Behind The Scenes: Iterative Combat Design In TLF

[Wingflier]

So out of curiosity, will there be any advantages to actually playing the battles out in the "action phase" of the game?

I ask because it would seem rather silly if the player could simply auto-resolve every battle scenario and come to equal or greater results than if they had decided to do the whole thing by hand. I would hope that, just as in the Total War series, a player's own personal contribution to the fight would yield significantly greater results on the whole than any kind of auto-resolve feature. Otherwise, wouldn't it make the action kind of fruitless?

The catch to that, though, is that if you do that course of action, playing it out is the preferred choice of action. So it really isn't a choice on higher difficulties / if you are testing your skills to the limit. Only if you make the outcomes equal in the end is it really a choice. For example, when I played total war, any meaningful battle I had to play, even if I won my units were wrecked so I had to play 90% of my battles even if I didn't really want to.

The solution to this is rather simple: Make two different difficulties for the game. A difficulty for the large-scale TBS portion of it, and the small-scale "action" part of it. Since these are basically two games combined into one, and since each player may have differing skillsets making them better at one than the other, it only makes sense that they would not be combined into a single difficulty.

In this way, those players who don't want to "fool" with the action aspect of the game, and who would rather just concentrate on the large-scale strategy part of it, can set the "action" difficulty to 'easy', and thus auto-resolve battles with positive results most the time.

As I said, this is probably the only way you're going to please both parties. If there is no logical reason to undertake the missions yourself, it makes no sense to do them.

But interest/skill mismatches aren't really something that the game dev can specifically eliminate, I think.

As I mentioned, it's easily handled. AI War does this in spades, allowing players to intricately tweak the difficulty to their own strengths and weaknesses. I can't see why this should be any different. Pigeonholing people after building such a dynamic combination of genres would be a horrible, horrible mistake.

[PokerChen]

 I do have lingering doubts about the transition to action. Space ship action generally means, to me, that you have multiple weapons firing away, grant both a form of eye-candy and tactical decision making. AI War and SoaSE are the sole exceptions in the games I play, where most ships have a single weapon and nothing else. I prefer schumps like Tyrian over Touhou, while games like Ikaruga come close the single-weapon spec, but are interesting when it comes to their perfectionist mechanisms, e.g., of absorbing all damage of one colour at any time and creating a super-weapon out of it. I can cite the embryonic Fraxy as another game that has single-ship weapon switch, but that uses the enemy boss itself as structure.

 There's this other identity problem in the action phase itself: it's not as deep as a (good) SCHMUP, but not as shallow as Asteroids. "Much as I love the action-oriented combat here in TLF, sometimes I just want to get on with it." describes too well the long-term appeal of the current action model. It's fine for a while, and then I uncover the limitations - it gets stale. The action phase has magnitudes less interest-lifespan than the solar phase (randomisation, depth, surprise, perfectionism) and I can't shunt all of that int polishing.

 Happy feedback is, you're not different from Total War in this respect. You're not going to be getting two games out of TLF, although you don't have to and can do with 1.5. You have an inverse Gratuitous Space Battles situation (fun experience, but still half the experience and says so on the tin, IMO), in which there is barely anything to plan and it's all about in-game execution and decision making.

 There a whole list of things that I suppose could be implemented for release. I'll throw it into Mantis.

So out of curiosity, will there be any advantages to actually playing the battles out in the "action phase" of the game?

I ask because it would seem rather silly if the player could simply auto-resolve every battle scenario and come to equal or greater results than if they had decided to do the whole thing by hand. I would hope that, just as in the Total War series, a player's own personal contribution to the fight would yield significantly greater results on the whole than any kind of auto-resolve feature. Otherwise, wouldn't it make the action kind of fruitless?

At the moment, one expected trade off is damage to your ship, which is designed to be not all that easy to repair (was still fairly easy in 0.808). You can avoid this in actual battle.

Getting differing amounts of BP from each mode would be problematic, IMO. So would, e.g., not getting all the bribe items in convoy raids...

[doctorfrog]

I like the informative post, thanks. Perhaps like Cyborg, I'm also not real fond of the arena shooter (Smash TV) aspect of the game. I keep wishing it were more like a squad-based pauseable realtime tactical type deal, like the Baldur's Gate series was. Or at least I could be able to pause, designate targets, and not have to hold target cursor everywhere. Or maybe have some tactical decisions available, like considering firing arcs (instead of unidirectional weapons) and weak spots (instead of flat hit points) on ships.

Ah, but this is why alphas are alphas, and we're all trying to figure out which wish-projection is going to result in the better product. I'm glad to see the game take an unexpected direction, and I look forward to the changes, and a more easily understood strategy layer.

I've got big hopes for this one, and I hope I can play a small part in giving useful feedback. For now, I'm going to wait for that next revision and see how it grabs me.

[Misery]

There's this other identity problem in the action phase itself: it's not as deep as a (good) SCHMUP, but not as shallow as Asteroids. "Much as I love the action-oriented combat here in TLF, sometimes I just want to get on with it." describes too well the long-term appeal of the current action model. It's fine for a while, and then I uncover the limitations - it gets stale. The action phase has magnitudes less interest-lifespan than the solar phase (randomisation, depth, surprise, perfectionism) and I can't shunt all of that int polishing.

To some degree, this bit isnt so much a problem with this game on it's own, but is an issue with the shmup genre as a whole.  One way or another, the entire genre is full of extreme repetition.  Your typical shmup, of any type, is often about 30 minutes long, from start to finish.  That's it.  But it can take you a LONG time... really really long.... to actually manage to beat the game.  You generally have to play each individual level about a billionty times in the harder games in order to make real progress.  Over and over and over and over and over.  Even moreso if you're playing for score, particularly with those games where the scoring is complicated as all hell.  The time investment is really extreme, yet the levels NEVER CHANGE.  Always EXACTLY the same, each time.

It ends up depending strongly on the player, wether or not they mind the repetition, and the problem is that this one small factor can decide wether or not a given player buys the game or passes on it.  A pretty major division.

In this game, it gets away with it somewhat by the fact that the shmup bits are not the ENTIRE game, but instead serve to further the main mode's purposes.  Yet still, that division is there.  No simple way to fix that.

Misery can probably speak to this better, as he's super into the bullet hell and SHMUP genre in general

True, though I didn't think I'd jabbered about the genre THAT much.... or maybe I have.... not sure.

The goal of the combat here is to be fun, and a diversion where you make interesting meaningful choices... but it's not remotely the bulk of the game.  Whereas, as I noted above, with a true bullet hell game that's all there is.  It's kind of like Quake II or Unreal Tournament or something versus DayZ.  Yeah, they both involve guns, but with the former two they are all about precision execution and never getting shot and always making your shots, etc, while moving as fast as possible.  Whereas with DayZ those skills help for sure, but there is other stuff going on as well.  From my understanding, not having played DayZ but having played the former two a ton ton ton (and those obviously being ancient examples).

Yep, that's a good way to put it.  Of course, potential players wont always think of it this way, which can be an issue when it comes to deciding on wether to buy or not, but still. 


One thing that might help is enhancing the fighting between AI ships.  Right now, an AI ship fighting another AI ship basically is like two old guys standing there ineffectually flailing popcorn at each other.  Neither seems to ever do any true damage to the other in many cases, there's never a point where you're like "Uh oh, I better go help this dude over here, he's getting pretty wrecked by that pirate there".  You have plenty of time to take out all targets without having to worry about that all that much.  Not much thinking involved here, as a rule, or decision making.

Some of this is balance issues, of which the combat side has many... for example, a bullet that does 127 damage doesnt mean too much when it's being shot at a ship with 120,000 health.  The ship doing the firing would have to be shooting these at a pretty absurd rate to be a threat.

But anyway, if that bit could be enhanced without dumbing down the difficulty (due to you fighting against an enemy with an actually effective ally helping you too much too often) that'd probably help alot.

[PokerChen]

 Indeed, but the very act of incorporating a SCHMUP-like into a strategy/simulation game should be justified by plugging the solar information usefully into the action phase. Every battle is a substage in SCHMUP, which can last all of 30 seconds, but you could be fighting 100ss of them in 1 strategy game.

 Misery,you play way too much SCHMUP. I don't believe TLF benefits in the end from having the same stage over and over, in the context of an ever-changing strategy layer. I'm fighting all the time, in various contexts against various opponents, but the 0.808 alpha doesn't have the variations that should be incorporated from the strategic layer. Everything boiling down into the same situation that the same loadout will work with, the same approach will work with. You don't need the mercenary outpost more than a few times. Once to get an energy blaster (anti-shield weapon) if you don't start with it. Once to get an anti-hull weapon that works against all races. That's it. There's no particular reason to get the other ones, except Gravity Missiles, which can take down 10 non-resistant flagships in the starting 30 seconds of the game. The only time the game fails you is for taking the default weapons against Burlusts.

 I would much rather each combat mission taking a time cost of 0.2 months (not 1 month), and give you much more freedom to retreat, regear (should takes a bit of time too if you did end up changing equipments), and take your grudge match. That is more mercenary. That's more pirate.
 I would much rather a situation where opposing flagships have many possible loadouts, and opposing fleets have builds and strategies, where they could choose as a fleet to stand-off and lob missiles and sniper rounds at you, or they could mob you with fighters at the beginning of battle, etc., and that you don't know which one it's going to be if you haven't spied or scouted upon them beforehand. Let the mission fail you for not taking at least one of gravity disablers and missile disablers, because every ship in the fleet focus on those two weapons.

 The action phase can be improved by many things, but as a schmup-like it lack by default the granularity of a space-combat sim. It's a difficult way to fix the division.
Chris may implement a command-room decision to disable the engines for your enemy freighters, but this can't translate into battle without giving ships subsystems. I can't disable the outpost without destroying it (not doable), or what if I just want to raid it for supplies (doable)? (Suppose I only want to hobble this race without creating -10000 relations.)

[nas1m]

Bullet hell is not how I would categorize it.  Yes, there are lots of bullets, and yes, you fire guns.  But you're not going for score or combos, this isn't the entire game, and it's a matter of taking tactical objectives via those mechanics, not just executing maneuvers as optimally as possible within those mechanics.  Aka, you're trying to tactically position yourself the best you can, and make approaches from good angles, etc, and hit key targets.  It's almost like a combination of a team FPS game and a SHMUP.

Typically my experience with SHMUPs is that you have no such decision making, and instead it's all about execution of your play through either randomized or hand-crafted levels.  That's really not the focus here; your actual execution can be a bit squiffy, as long as you are thinking tactically and doing the larger approaches right.  It's still a tactical combat, just with a single ship, direct firing, and some abilities, instead of a fleet of ships doing essentially the same thing.

Misery can probably speak to this better, as he's super into the bullet hell and SHMUP genre in general, and has played this extensively.  The goal of the combat here is to be fun, and a diversion where you make interesting meaningful choices... but it's not remotely the bulk of the game.  Whereas, as I noted above, with a true bullet hell game that's all there is.  It's kind of like Quake II or Unreal Tournament or something versus DayZ.  Yeah, they both involve guns, but with the former two they are all about precision execution and never getting shot and always making your shots, etc, while moving as fast as possible.  Whereas with DayZ those skills help for sure, but there is other stuff going on as well.  From my understanding, not having played DayZ but having played the former two a ton ton ton (and those obviously being ancient examples).

So out of curiosity, will there be any advantages to actually playing the battles out in the "action phase" of the game?

I ask because it would seem rather silly if the player could simply auto-resolve every battle scenario and come to equal or greater results than if they had decided to do the whole thing by hand. I would hope that, just as in the Total War series, a player's own personal contribution to the fight would yield significantly greater results on the whole than any kind of auto-resolve feature. Otherwise, wouldn't it make the action kind of fruitless?

If I get Chris' meaning right he is confident enough in the depth of the solar map game to provide enough "meat" to make the new command room combat the default. The action combat is intended to be a fun diversion, a route to take once in a while if you just not feel like using the command room combat.

The combat of MoO2 was just that as well - a fun diversion. IIRC .

Thus, there really should be no incentive to favor one combat model over the other. It's simply a matter of choosing the diversion you most feel intended to play at a given moment. Since the whole combat seems to be intended to break up the main solar map game (again if I get Chris right) this seems fine to me.

I also feel like the action part already seems to be fleshed out enough at the moment to provide the intended diversion (with the disclaimer of me not yet being in the alpha). I say this because zharmad and others (I think) seem to think otherwise. I think the "error" here might be that they still conceive the (action) combat as a second game (which it indeed seemed to become in the meantime, I have to grant), while it is already in the process of being "reigned in" to be more of a (considerably beefed up) minigame in terms of the design Chris describes (and how I get it).

Just my two cents anyway .
I can't wait to play this .

[Tridus]

The catch to that, though, is that if you do that course of action, playing it out is the preferred choice of action. So it really isn't a choice on higher difficulties / if you are testing your skills to the limit. Only if you make the outcomes equal in the end is it really a choice. For example, when I played total war, any meaningful battle I had to play, even if I won my units were wrecked so I had to play 90% of my battles even if I didn't really want to.

The inverse is also true. If autoresolve gives you the best possible result, reliably, the only two outcomes of doing manual combat are:
1. The same
2. Worse

So that really isn't a choice on higher difficulties / if you are testing your skills to the limit.

They can't be made equal. Equal to whose results are we talking about, anyway? Equal to the best player on the planet, or equal to me? If it's one of those, >90% of the playerbase will get clearly *better* results with autoresolve, and if its the other someone will complain that the battles are "mandatory". And course, if almost everyone is getting better results out of autoresolve and thus not bothering to use manual combat at all, why did all this effort go into its existence?

At the end of the day, if you're only playing half the game then maybe you have to live with the tradeoff of not getting optimal results at the absolute highest difficulty.

[Misery]

Indeed, but the very act of incorporating a SCHMUP-like into a strategy/simulation game should be justified by plugging the solar information usefully into the action phase. Every battle is a substage in SCHMUP, which can last all of 30 seconds, but you could be fighting 100ss of them in 1 strategy game.

Yep, that's the repetition problem.  It could be worse.  That genre usually doesnt just want 100 repetitions out of you for each game, it wants 100 HOURS out of you for each game.  Or more.  Too many more. 

But that's where the alternate thingy should help alot.

Misery,you play way too much SCHMUP. I don't believe TLF benefits in the end from having the same stage over and over, in the context of an ever-changing strategy layer. I'm fighting all the time, in various contexts against various opponents, but the 0.808 alpha doesn't have the variations that should be incorporated from the strategic layer. Everything boiling down into the same situation that the same loadout will work with, the same approach will work with. You don't need the mercenary outpost more than a few times. Once to get an energy blaster (anti-shield weapon) if you don't start with it. Once to get an anti-hull weapon that works against all races. That's it. There's no particular reason to get the other ones, except Gravity Missiles, which can take down 10 non-resistant flagships in the starting 30 seconds of the game. The only time the game fails you is for taking the default weapons against Burlusts.

Alot of this is mostly balancing problems.  Like the fact that Pirate Ravens can be deleted with nothing but energy blasts, which wrecks the shield AND the ship.  Stuff like that, I expect, will be fixed before release.  Chances are I'll constantly complain about it until it is

I would much rather each combat mission taking a time cost of 0.2 months (not 1 month), and give you much more freedom to retreat, regear (should takes a bit of time too if you did end up changing equipments), and take your grudge match. That is more mercenary. That's more pirate.
 I would much rather a situation where opposing flagships have many possible loadouts, and opposing fleets have builds and strategies, where they could choose as a fleet to stand-off and lob missiles and sniper rounds at you, or they could mob you with fighters at the beginning of battle, etc., and that you don't know which one it's going to be if you haven't spied or scouted upon them beforehand. Let the mission fail you for not taking at least one of gravity disablers and missile disablers, because every ship in the fleet focus on those two weapons.

 The action phase can be improved by many things, but as a schmup-like it lack by default the granularity of a space-combat sim. It's a difficult way to fix the division.
Chris may implement a command-room decision to disable the engines for your enemy freighters, but this can't translate into battle without giving ships subsystems. I can't disable the outpost without destroying it (not doable), or what if I just want to raid it for supplies (doable)? (Suppose I only want to hobble this race without creating -10000 relations.)

As far as stuff like fleets coming at you all at once or organizing themselves or whatnot, there's only so much that can be done there with gameplay of this type during combat.  A player with enough dodging/aiming skill will have a dramatic advantage over the AI regardless of much else. Even if the AI all gangs up on them at once, which it already often does.  Heck, I havent lost once yet, on the highest difficulty, with everything focused on me. I simply dont get hit much by anything.  Only problem I run into is that it can take awhile to whittle down groups that are particularly massive just because they need so many hits to be defeated.   Too much AI though might frustrate players that arent so good at the dodging and stuff to begin with.  Hard to say.

Aaaand I've blanked out and forgotten whatever else I was going to say.

[x4000]

Whew, lots of commentary.  I've read all of it, but don't have time to respond in-depth.  I think it's kind of simple, in some respects:

1. When we are doing trailers and such for the game, we can't be emphasizing the action combat too much or people will think that's the meat of it.

2. There is generally no advantage to doing the action combat.  Either way turns out equal, in the main.  Unless your skill level at action combat is very high, I suppose.  But...

3. The action and strategic difficulties are already separate.  And...

4. Your combats are typically not a deciding factor in the overall game.  They are a clutch event here and there, but you are not winning the game based on combat.  If a whole bunch of armadas show up at an allied planet, you at best can whittle them down or buy the local defenders some time.  If that planet gets lost, it's not because you didn't click fast enough -- it's because you let things get into such a situation where a whole bunch of armadas could show up at an under-defended allied planet in the first place.  That said...

5. Even so, sometimes it will come down to a key combat or two, sure.  The combat isn't pointless.  If the forces of the enemy planet are nearly equal but not quite so, you can even out the odds one way or the other.  Or you can be a one-Hydral wrecking crew of key infrastructure, or a per-emptive force, etc, etc.  So combat matters.  But it doesn't matter to the level where being an excellent fast-clicker and always doing the action part to gain a slight advantage, versus instead playing the combat room version, would just bowl over the game for you.

6. In fact, thinking about this, the game should give a slight advantage to using the combat room, and even mention that advantage prior to starting action combat.  That way doing action combat is a risk/reward sort of thing.  It's harder to get the same result as the combat room, but if you play extra well then you can actually go beyond it.  But the risk of death is much much higher.  Which brings me to...

7. Action combat balance.  Right now the individual enemies are more anemic in their weaponry, as Misery noted.  This was to give players a chance to actually do anything, since things move kind of quickly and you don't have full tactical awareness, etc, etc.  And to keep you from getting your own ship killed instantly all the time.  This is one thing that would inherently give an advantage to the command room, because there you have much better tactical awareness and can choose your targets yourself before they get killed, etc.  Which means...

8. All in all, yeah, I'm thinking that most people most of the time would not play the action combat.  But when you do, it's really fun and tense and exciting.  Repetitive?  Less of an issue when your palms are sweating from being in the fight of your life.   That's what most SHMUPs get by on, in my opinion -- the 30-minute playtimes Misery noted are paired with that extreme difficulty and requirement of twitch response.  Not that we'll go THAT extreme here, but right now it's super tame.  Because frankly on the default we didn't want to scare away the grognards.  Or just make it so people were randomly losing in minor combats all the time, which as Cyborg has pointed out before is super demoralizing.

9. All of which is to say, I don't see beefing out the action combat section too much more right now, but I do see really putting a lot of spit and polish into the command room (yes, animations and whatnot as possible, and certainly sound effects).  And I do see really rebalancing the combat side to make it so that, in action mode, you are still "Batman in space," but the guys around you are actually fighting each other with more than whiffle bats.

Thoughts?

[PokerChen]

I also feel like the action part already seems to be fleshed out enough at the moment to provide the intended diversion (with the disclaimer of me not yet being in the alpha). I say this because zharmad and others (I think) seem to think otherwise. I think the "error" here might be that they still conceive the (action) combat as a second game (which it indeed seemed to become in the meantime, I have to grant), while it is already in the process of being "reigned in" to be more of a (considerably beefed up) minigame in terms of the design Chris describes (and how I get it).

 Chris did put a lot of effort into the combat engine, and it seems rather harsh to relegate it to a purely bonus addition to the base solar game. It's at the point pretty much all of the base elements exist, but needs to be put together to make it a more long-lasting experience. However, he's the one who has to decide how much more effort to invest between the two sides at this point. There are things to be worked out in the solar map as well that could personalise the various faction far more than what you'd see in a standard 4X.
 In terms of priorities, well, the solar map has to come first. However, as RPG review stated, people like me want to see that fusion between genres work.

 = = =
 I guess we'll find out more after the next round of additions and balancing. If the ring is staying, I would prefer the player start in the middle of the ring surrounded by enemy ships during assassination mission. Not that it would defeat anyone near Misery's skill level - the player ship is too fast and nimble to get hit by more than stray shots at the moment. I think you'd have to attack his computer instead with an overflowing number of ships. Difficulty here should IMO be measured by the tactics used to achieve your goals, rather than by the pure number of shots to defeat an enemy. Now, if only there was a way to demoralise an 100-armada fleet.

 = = =
 Oh hi Chris. Ok. Perhaps I'm a bloodthirsty bastard, or Eddard Stark. I prefer to end things with my hands, rather than a button. IMO your current alpha videos already over-emphasise the action aspects of TLF, so you'd better polish up the solar map so as to show the depth of the strategic sides.

 For the grognards, you could just put in an emergency repair function that refits your ship with some health and a long cooldown when they are critically wounded. It's difficult to randomly lose a minor battle with such instant resurrect devices - which is what Diablo 3 does so that all the noobs can go play Torment without not being crippled by a serious tactical mistake.

[x4000]

Couple of things:

1. The combat was never ever ever supposed to be a "second game," so to speak.  It started as a minigame, and it grew larger than intended, but never to the point where I felt like it was a standalone thing.  I love playing it, though.

2. As you say, I love doing things with my own hands, as well.  But not always, every time.  I mean, it really depends on the circumstances.  Sometimes Fast Travel is a nice option rather than taking the more realistic way of walking, right?  I mean, I love the feeling of distance that is given to me by walking, and having the option to always walk is great.  But once you get fast travel, you tend to use that to skip places that you've already been enough times, unless you just feel like going somewhere on foot to discover hidden goodies or for the fun of the combat, etc.

3. For me, I play FPS games in particular to relax.  Sometimes it is just freaking fun to go shoot up some guys.  I loved Far Cry 2 because it was really good at delivering that.  Others panned it for various reasons, but I mean it did everything I wanted out of it.  For me, having an action style of combat is kind of the same idea.  Getting my hands dirty and just having some fun.  But that sort of fun isn't for everyone, and if I'm not feeling in the mood for that kind of fun, I want to be able to just skip that and have the more cerebral kind of fun.

4. The prior videos of course over-emphasize the combat stuff, because that's what we've been able to show the most of in terms of what got visually polished first.  For the official trailer, I intend to keep that short and sweet (for once), and emphasize different things.  How that will work is something that is shaping up in my head pretty nicely, actually, and I think it can send the right message.

[Teal_Blue]

  Really really nice article! I really like the Command Room pic and am really looking forward to trying it.   

I think this is really going to make the game something special and very intriguing to play!! 

-T