Space combat: next steps

Past public reviews and discussions.
Space Krill
Posts: 8
Joined: Mon Jul 11, 2005 7:01 am
Location: Canberra

#16 Post by SmellyTerror »

Other stufff, in a separate post becuase it's got nothing to do with the other thing:

1. What kind of scale are you looking at? I was thinking that, if combat was to be on, say, a system-wide scale, then perhaps the combat map could actually be used in the strategy part of the game - place minefields, build stuff (refineries, shipyards) in asteroid fields, etc. That could make the battles pretty interesting, because the system's infrastructure would also be part of the terrain.

I guess the problem would be that, to maintain any kind of scale and/or complexity, the map would have to be pretty damn big. How long do you want battles to go for?

2. Is the time system set in stone? That is (or so I gather) fairly slow real-time, no (?) ability to pause, but no real need either. Simulated real-time meaning turns are happening every few seconds. Is that right, and is it still up for discussion? I like it, I just want to see if I should be addressing points raised.

3. The thing I liked most about MOO2 was that ships could be purpose designed for special tasks. They weren't just attack/range, defence, and move values - there were extra things they could do, like boarding, tractoring, shield breaking, getting behind, etc. I guess the biggest problem I have with task-forces in tactical combat is that it makes stuff like that difficult. That's an influence for the mothership stuff above, of course, but even on a larger scale I think it's important to maintain the range of utility that MOO2 ships could show.

MOO3 combat felt souless (I think) because, first, ships pretty much just tried to stay at optimal range for their weapons, making them nothing more than stat packages, and secondly, because you had so many that you didn't care what happened to them.

4. I really like the ideas about pre-defined behaviour. Target priorities and "role" presets would make the simulated-real-time work much more smoothly, and reward players who plan well.

What I'd most like to see is some interface for the tactical AI. Perhaps a vastly simplified version for the casual player to pre-set combat behaviour for their own ships, but it'd be sweet to extend the same concept to permit players (modders maybe) to significantly alter the AI's behaviour in battle, to improve the way it operates. We could maybe see players making better AIs to beat each other as a variation on multi-play. :)

Space Krill
Posts: 1
Joined: Thu Jun 29, 2006 10:37 pm

#17 Post by mjb »

I think space combat should be quiet complex so that the player with the disadvantage still has the chance to win the battle.
large battle fields, the whole solar system including planets, moons, etc. would allow for a lot possibilities
hiding behind planets etc. would allow for surprise attacks and so on
formations could help minimize damage, if one ship is damaged, it falls back and is shielded from the enemy by the other ships
in terms of numbers I think battles should be between 1-50 ships

Master Of Puppets
Space Krill
Posts: 1
Joined: Sun Sep 10, 2006 10:26 am

Space combat

#18 Post by Master Of Puppets »

Hello out there!

Great project - I really appreciate your work although I'm afraid of not beeing able to play or even contribute because my "job" of beeing a student of mechanical engeneering takes most of my time and energy.

But nevertheless I would like to try explain my understanding of good space combat just to give some feedback to the ongoing discussion. I've played "MoO3" and "Space Empires 4" quite a lot..

To the point:

I always disliked Moo3s unpersonalised point of view: things I like to see are
1.) individually named ships; they get a auto-generated number upon construction but the player should be able to rename it. Its just cool to have your flagship named according to your favorite TV-series, movie or grand mother
2.) the orientation of ships should matter when hit and when firing
3.) destroyable components (at a grand-scale): fighters get a whole new role when they are ordered to take down the shield generators
4.) heavy influence of crew quality

Point 2), 3) and 4) are heavily related to ship design: when designing you should be able to place the weapons and some very important - destroyable - components (Bridge, Shield generator, Hangar, Engines) How these things can be arranged should depend on the ships hullsize and shape. I think it all is best explained by example:

Ok I want a all purpose destroyer capable of using beam wepons to fight capships and sufficient PD-weapons to defend itself. So I enter the ship design screen and:
- select a hull shape: in this case I've got three diffent shapes made available through research or even racial traits. These are: cylinder, wedge and ball. The cylinder would allow for a low profile making the ship hard to hit from the front and allows easy deployment of one very lage spiral mounted weapon but PD-wepons can't cover the direct front and the rear-section (cause the engies are located there) and other weapons such as turrets are not allowed (at least on a small destroyer-sized cylinder).
The Wedge has similar ailities as the Cylinder but allows for best PD-covering in front and also can support turrets (270° rotateable) The rear section is also mostly unprotected.
The Ball would allow best coverage of alle portions but can't possibly sport a large spiral mount.

- select propulsion systems, shielding and other componets. Every of these critical ones (bridge, shielding, hangars, engines) have to be placed on the hull in some way to determine the direction from which they are vulvernable. It wouldn't make sense to me if the engine are destroyied due to a direct torpedo hit in to front...
- select your weapons and layout the firing arcs. The values of all weapons should be summed up to some kind of overall effectiveness againts different target types (fighters, missiles, capships) at certain ranges.

My example: a wedge shaped ship with one spiral mount, two turrets and 20 PD-cannons distributed evenly. It have is maximum effectiveness against capships direct ahead with reduced effectiveness in the flanks (due to the turrets) and would not be able to fight capships in the rear. The PD-defense would try to disable fighters at quite long range but could get in trouble when catchted from behind. Additionally there would be a small hangar for recon-fighters included some sensor-systems and the bridge placed roughly in the middle of the hull.
By the way: it's someway inspired by the Star Desroyers seen in Star Wars but that doesn't matter that much because it's all about the general principle of "layout and shape determine effectiveness in certain angles of approach".

The next thing related to ship design is weapon characteristics. I'm not a friend of "rock paper scissor" (RPS) that much cause it simplifies things often too much and makes for odd results. In fact there is some kind of RPS out there but there are better ways of implementing it than just say "fighter against bomber, frigate against fighter, bomber against capship and so on. I would prefer the following approach:
- weapons have a prenetration value: you can't take out a capship's amor with fighters but their tropedos and missiles can do.
- weapon systems only able to aim for targets of a certain minimum size. This way the heavy spiral mounts wouldn't be too effecive against the frigates in the picket ring and one is forced to use small hull sizes troughout the game.
- the accuracy should be heavyly dependant on crew and sensor quality

In the end this is "rock paper scissor" but in a much more natural fassion I think.

5.) let the AI do the job according to your general orders

When dealing with a group of 20 ships one can easyly manage a battle alone. But if there are 500 ships around things become weird. To avoid the reduction of the player to an mere "viewer" we should have a high degree of automation but with the ability to jump into the action at any point. Again an exaple:
- I've got some recon-fighters on-board of my destroyer so there should be an order to launch these an spy around; related to this would be the order to spy out a certain location such as the region behind that moon.
- there are fighter-bomber approaching directly ahead. I would order to lay barrage fire on them instead of individually all PD-cannons
- the enemy capship is ahead: I can order to fire at certain components or just attack at will (let the captain decide which section to concentrate fire on) (same for fighters)
- same principle for special actions such as boarding. When there are boarding parties ready there should be a command to board an enemy ship.

All this applies to TFs but "single ship" TFs should be possible to.

6.) It should be possible to move ships around between TFs

7.) It should be posible to move the individual ships around in their TF.

8.) The formations of all TFs should be user-definable. This goes hand in hand with the concept of ship design above.

9.) I prefer simulated realtime with an option to decrease turn-time to an point where the action is "close to realtime" cause this way you have the chance to see some really nice fireworks. The ability to pause and give orders meanwhile is a must.

10.) The combat area should inclde the whole system but space is big and therefore getting the scale of things right is tricky. Maybe some kind of "fast SLT in-system travel" has to be included. While traveling weapons and shields are down so can't fight (but die of course) but distances between in-system objects could be propery scaled which would add to the atmosphere. Obstacles such as asteroids and mines would of course prevent the use of this travel mode.

11.) Balancing should not allow for too many big capships. Destroyer, Cruiser these should be the main battle forces. Only a few battleships and only one or two really big commandships per empire.

12.) Include some kind of sensors that roughly give directions where the enemy is. Lets say a marker on a strategic mini-map.

So that's it. Sorry for the lenghtly explanations but english isn't my native language. Additionally I didn't come across a game roughly similar to the ideas above.

with best regards

Large Juggernaut
Posts: 938
Joined: Fri Jan 20, 2006 9:34 pm
Location: GA

#19 Post by marhawkman »

I LOVE the idea about ship shapes. That would be a REALLY cool thing to implement.
Computer programming is fun.

Space Floater
Posts: 26
Joined: Wed Aug 30, 2006 7:56 pm

I like the idea of the ship shapes as well

#20 Post by Airshipjones »

I had thought of something like that as well, but I was thinking of it more from the propulsion and detection/defense capabilities. Ships with a narrow design will be harder to detect and hit, but will not use space as effeciently. The spherical ships will have the maximum space (great for dreadnoughts or carriers and supply ships) but easy to detect/hit. The wedge shape will be a useful intermediate, with the interesting limits in firing arcs and moderate detection/hit bonus/limit.

Large Juggernaut
Posts: 938
Joined: Fri Jan 20, 2006 9:34 pm
Location: GA

Re: I like the idea of the ship shapes as well

#21 Post by marhawkman »

Airshipjones wrote:I had thought of something like that as well, but I was thinking of it more from the propulsion and detection/defense capabilities. Ships with a narrow design will be harder to detect and hit, but will not use space as effeciently. The spherical ships will have the maximum space (great for dreadnoughts or carriers and supply ships) but easy to detect/hit. The wedge shape will be a useful intermediate, with the interesting limits in firing arcs and moderate detection/hit bonus/limit.
that works too. I suppose we could make it so some techs fit into certain hulls better. but that'd be last on the list. So far we don't even have a combat thingy.

and just to make it interesting. throw in a hull designed entirely for Stealth. maybe something shaped like a quartz crystal. Maybe like this thing:
Computer programming is fun.

User avatar
Design & Graphics Lead Emeritus
Posts: 3858
Joined: Sat Sep 23, 2006 7:09 pm
Location: USA — midwest

Re: Space combat: next steps

#22 Post by eleazar »

Aquitaine wrote:I think the next important issue is to determine how we manage fleets -- whether we use 'units' of many similar ships (ala Total War) or identical ships (ala Empire at War). Should there be a maximum number of units on the field at one time? A maximum number of ships? Does having a finite number allow for a more effective UI? What are some examples of tactical engines with good UI? Talk about your favorite strategy game, whether it's Total War or EaW or something else entirely. Pinpoint what makes the fighting part of it so great.
I like davebaby's propsal of a potentially infinte number of ships divided into a limited number of "task forces" or "battle groups". It seems much more natural that artificially limited the number of ships, or requiring individual control for a massive armada. That "epic" feel earlier alluded to seems much more sustainable, when like a general you issue your orders to eventually massive battle groups. You can spend you attention on the broad sweep of battle.

This article on gamasutra is applicable. While he uses the Civs as his example, and some of his complaints are spurious, the ideas about battle management on page 2 seem to fit into FO well. And i agree with his basic conclusion at the end, they fewer clicks (or any human/computer interface actions) required to accomplish one meaningful task, the better.

For fun's sake, a battle group:
• should be remebered from the last battle.
• should be recofigurable at the start of every battle
• should be build-able by the AI, if the player doesn't want to bother.
• individual control of ships should not be allowed, to simplify the UI and stop micromanagement.

I am leary about giving battle groups as a whole complex tactical orders, perhaps becuase i have never seen such a system, and can't imagine it being easy to use.

Though i do think battle groups should take orders more complex that
"go here" & "shoot this." Perhaps a list of less than 10 orders, including such things as "use special weapon", "shoot with long range weapons," etc.

Another simple control that would be valuable is the "aggression" of each battle group, as in Homeworld. Sometimes you want a battle group do do as much damage as possible, sometimes you just want it to survive. The fleet would interpret this by giving more or less "power to the shields" and by the percentage of fire devoted to self defense.

Under many circumstances ships should be able to flee the battle, though some sort of "interdictor" tech would be cool.

Finally to keep battle fun, the player should be able to opt out of viewing a battle (at least against AI) at any time, and let the computer autocalculate and display the result. Not every battle can be tactically interesting if the player has sufficiently superior forces.

User avatar
Design & Graphics Lead Emeritus
Posts: 3858
Joined: Sat Sep 23, 2006 7:09 pm
Location: USA — midwest

Re: Space combat: next steps

#23 Post by eleazar »

Aquitaine wrote:...But I figured it's not too fair to all you guys who care about FreeOrion to leave you totally in the dark.

Here's what we know about our space combat engine so far. It's not much, but it took us forever to get this far, and I think it's important.

Combat will be 'simulated real time' ala KOTOR -- that is, it will operate in turns under the hood (of approximately 3-5 seconds) so response to orders will be minimally delayed. Both sides give orders whenever (though they are obviously processed 'on the turn', there is no 'end turn' button).
• I still feel rather "in the dark." How 'bout a brief explanation of why this was chosen, what advantage are foreseen. If brainstorming here is to be of value, we on the outside need to know "why," not just "what".

• What about pausing or auto-pausing? These might be problematic in a multi-player game, but for a new player trying to learn the game in single-player mode, a chance to look around and digest what just happend would be invaluable. I found myself frequently pausing KotOR (a game i like :) ) but especially as a new player.
Baldur's Gate II, (the non console-ified ancestor of KotOR) had somewhat similar combat to KotOR, only it had a stronger tactical component. The game could be set to autopause on a variety of events (chosen in the preferences). This could alert the player when something happened, (like when the current target was eliminated) and a character needed new orders, or when a character became gravely wounded, and probably should have his orders changed.

Space Squid
Posts: 54
Joined: Wed Oct 25, 2006 8:59 pm

#24 Post by SowerCleaver »

I figure the question of tactical combat design can have multiple "right" answers, and it is ultimately a question of personal preference. In that regard, please register my frenzied support for MoO2 type micromanagement battle system and my vehement objection for large-scale macro-simulated battle system.

As many commentators like myself noted, there are many virtues to MoO2 style micromanagement combat system. Just to name a couple: Ship design is one of the fun part of space combat sim game that is not well-resprented in historical or fantasy-inspired sim games. After spending quite some time placing different type of weapons and special devices into a ship, I don't want to see 183 of so-painfully-designed ships vaporized in two seconds after the start of a battle. Also, giving vague orders or setting up algorithmic instructions to AI so that it can carry out space battles for me sounds closer to a movie-watching experience than a gaming experience. In a space sim combat game, I want to exercise tactical judgment and beat AI's larger force with cleverly manuevered smaller force.

Obviously, I'm not suggesting we follow MoO2 system in every detail. For example, I have no qualms with the Phased Turn approach (is this the same thing as Simulated Real Time? I couldn't understand what SRT is from Aq's post). I also acknowledge MoO2 system forced user to suffer extremely long and tedious battle if more than a handful of ships were involved. However, this can easily be overcome by UI design - e.g. enabling the user to issue commands to a group of ships with the same design (or alternatively "fleet" or "battle group" consisting of different design).

I noticed that sizable group of people actually prefer macro-scale simulation type combats (e.g. "direct user control of combat is boring"). I respect that perspective and totally understand that there may be awesom space combat game based on such design. However, direct user control of combat is very fun for me, and I would assume many MoO2 fans would agree with me. MoO3 or GalCiv2 didn't really work for me because of the combat system. I just hope that FreeOrion would follow the glorious MoO2 tradition in this regard.

Large Juggernaut
Posts: 938
Joined: Fri Jan 20, 2006 9:34 pm
Location: GA

#25 Post by marhawkman »

hehe.... I've taken out an enemy fleet of 20+ doomstars with a fleet of (around)10 doomstars. It took about 30 minutes.
Computer programming is fun.

User avatar
Creative Contributor
Posts: 643
Joined: Mon Sep 01, 2003 2:17 am

#26 Post by skdiw »

What if you had just a few big, awesome ships? Ships you knew the names of, which you designed from the ground up, which you take the time to repair, refit and upgrade as tech becomes available? What if you had emotional investment in your ship?
wonder unique ships have been brought up before. FO isn't there yet. i believe this thread is just sorting out and generating ideas for tactical combat.

one way to satify more people is maybe develop two independent routes so the player can chose where to invest his resources. for example, one player may chose to invest large amount of identical ships and another is more focus on developing large unique ships. the game could be balanced that way. for example, galciv 2 has logistical points to make a fleet where each ship size takes some number of logistical points to limit the size of a fleet. the larger the ship size, the smaller amount of ships in a fleet. the game is balanced in an excellent way so that small ship fleet isn't inferior to large ship. by researching in logistical technology, that in effect is similar to one player researching better hulls in terms of tech edge. so player A who likes large quantity of ships and likes to manuever different ship roles to target shield generators on large ships can do so, while player B can be happily chose where to fire his massive black hole generators from his few capital ships. path B maybe more micro-intensive to satify tatical players and path A is more developing types of forces from macro level then point-and-shoot-then-watch at tactical level to satisfy macro players. both is fun so i might do a combination of the two. just an idea. i believe we can do both and still have FO be streamlined.

note: i support ship "class" rather than ship "size," even though i use the word size a lot because more people are familiar with size than class.

Krill Swarm
Posts: 11
Joined: Mon Dec 19, 2005 1:57 pm

#27 Post by zobo »

There is something to be said about having few ships to get the interest of the player. If we have large fleets of ships most of them can never have any emotional involvement from the player - I'd say we could have about 7 ships per empire that are special. The number does not have any real in-game reason, it's just something to do with human cognition.

Even if your fleets consist of huge battlewagons it's rather easy to imagine your empire producing a lot of them. And with numbers goes the identification.

And a point to consider is.. I really want to see the massive apocalyptic battle between massive allied fleets on both sides, thousands of ships in all.

And, on my fleet, I have the Battlegroup Revenge, commanded by Admiral Bsqraq from his Dreadnaught Revenge, the flagship of my empire.

I have the Battlegroup Aurora, headed by similarily named Supercarrier.

Thus, your empire has several ships that are special, that act as command ships in fleets and get certain bonuses. When one of them falls you can designate (or build) a replacement, but the number of those you have is strictly limited, perhaps it could be a square root of your colonies or even a cubic root. Perhaps something adjusted to galaxy size.

What makes them special? Um. Possibly the command staff?

Usually these flagships are among the last ships lost when the battlegroup gets damaged, but not always. Destroying the flagship could be a tactical element in a battle, difficult to pull off but potentially having a big pay-off.

The ships should also mostly give bonuses to the actions that shiptype is good at - so you'd get supercarrier in charge of carrier force, a dreadnaught leading the battleships and a stealthy cruiser leading a fast attack group.

User avatar
Cosmic Dragon
Posts: 2175
Joined: Fri Jun 27, 2003 12:37 am
Location: Auckland, New Zealand

#28 Post by utilae »

zobo wrote: Usually these flagships are among the last ships lost when the battlegroup gets damaged, but not always. Destroying the flagship could be a tactical element in a battle, difficult to pull off but potentially having a big pay-off.
Unfortunately the classic hero unit always has the problem of being the first to die. As players will seek to kill the most powerful units first.

Krill Swarm
Posts: 11
Joined: Mon Dec 19, 2005 1:57 pm

#29 Post by zobo »

Obvisously it is, if your targetting is done MOO2 style.

It would have to be something else.

If we can make the combat system have targetting hero ships be really a significant choice, one that will lead to bigger casualties on your side, and probably cause you to loose this combat we have just made a nice strategic choice.

Obviously, this only works if we have less individual ship control.