throwing a dart and hitting an insect. So then engagement range becomes a question. Do you want to keep your beam barges out of harms way, even though it will result in far less damage? Couple this with limited use weapons: do I want to pepper the enemy from long range with missiles, where most will get shot down, or do I want to close to point-blanke and launch them into the enemies teeth?

Finally is the differing attacks and defenses. This is kind of a RPS system, but not quite. Different attacks should be easier to repulse/avoid with differing systems. Dodging a lightspeed beam is going to be hard at close ranges, so energy shields are your best defense. A projectile cannon will only be slightly affected by shields, but they can certainly be dodged. Missiles will just laugh at your manuverability, but they can be shot down, and fooled by ECM. Fighters can outmanuver you, slip through shields, and ignore most ECM. But they can still be shot, and odds are there aren't going to be nearly as many Strike craft as members of the other weapon categories. Size should also matter, small ships being more evasive, but obviously less able to mount other defenses. You can't change the kind of defenses you have (though you might use a doctrine to augment some of them), but you can try to outmanuver the enemy to minimize your defenses weaknesses, and strike at theirs. This should also foster combined arms.

Well, I'm going to stop ranting, as I think I started this post a half hour ago....

I think my favorite 4X combat system has to be BoTF. It's just so damn simple. You can choose between basic approaches like charging and strafing, but that's about it. The entire battle usually takes around a minute. With all that simplicity, it's still pretty fun to watch the battle take place. I would like to see more RPS elements, but use a basic interface like BoTF's. I agree with the post about MOO2 combat being tedious with more than a few ships, but I'd like to see much larger battles. Though you should be able to select groups of arbitrary size and give them simple commands, we should have no facing, targetted firing, boarding, power level adjustment, etc. You're playing an emperor, not a ship's XO.

I agree with tzlaine (I think) -- I don't want to be dealing with minutae on individual ship levels, but I think we should emulate Total War here. We need to be able to group like ships so that if you click on a group of ships, they're all the same kind and they all follow the same order you give them. This is a slightly different discussion though, and not something I expect to solve in the first draft of the combat document.

_________________
Surprise and Terror! I am greeted by the smooth and hostile face of our old enemy, the Hootmans! No... the Huge-glands, no, I remember, the Hunams!

For every offensive move, there should be a defense move.

This is usually the key that makes game's like Street Fighter II fun to play. It doesn't just come down to using one certain attack all the time (although it did for quite a few fireball-hurling button mashers at the arcade. I know)

The strategy is therefore in the fleet composition, not the outworking of the battle.

The battle itself is tactical fun, but the strategy is making sure your vulnerable points are covered and that your enemy's vulnerabilities are exploited.

So if you want strategy during combat, you need to be able to address fleet vulnerabilities on the fly. That means that you need to be able to issue generalized orders like "target all enemy bombers" or "defend allied fighter-carriers". "Shoot this guy with such-and-such gun," that is, the "MOO approach", is getting a bit too specific if you want large battles.

I'm not sure what you've got planned, but i'm thinking that if you really want true strategy, it's going to be a game of setting priorities for targets and defense and setting general behaviors and letting the AI do the rest.

Facing is good to have. Maybe ships could adjust their own facings based on whether they are attacking in a direction they are not facing.


Have to decide on targeting, but do you have to select your ship, then select the weapon, then the enemy.

Can't you just select the enemy(s), choose possible attacks/abilities that can be used. The Ai will then decide which of your ships can use their ability/weapon on the enemy.


the ship could do this automatically. Maybe when designing the ship, you could set out various AI decisions, ie always try and board when in range.


not really needed, but if it was make it so that the ship adjusts its power based on the AI decisions set out during design, eg convert power to shields when hp low/weapons ineffective.

In ship design we could have some kind of condition=>event system where you can create 'five' rules that determine the AI behaviour for a ship. If you have a limit to the amount of rules, it would be simpler and would be more challenging (no fun if you were aloud as many rules as you would like).



While it would be useful to have a 'shoot all enemy _____', you would have to cater to each and every type of target, eg carriers, bombers. What if the player wants to shoot all enemy ships with low shields or low armor? There may be too many types, but I guess it is always good to have a minimum. Better to be able to 'attack all carriers and bombers', then not being able to 'attack all anything' at all.


Cutting out this level of choice would be bad. What needs to be done is to make it so that the choices are very easy to implement.

I seem to remember writing quite a bit of stuff about this over on OS. I was trying to figure out exactly this - ways to get total war style tactics into a space combat engine. The conclusion i came to was along the lines of: you need to add more stuff to the mix other than just a bunch of spaceships which shoot at each other, i.e. you need to have strategic objectives other than just 'blow up all the enemy ships'. There are several things which can be incorporated into space combat to try to achieve this, e.g:

Use of Terrain: this is essential in the total war games (though it has less of an effect in rome:total war, to that game's detriment). Its more difficult to generate terrain effects in space, but we could use things like Nebulae, Asteroid Fields, Planets, Moons - if these features conferred advantages or weaknesses to ships in those locations, then you immediately have tactical objectives along the lines of 'hold the high ground'. Also, varying terrain features from battle to battle keep things interesting, and can force players to vary their tactics.

Scale of Combat Arena: If you make the combat arena the entire system, rather than just around one planet (as in the moo games), you open up lots of possibilities. If a system has 2 or more key planets, do you split your forces and try to attack/defend them both at the same time, or keep your force as a coherent whole?

Balancing Detection and Weapon Ranges, and stealth ships - not knowing where every enemy ship is at the start of combat can make things interesting - it forces you to send out scouts, and maybe to keep some defence back to protect your vulnerable carriers. Its important to balance this, however, so that every combat doesnt end up with 2 fleets of stealth ships wandering around blindly looking for each other. Its also important, IMO, that ships cant hit each other from opposite ends of the arena. Force ships to manouevre for position.

Hit and Run - allow players to perform actions which dont necessarily involve defeating the enemy fleet. A good example of this would involve hitting an enemy planet, either bombarding it, or landing ground troops, and then running away. This might encourage interesting ship design - e.g. fast, stealthy troop transports, and interesting tactics - i.e. have a small force to distract the defensive fleet and draw them away so that your stealth bombers can sneak around the back.

The other thing i thought of to try to improve space combat tactics was Shield Facings, which is discussed here.

_________________
The COW Project : You have a spy in your midst.

Might i suggest that we utilize a system of dynamically generated buttons that float over the right or left side of the screen. Depending on what types of ships (or even designs if we limit the #) in the combat, you would have a button to represnt each one. With one click you could select all carriers or all missle ships, etc... If you don't have any carriers, there would not be a carrier button generated.

This could allow qucik grouping without slowing the pace of a battle & even allow selecting ships w/o knowing where exactly they are in a large task force.

I know this is sligtly off the topic at hand, but I wanted to put it out there.

TO answer your general question, something similar to Total War would be excellent. Even a slightly more interactive version of BoTF would be nice.

You can't catagorize everything (easily), but you would certainly have to boil it down to a few certain option. The player would be forced to use those option, but so would the AI, so there is no disadvantage for anyone.

But on top of that, it might be usefull to specifically target a ship or squadron as a particular threat, possibly giving them a "priority number"