Excellent game, would like to see improvements

Home Forums General Discussion Excellent game, would like to see improvements

Viewing 5 posts - 1 through 5 (of 5 total)
  • Author
    Posts
  • #2032
    Magitex
    Participant

    I had put off playing Diaspora Ex Astris Renatus for awhile now, mostly because it looks like a fairly dull, standard-fare game where you stare at circles and triangles pretending you are exploring the universe. Here’s some extended feedback.

    Well, for starters I wasn’t quite right. This is more like Aurora 4X Factorio edition where the triangles and circles are rather epic in scale and as such have been glued to it for the past week or so.

    I started with a scant sized fleet with the notion of attempting to understand the game easier which was maybe a little overzealous of me because it actually created the opposite problem, even now, I still haven’t even completed construction of the basic facilities from the tutorial (I’m thankful for the scattered fleets and hulks). Nevertheless, I’m taking it slowly, optimizing everything I can in order to make progress.

    I’m at the point where you are allowed to settle a second system – which I assume is basically the very start of the game..
    With the preamble of my horribly naive experience with the game out of the way, I’ve quickly come to some conclusions:

    The performance is great so far, this is nothing like Aurora 4X where the game immediately slows to a crawl and crashes every step of the way. I’ve noticed the game can slow down here when you zoom in over a lot of circles so the overdraw is killing performance but that’s about it. The scale and AI factions working in real-time is tremendous.

    There seems to be a lot going on in the background as you play which I’m a huge fan of, although I’m not sure what fidelity AI are playing at (i.e are they playing by the same rules, do the AI have unlimited resources, are they randomly directing fleets around with no purpose in mind).

    So far it’s been really fun, but as I get closer to the core of the gameplay, the more combat irks me in comparison to everything else.

    And so the criticism:
    Combat in Diaspora, is simply put: horrible in execution. I can’t wrench any excitement, enjoyment or tactical depth out of it. The combat could literally be resolved by a dice roll and the game would be the better for it.

    1. It interrupts gameflow tremendously. We have this amazing agent system for logistics, but only a cookie clicker for a combat system. You can intercept ships automatically, but not resolve combat automatically without a pop-up and being forced back to 1x timescale.

    2. Automatic or manual resolution in combat, I can’t see any strategy so I don’t understand why it would even interrupt the player because there is no decision making to be made. All my ships move forward and die without being able to ever move back. My expectation here was that we could rotate ships to allow for repairs, shield regeneration, retreat, spreading damage between vessels.. this doesn’t seem to be a thing at least in early game, we can only move forward and pray ships don’t explode (particularly small vessels, acceleration and evasion don’t make the difference).

    3. Even if you do use manual combat, you’ll be hard pressed to understand the moment-to-moment effectiveness of your ships. Damage swings wildly regardless of what I’m targeting, certain weapons are effective in X situation, but I can’t even see the pattern of this in combat. It needs to show each weapons effectiveness on target and the to-hit calculations as well as feature a combat log/post combat alert log. Different colored bars would help also.

    I did like the fact we have additional defensive munitions, if nothing else, although it still feels like combat lacks variety and technologies.

    I’ll come back to my suggestions for combat, moving on to the UI.

    For the most part the UI is good, largely responds the way I expect, it could be expanded however:
    It’d be great if we could double-click objects on the map to bring up their information in situations where there is only a single object/vessel, rather than clicking the information button in the top right every time. The same would be good for the cluster view for fleets and systems. It would speed up the game tremendously for me.

    We could also use a button to pause ALL development of a structure/fleet.
    Right now what happens is that if I’m running low on supplies for an extended period of time, I have to go to a vessel, pause its hull construction, and then pause all the sub-modules individually (you can multi-select in some situations but not in every instance). There are also a few cases where you just can’t stop construction – I think in the shipyard you can’t stop the modules at all. This is more of an early game woe, but at a certain scale and situation I also imagine it’d be a pain to manage so I thought I would mention it.

    When we select a fleet on the cluster view, we could use a toggle that might show little green dots for each of your fleets in a system. Clicking the dots/insignias might bring up the fleet or the context tooltip for selecting one of them. Right now I use ‘Attitude’ to get an overview of where fleets are visually, but it’s not the best.

    While a fleet is selected (in the top right of the screen), it’d be good to see it’s current supply %, its current order, its heading and speed, rough ETA to destination, as well as the overall health of the fleet.

    The UI is not bad really, I just wanted a bit more from it. Perhaps being able to set our own icons for different fleets and what not.

    As for the agents system:
    Even as a programmer this confused me for a fair bit and while it’s a reasonably powerful system (the tutorial was good), I thought it was lacking some functions.

    A few of the things I wanted to do:
    Get the distance, speed, heading between fleets or celestials.
    Change a fleets refinery/extraction type based on an objects yield/quantity.
    Construct facilities or ships and the ability to split, join or create fleets.
    Add and use variables attached to a star system, I notice we can add a product_source of type fleet, but I wanted to be able to add string or number variable that my fleets could read/write and respond to.
    Set up a custom alert or timescale trigger, with custom message or ability to display a variable would be great (in post condition or active). I didn’t see anything like this but it would be really handy to get reports from your fleets using the agent system!

    I won’t comment on factions and trading because I’ve really experienced very little of it. It all seems well thought out and ripe for a huge epic confrontation, which is all the more reason why I’m so disappointed when I get to the combat phase.

    My suggestions on combat are a little spartan, but I’ll keep it as brief as possible, in short:

    Move combat to real-time rounds, one tick every game day (or week). The combat would use the existing system, only now it takes a lot more time to resolve every round and the player doesn’t control fleets directly (I guess this is still quite viable in slower timescales if desired).

    If you click on the ongoing combat:

    You can see a rough battle log of what has occurred thus far.

    You can prioritize how long a ship will spend at the front before being rotated (or % damage it sustains before attempting to move to the rear).

    You could disengage specific ships that are at the rear from combat, you would set this up manually, or with a % damage trigger before battle (or a click at any time). You could also program an agent to automatically disengage from battle based on certain parameters and transfer to a dock (assuming they can move) or join another fleet.

    You would also see ammunition, spares, supplies, because now you’re using them over the course of the battle.

    And to top it off:
    Every ship destroyed should become an in-world wreck for salvaging (basically a neutral fleet), so we can set up salvage agents to collect debris. You could click the wreck to see the contents, but salvage fleets would not question the player when looting (maybe a manual screen if you command them directly?). I get a huge kick out of salvaging in Aurora 4X after a battle and I like recycling in space games.

    At the end of this, ideally the objective is to be able to reinforce and resupply fleets DURING combat with sufficient time frames that make reinforcing/repairs possible. This is really what I want to see in Diaspora, with epic clashes of attrition warfare fed by the agent system.
    After combat, you would get a log detailing the individual effectiveness of each ship, how many components were lost on who and how much ammunition and supplies were expended. Similar for the enemies statistics, basically providing any feedback that gives the player detail on how they could improve their fleet versus a specific opponent (i.e overall resistances to our attacks, how our defenses faired versus their lineup).

    Ultimately I would prefer a much more complicated combat system (AKA Aurora 4X) but I think a fairly simple agent/attrition based system would be the superior option in Diaspora’s highly logistic world.

    This got fairly long-winded but I wanted to be sufficiently elaborate on what I feel lets this game down a lot, and although I didn’t talk about a lot of other minor things, they really were inconsequential compared to the combat. I still wouldn’t hesitate to recommend Diaspora if it happens to make it to Steam or another platform. It seems every bit as good as other 4X titles there.

    TLDR: I am not the biggest fan of the combat in Diaspora, but it is still a great game.

    PS. is the source available? I wouldn’t mind trying to run some combat experiments!

    #2039
    Zorro
    Keymaster

    Hi,

    yeah combat in Diaspora hasn’t worked out as I expected. My original plan was for the game to be distinctly “small unit actions” and have much more tactical decision making to make, but unwilling to impose arbitrary limits on fleet sizes I have always ended up with huge stacks of ships which makes the original idea of manually targetting things essentially impossible. Some of my first major forays into serious gaming were roguelikes such as Moria and later Angband. The turn-based combat in Angband is gripping when you’re in a tight spot and trying to work out what spell to cast next to improve your chances, or taking a punt on a phase door/teleport and hoping it doesn’t make things worse. I’d like combat in Diaspora to get a bit more like that; at least additional unit/effector types that might be able to boost shields etc…. Unsure about changing combat to be “real-time” as you state it, my main Sci Fi influence was Iain M Banks “Culture” where I imagine combat is over in the blink of an eye, with munitions displaced to targets and taking effect instantly, so technically the combat here is already “real-time”, it’s just slowed down into turns to give humans/players a chance to see what’s going on, but yeah right now there’s no point doing anything but sticking everything on Auto and seeing what happens, and combat tends to end up “biggest fleet wins” (assuming equivalent tech levels) with no chance of clever moves to save a fleet that finds itself at a disadvantage; i’d like to fix that.

    No, the AI does not play by exactly the same rules. I had intended to use the agent infrastructure for AIs to use and be 100% simulated, but it’s just too CPU intensive for any non-trivial sized cluster. As it is the game simulates too much (all the systems are AI populated and grow in the background even if you’ve never visited the system), factions can be wiped out without ever seeing them. The AI does not have infinite resources, each system has a certain carrying capacity that the AI will not exceed so they can be worn down.

    #2040
    Magitex
    Participant

    So just continuing to talk about combat design (more of an unsolicited rant), I think having huge stacks of ships is fine in turn-based if ships have specific group roles that overlap, like:

    Frigates assigned to anti-fighter/area defense.
    Brawlers with high armor designed for anti-capital and damage soak.
    Fighters for catching fast-movers attempting to flank and damaging key capital systems.

    In the midfield behind these, you might have heavy artillery ships with the highest general damage, featuring no armor, no sensors, packed with volatile ammunition locations and vulnerable damage boosting modules.

    To the rear you’d maybe find rearm and supply vessels, tasked with bringing ships back into the fight, these could be specialized for module repair, salvaging, ammunition regeneration/rate of fire boosters etc.
    Finally, the command and sensor vessels providing the fleet with superior target acquisition and tracking, without these your fleet either can’t track specific targets or aim accurately. Perhaps you can’t even pick your preferred targets without them.

    These roles just by themselves require a fair amount of specialization mechanics, as there needs to be a good reason not to just have every system on a single monolithic ship. Right now I don’t think it matters at all what a combat ship in Diaspora is comprised of, and each faction looks like it has rather even spread of everything so it’s difficult, if not impossible to strategize against. You’d bring a bigger fleet and not a more specialized one to every encounter which leads to battles feeling similar.

    Three options I can think of to help with specializing ships in Diaspora:

    Non-linear scaling for modules (larger modules better than more modules, for instance: 2x shield (+10) and 2x damage (+10) versus 4x shield (+25))

    Value efficiency (making it logistically cheaper to build one module category on a ship and dedicate its entire purpose to that, sort of like the dredge vs infrastructure is now, but extended to combat or technical fleet abilities)

    Booster/assist modules (one ship provides functionality for the fleet – i.e targeting, sensors, electronic warfare) these type of ships give huge advantages to fleet efficiency (by not having to build these subsystems for each ship individually) but as a side effect it also introduces an actual weakness into fleets that can be interesting to exploit or defend against.

    And I think that’s still really the start of more interesting combat, at least from my angle. Each of these roles then need multiple choices in the midst of combat that could tilt the battle, otherwise it’s just more of the same.

    To take the examples earlier to an extreme:
    Frigates could switch bonuses during combat between evasion or targeting (prediction modules). They could function as a temporary defense screen for a ships retreat, or ambush smaller evasive ships and fighters targeting a valuable capital ship in the rear.

    Brawlers could have an overdrive module where they are able to advance sooner in the turn order, either to reposition at the front or the rear quickly, while negating their evasion or hit chance. You could theoretically bolt these overdrive modules to the frigate if you wanted more of a limited strike craft that hits a specific enemy quickly and gets out before being locked (if relatively small).

    Artillery might have fire-control modules that boost the chance to hit specific enemy systems, rather than boosting general to-hit chance, these might hit shield generators or engines or specific modules the player specifies. This sort of thing allows you to execute strategies like eliminating enemy AAA targeting before striking with small ships.

    Opposing these system could be ECM modules which reduce the chance of fire-control modules working, you could put ECM directly on the ship you don’t want enemy hitting modules, or on an ECM projection ship where you can choose to counter specific EW attacks but the effectiveness is not as high.

    Back to Diaspora:
    So I think if you want more gripping turn-based battles, fleet design needs to be a bit more detailed than it currently is, with more clearly defined roles for ships or fleets. Perhaps it’s best to think of the ships themselves as skills or spells in a roguelike. You use each ship or subfleet for a specific task or to overcome a specific problem. Right now, it’s a bit hard to pinpoint what these problems actually are in Diaspora, which in turn makes it hard to design interesting combat.

    Here are some of those factors:
    No tactical movement, range or facing, eliminates a lot of strategy right off the bat. It’s not necessarily a minus, but you have to be creative to keep choices interesting in combat. Basically it’s like going from an overhead rogue-like or submarine warfare, to a round-based MUD.

    Time is a little too tactile. It’s hard to have any difference between a large, slow ship, and a thousand fast drones when most things have very similar turn cycles for targeting and movement. It’s not always a drawback, especially if you have complex abilities like a card game, but in something of a war game of numbers it doesn’t provide much granular detail.

    No logistics system can provide any benefit to combatants. If you’re not carrying all the correct tactical tools, fleets, ammo you need in advance, you lose. Meaning you (or the AI) cannot adapt to problems and there is little advantage in fighting in defense. You could rebuild an entire fleet when you lose to a specific strategy, but ultimately things are more limited without logistical aspects and a lot more binary for outcomes. You can’t for instance, perform a fighting retreat to buy more time for reinforcements, and you can’t bank on the enemy running out of munitions before you do.

    And finally the strategy at the tactical level. Although we can target specific vessels, no specific vessels matter in a fleet. We need reasons beyond DPS optimization for targets. I already touched on this earlier, but it’s one of the limiting factors for making combat more interesting. Faction specific modules could work here possibly with several factions fielding trademark technology on very specific ship designs (might be fun to provoke some faction vs faction warfare, especially if you have to trade to win). Bonus points if we can roll over their homeworld to illegally produce tech ourselves (or trade for them I guess!)

    That’s about it from me about combat, hopefully at some point others will chip in as to what kind of combat they want to see or if I just missed the point entirely.

    I should mention although I was baffled at first about the soul element of your ships, I think they’re a pretty cool feature and it’s nice having a defining advantage we have over the other factions. Although the souls don’t quite feel impactful enough because they automatically gain experience and there isn’t much you can do to enhance them, I like the idea they are bonded to specific ships. Having ship modules or research centers dedicated to the development of souls would be cool.. or some way of mutating/merging/breeding souls haha!

    Well, hopefully that was enlightening or useful in some way. I hope if or when you decide to work on the combat side of things you can pull some inspiration from this rant 🙂

    I think the AI seems pretty cool so I’m not too bothered about it not being 100% simulated, there’s something to be said for the epic scale that Diaspora has achieved. I like that it feels something like an actual space opera and not just kill on first sight (yet), not many games can get to this scale without getting bogged down by one thing or another.
    On the other hand, if the AI did have unlimited resources I’d probably throw in the towel, just from looking at just how stellar their empires are compared to mine!

    Just a question at the end of this, what the heck determines how fast structures are built and can I speed it up with more population or something? Shipyards are equally a mystery, should I simply build more ships simultaneously?

    I can see there are building skills for souls (I think?), but there’s no way to capitalize on those because there’s no mobile construction vessel or any way to transfer souls. Construction fleets would be nice to see if there’s currently no way to speed up constructions, it feels a little odd we can develop an infinite amount of structures but there’s no research or anything we can do to speed it up directly.

    Thanks for taking the time to read to my overly long posts.

    #2044
    Zorro
    Keymaster

    hi,

    thanks for the detailed thoughts. I’m definitely considering a lot of what you said. Even after the first post i did start wondering whether real-time combat would work/fit… it kinda could since fleets might end up barrelling towards each other at such relative speeds that even with displacement munitions they might not be in range for too long and have a quick mutual exchange of fire and then have to twist and turn to get back in range… OTOH if a fleet is fleeing then the relative velocities would be lesser and easier to keep in range so would hopefully not penalize asymmetric combat too much. I’m definitely considering it.

    for the turn-based combat, i decided that if there is such a thing as teleported/displaced munitions and AI are in control then the amount of incoming munitions you’re going to have is so overwhelming that whoever had the “longest range” would generally win so bodged that into being a universal physical constant and everyone gets exactly the same range so mutual bombardment at all times. With that in mind i’d avoided multiple range stages and everyone just ends up in a single bucket that can launch things at everything else. It’s pretty crude and could certainly be adjusted to have multiple “buckets” and some rule such as each bucket can hit things in the same bucket or adjacent ones, then you might have something like -3,-2,-1,0,+1,+2,+3 to play about in… would like some kind of support functions in the game, so limited only by imagination at the moment, could imagine most of the combat happening in bucket 0 but support ships in bucket -2 throwing “friendly” munitions (shield regen? warping space-time to deflect munitions? sensor fusion/triangulation for targetting bonuses?)

    In the end i might end up doing both (enhanced turn-based AND real-time), and then just have a flag so you can pick what style you want (maybe other flags for modifying the flavour of turn-based too)

    The “Souls” in-game are a rip-off of Iain M Banks “Minds” from the Culture series. I’ve thought about doing more with them, having them develop personalities (bonuses) over time that might have certain beneficial (or negative) interactions with other souls in the same fleet for “set” bonus type stuff. It’s come and gone in the dev branch for years now as i worry it might end up too micro-managey… oh damn this idiot has gone and turned eccentric now i need to ship it back off home and replace with other.. jiggle jiggle etc. Have also gone to- and fro- over whether to make souls transportable between ships, possibly being able to back-up their “state” somewhere as a snapshot in case they get wiped. In conjunction with getting “quirks” it might work where the Soul of your most prized ship/station goes a bit wonky and would be better off swapping it out…. if it goes ahead it might be another one to put on a flag so players that find it tedious can turn it off.

    #2046
    Magitex
    Participant

    I have to admit I don’t quite understand the concept of instantaneous combat especially when range, speed and firepower are still very limited. I can’t really come up with a sound strategy that isn’t jousting or just avoiding engagement.

    Just a thought on the souls, what you could do is allow manual migration to only larger vessels (30% larger than the last or there abouts), and each time you do so, it creates new quirks (perhaps based on scale of their upgrade and the previous skills). There could be a skill for migration so you can’t just swap souls until about a year or so (or a milestone required) but the process would be ultimately endless until you could not physically create larger ships (or the quirk maluses are too much to want to).

    So it doesn’t affect general gameplay, particularly of any first generation soul born with a ship, but the player could create some monster flagships over generations.
    It could get micromanagey but I think it’s a pretty fun concept to explore. It’d be cool to have our ‘own’ homebred ancient hulks with minds of their own a couple of hundred years down the track, perhaps even spawning their own twisted empires. I also love the idea of having ‘set bonus’ fleets, I would have never thought of that, although I understand it could get tedious to manage at times.

    As for the real-time side of things, I took the liberty of mangling a real-time prototype as a mod of Diaspora if you want to take a look at it. I’m actually still using the turn-based battle interface, only now you can minimize it and fleets continue about their business (they can also flee out of range or have more fleets join an on-going battle). The player is informed about the battle progress through reports on the alert system, so one can intervene still.
    So far I like it in real-time but it doesn’t actually address combat in itself, it just lets me up the scale of conflicts by having multiple combat fronts, and it makes combat a bit more organic. I also have the player set a fleet strategy by changing formation prior to starting battle just to try and inject more player agency into it (+damage, +evasion or rotating damaged ships to the rear automatically etc). My next idea was to try and add sensor/target systems and buff mountings (fleetwide resistances/debuffs etc) like you mentioned, but I feel like the AI would need to adapt or I’d have to reduce the players combat potential somehow. I guess throwing ECM ships and other tricky ships into AI fleets at random might be fun though.

    I’d love to go whole hog on positional combat, i.e submarine warfare style with formations but that sort of thing requires a ton of AI work so that the player can’t just run rings around it, so it rarely works well in larger scale conflicts. Having jousting passes in real-time wouldn’t be the worst option, but the question is how many ‘turns’ of combat is a single pass and how long does it take to turn around.

    If there’s one good thing about Diaspora’s combat right now, it’s that the only real advantage(or disadvantage) over AI is actually your own empires growing build power (although there’s no reason not to build a singular giant doomstack, but the AI is generally huge enough that it’s not a game-winning strategy in any sense of the word). You currently can’t easily side-step the AI like in most games, for better or worse.

    I’m not sure about splitting it between turn-based/real-time myself, from a workload perspective, although it doesn’t work too bad just dragging the turns out in real-time. It’d be difficult to be able to create an extensive combat system for both, assuming that was the goal. I can’t complain about having the option though.
    I guess these are not getting shorter 😀

Viewing 5 posts - 1 through 5 (of 5 total)
  • You must be logged in to reply to this topic.