Hello everyone! Like many of you, I am devoted to the "Annihilation subgenre" of RTS games, which in my book encompasses TA, SC/FA, SC2, PA, and now AOTS. Since SC/FA, I've been longing for a physical/simulated RTS that tackles the challenge of "massive scale" and "macro vs. micro" more profoundly, by innovative UI and advanced game mechanics. I'm hopeful that Ashes of the Singularity will be that game — or at least a big step in the right direction.So I'm not in it for the graphics, balance or competitive play, but for the possible advancements in visualizing and controlling large (real-time and physically believable) battlefields. I realize that I'm joining the discussion rather late — but I hope not too late to provide some useful ideas and feedback.TL;DR: The one most important point I will make is that AOTS can be, and should be, more innovative and groundbreaking in its attempt at "massive scale" and "macro vs. micro". (Stay tuned for posts on this matter.)In this post, I want to present feedback to various other points that have come up so far.
Brief
Positive:
Negative:
Worries:
Wishlist:
Detailed
Frogboy's relentless fight against the One-Icon-Per-Unit meme (Positive)A "massive scale" RTS clearly deserves (and needs) better ways of abstraction/visualization. When playing vanilla SC/FA (without Icon mods), it does feel like the icons are hardly helping; it's frustrating to keep an overview. PA improves the situation a bit by much better icons, but it's still one-icon-per-unit there. I hope AOTS will have something that's better (provides more information) than just dotplots or heatmaps (those certainly are valid options however in some cases)."We are working on an abstract map view mode (full map zoom out)." [Frogboy] (Positive)Alright! I am very anxious to see what this will look like. Can you share info about what you're planning to do here? Is it not too late to provide ideas in this area? (@Frogboy)"Overall, we do not want micromanagement to be part of the "skill"." [Frogboy] (Positive)Hallelujah!Armies/Battlegroups/Meta-Units and -Reinforcements (Positive)This is really what made me spend more time with AOTS in the first place. Especially army reinforcements is the kind of "non-trivial" feature I like to see more in a "next-gen-RTS".By the way, could someone settle the choice of term for this one?Global Abilities (Positive)Same here. This certainly helps reducing the micro. On the other hand, it might be that they will feel somewhat "soulless" (excuse the term), compared to effects that visually originate from units directly. We'll see.Slow pace. But NOT the slowdowns! (Positive)I do like the pace generally. In some cases however, units seem to slow down to a crawl. When it's a dreadnought that is desperately needed at the front line and it's just crawling there (slower than usual) — Well, that's annoying to say the least.Pseudo-Units (Archer/Brute): What a huge disappointment. (Negative)This is the one "feature" of AOTS that is truly disappointing and confusing. Confusing in gameplay, and confusing as to why the heck someone comes up with this idea. It does complicate the game a lot and makes it a harder to understand and use, without any apparent benefit whatsoever. It feels like deceit, to be honest (not to be overly dramatic), or a "betrayal" of the core "massive scale" promise, to be precise. A game like this can only be (feel like) massive scale, if it is truly simulating every single unit, and if every single unit can actually be controlled independently of others. These Pseudo-Units or Forced-Groups violate this, and make the "scale" feel "faked".With this feature, it is possible to see four Brutes that could be four quarter Brutes (using 4 logistics) or just one full Brute (using 1 logistics). See, it's not even possible to properly talk about "units" this way... It doesn't even make sense to ask: "How many Brutes are in this Army?" Moreover, the health info is irritating (240/320 for three-quarters-of-Brutes, but repair is not possible, because this would have to "bring back" one "quarter-unit"). A quote:"While the game will allow you to control individual units if you'd like to, you won't want to. Ashes is not designed to be a game that rewards clicking speed." (FAQ)This is currently not the case. Firstly, we can't control every single unit, and secondly, one DOES want to, mostly because meta units are currently rather broken (see below).PLEASE REPLACE THIS BY PROPER SINGLE UNITS. Now, if you really, really need to have units "come out" in groups, then just let them be Armies (of 4 single Brute units, e.g.) that then can be split up!What was/is the reason behind this "feature" anyway? (@Frogboy)Current Army/Battlegroup/Meta-Unit Behaviour/Robustness (Negative)Others have complained about this already. I'm aware that we're in Alpha, but the implementation of Army behaviour should receive max priority, more than anything else, in my opinion. This is the one feature that needs to work flawlessly on release. Players must never feel like it's good idea to break up armies to improve unit behaviour. Because when they do, this core feature can be considered a "fail".Meta-Unit Stat Buffs (Negative)Having this is a valid design choice. It feels misplaced in the AOTS "niche" however. I'm not strongly advocating against it, but it really does feel "magical" instead of "physical". Having unit stats buffed by "assigning" units together makes absolutely no sense (in a physical world), and that hurts immersion. A player should want to assign units together because of their respective function that they perform (offense, defense, protection, repair etc.), not because they magically alter the other units' stats.Right mouse button for panning (Negative)Actually, the problem is having the already "overloaded" usual right-click and camera movement mapped to the same button. If panning really has to be on the right mouse button, then the "move/attack/guard/join" command needs to be mapped to something else. This is really a problem while playing; I feel like I have to wait a while after moving my mouse and then click "carefully", so that my orders won't be misinterpreted as panning. As others have mentioned: At the very least, this must be configurable.Upcoming BETA (Worries)Well, this worries me because, as a software engineer, when I hear "beta" I understand "no more changes except bugfixes". I hope very much that there's still time (and courage) for more profound innovation. (@Frogboy?)Resistance against abstractions (Worries)"Ashes of the Singularity games can be described as a war across an entire world without abstraction." (FAQ)Well, how do you intend to visualize (get a good overview of) a "massive scale" battlefield without any kind of abstraction?Extent of moddability (Worries)"Ashes is very moddable. To mod, you will need to get a good CSV editor as that is the format we're storing everything in." [Frogboy]CSV, really? I have only quickly glanced at the AOTS folders, and seen CSV and XML so far. What do you think moddability will be like, compared to SC/FA (Lua) and PA (JavaScript)? (@Frogboy, @tatsujb, @ColaColin)Chrono Cam (Wishlist)Chrono Cam is perhaps the most impressive feature of PA. AOTS totally needs to have this! At least, Chrono Cam must be available for reviews. Better however if it were available in-game. Ideally, Chrono Cam would be tightly integrated with events. For example, "enemy dreadnought detected" could show up on a timeline in the UI (which visualizes events of the last minute or so), and a click on the "EDD" event would then take the player to the place AND TIME of the event! How great would that be?!Borderless/Infinite Maps (Wishlist)I realize this is probably out of scope, if only because an engine not built to do this will probably not be enhanced accordingly. Anyway, let me still advocate for borderless/infinite maps. They work by simply repeating a rectangular map vertically (without any offset) and horizontally (with an optional vertical offset). The effect then is that the map has no "beginnings" and "ends". This would, in my opinion, help much to convey the message of "battling over a world" instead of some backyard with artificial straight borders. Also, it would allow much more symmetric and interesting geography, e.g., having each player have exactly 3 players around them (120 degrees apart), or two players who can both attack the other by going north, east, south or west!No Upgrades (Wishlist)I'm aware that this most likely won't change, but still: One of the key ingredients of the "Annihilation subgenre" is the absence of research trees, in my opinion. (And this is why I consider "SC2" not to be truly "in line".) I'd rather have buildings and units perform functions, instead of hidden stats being tuned. And it feels weird too: Needing certain factories to build certain units makes "sense", while having to research the same things over and over again in every battle feels "unrealistic". And why should it be possible for the armor, health, strength or even armament to suddenly change for all the units in the battlefield, anyway?! (It's magical!)Weighted pathfinding (Wishlist)Currently, units seem to drive to their destination the shortest way possible, completely ignoring the facts of the battlefield. In particular, units that drive towards Armies as reinforcements might want to take "safer" detours. This could even be configurable in-game. A player might define the pathing penalty for different kinds of territory. For example: Captured and visible territory (100%, fixed), territory known to be free of enemies by vision or radar (80%), fog-of-war territory (50%), territory with enemy presence (20%). This would mean, for example, that a unit would take a path up to 2.5 times as long through fog-of-war territory to avoid travelling through territory with enemy presence.Area commands (Wishlist)Another reason not to have panning on the same button. Area commands in PA are utterly helpful. Why not tell an Army to capture an entire area on the map? And the Army will capture every node in the area, in such an order as to minimize the time it takes to complete the full area command."Imbalance in Power" game ending condition (Wishlist)To reduce the "mop up" problem, I'd like to see a "power balance" game ending condition. For a two player game, for example, one player wins as soon as they have superior economy and military, where "superior" would be configurable (factors 1.5, 2, 4, ...). Of course, the "imbalance" factor would need to be visualized (always, or only when close to the configured threshold).
Very well written. I am meant to be doing something else right now so will just agree with one thing right now:
I also wish all the units were individuals, when you can't select individuals there is a disconnect and you automatically have a sense/niggling feeling of not having full control. The more barriers to control (incl. fighting with the UI) creates a cumulative effect which risks creating a feeling of frustration and alienation with the game.
A lot to agree with here.
Yes, please let's settle the terminology of armies/meta units/battle groups.
The pacing of the game really has to wait until a teleport capability is added to the game. If you can teleport that new Dreadnaught to the front line, unit speed is less of a factor.
I wonder if the pseudo-unit exists to reduce the overhead of pathing. Four units that have to take the same path is less expensive than four units that can take different paths. Your second paragraph on this topic has good points.
Yes, very solid and predictable battle group behavior is absolutely vital to this game. Making it clever will be an innovative feature that will be a selling point for the game, but only if the behavior is reliable.
I am also worried about the future of the abstract view, or strategic view. They plan to make it better, but I don't think they really know exactly how yet. First, you cannot assess terrain from that view, and terrain is pretty important in this game. As for units, they want heat maps, but heat maps provide very little information to help make strategic decisions. "Uh oh there is a big enemy blob coming" is about the limit of what you can do, and even then you have to shift your main screen to that area to see if you have the right composition of units and structures to handle it.
I'm not so sure modding is going to be a selling point of this game either. I've looked at many of the CVS and XML files and while there is a lot more there than people realize (like formation parameters, AI group templates, objective priorities), core gameplay mechanics are hard-coded. At this point modding is NOT a core design criterium for the game, which makes it an afterthought, and we know how that usually ends up.
Area commands - YES PLEASE!
Many good points/ideas here. I must strongly agree that the right-mouse-button panning is a negative. In trying to pan I've changed move orders accidentally requiring me to move back to the last unit(s) I gave orders to and have to re-issue their move correctly. I've taken to left-clicking a clear area after giving move orders in order to de-select the unit before using the right-mouse-button to pan. Perhaps WASD or pressing and holding the mouse's scroll button? Just a couple of quick suggestions.
Great stuff. Will respond in detail later but what is chrono cam?
Re controls. We plan to make this configurable.
Beta has been pushed to January.
No rush! Thanks!
Thanks! It's appreciated.
Exactly. It's not about actually needing control of every single brute (tactically speaking), but more the feeling ("sense of control" and "scale") that it evokes.
Hm. Very well possible. However, considering the abundance of disadvantages it brings, that really doesn't seem worth the hassle. And more to the point, optimizations like that should be hidden from the player, and certainly not cause usability issues. As a solution, why not do this "fregatte-pathfinding-subgrouping" implicitly when Armies are created? Within an army, I don't mind seeing groups of Archers/Brutes acting together, just as long as the UI communicates one-unit-per-visible-object, without exception.
Indeed. It is not to be underestimated how important this one feature is. Armies need to behave "they way should, in an optimal manner" (effective/efficient). And I'm not saying that this would be easy to implement ...
About the "terrain" issue: Well, then we need better abstractions! Zooming out, everything needs to be abstracted sooner or later, maybe even at different steps (distances), in different ways. Terrain obviously can be abstracted as well, see real-life maps.
Oh well. We can't have everything. I'd rather have flawless and super-powerful Army mechanics, and I'm willing to sacrifice modding to that.
Actually, I was asking exactly because you're busy with FAF . So I assume you know a thing or two about moddability of SC/FA.
To be honest, I haven't looked at modding SC/FA or PA in detail, didn't have the time. But I certainly do like moddability in general.
No worries. I wasn't trying to ask you for modding support or anything. More how you, as someone who does mod (FAF), would characterize the "moddability" of this game, given the CSV/XML scheme.
Right. It's a problem in both ways, orders not issued when they should have been, and orders issued when they should not have been.
Oh. You really have to try PA if you haven't already! Chrono Cam is basically a review tool, or "interactive movie player" to explore the past of a game. It is normally and extensively used for reviews, but it is available, with the same functionality, in-game! That is, for example, when I see one of my outposts obliterated, and I missed it, I can very quickly open Chrono Cam and go back to the time when it happened, and watch the event, to see who attacked and how, and where the attackers went after that! Very useful, and very impressive, technology-wise.
The most important thing here is really that the internal review tool of Ashes absolutely needs to support "scanning" or whatever the term is (going back and forth in time, selecting any time efficiently, as in a audio/movie player).
But having an actual in-game Chrono Cam would be awesome, of course.
Good to hear. However, I'm worried that if the current assignment stays per default, some "decision algorithms" stay in place that make the UI feel unresponsive. I mean, it sometimes feels like the UI is waiting to see whether I'm trying to give orders or do panning (it "disambiguates"). But it really shouldn't, panning and giving orders needs to be super responsive and exact. Again see PA as a good example, its UI is very precise and latency-free, much in contrast to current Ashes.
Hm. What I wanted to say is: Is there still time (and budget and willingness) to try out "new things", UI-wise?
[...]it's pretty mind-blowing actually.[...]Uber had plans to use this feature to unlock "Pick-up-anywhere-as-anyone-in-the-game" in replays. Which honestly I'm dying to see in Ashes.
Actually, PA does indeed allow to "rewind" the game to any point in the past in single-player. So it conveniently allows "trying an attack", then rewind, then try again differently to explore the mechanisms while playing against AI.
And yes, "mind-blowing" is the appropriate term for this feature.
When I first heard of aots,I heard we could create and tailor our armies/meta groups how ever we want I thought it would be a little bit more then just click to build what units we wanted to add to a certain meta unit,I imagined it would of been something similar a stardrive(space RTS)type interface, it that interface you could for example set default behaviours for each unit type,so low health torpedo ships would attack only from a certain distance and your repair ships could be assigned to only help your weak units leaving heavily armoured ships to take the hits, would anything close to this type of individual unit role assigment be possible?
For anyone interested in what I just said check out stardrive fleet designer
Re Brutes/Archers
This is a bit of a trade off. Those brutes actually are 4 different units in every sense of the term. So why did we do this?
Balance. We don't want the factory cranking out a unit every 4 seconds. It's overwhelming to new players. But we also wanted to balance the game around rate of fire. So that some units could do a lot of damage but have a slow rate of fire.
Possible Solutions:
1. Allow advanced players to ungroup the unit via the same UI as you create an army (they are literally 4 different units as it is now, we just send them out as a meta unit).
2. Have them come out of the factory as 4 individual units at once. I don't like this because it's just overwhelming. We already have players accusing the game of being about cranking out blobs.
You can actually play around with this for yourself in the unitmodules.csv file.
Re Armies
This is a high priority for us too to fix. It's still a bit buggy. It's a very very complex animal.
Re Unit Buffs
Agreed. This is a bit of a hold over from an earlier design concept of how we were going to make meta units work. These buffs will likely be moved into the T3 unit abilities.
Re controls
With regards to defaults, other than the middle mouse button (I'll veto that, I hate it) which do you prefer? Left or right? I'm okay with it being left to move camera with shift key held down to select units.
more to come.
I believe settings per army/meta-unit like "stances", or perhaps "autoreinforce" or similar mechanisms are under consideration. However, don't you think individual settings per unit type might be too much for the player to (reasonably) control in a large real-time battlefield? (Is the game you mention real-time?)
I'm not sure I'm following -- what does rate of fire have to do with this, exactly?
Hermes finish in 5 seconds; would that be such a big difference?
In the end, having units stream out one by one (i.e. no special treatment for Archers/Brutes) is the only real solution I see to this problem. Why do you think that that would be overwhelming to anyone?
Agreed, "2." is suboptimal.
Like I (somewhat emotionally ) indicated in the OP, "1." would be a compromise, but it would only solve part of the problem. I see the following aspects to the issue:
The first can be solved by making Archers/Brutes "Disband"-able. To solve the second, you would need to count logistics per Archer (1/3), and per Brute (1/4), instead of "rounding up" per Archer/Brute-group. (Of course, logistics numbers would need to be scaled or adjusted to avoid fractionals.) The third can only really be solved if you make the factory produce the exact number of Archer/Brute-units requested in the UI (the number of clicks on the button), and not a per-type multiple of it.
Why is this an issue? Because if a single click on the factory button causes more than one unit to be produced, it will cause confusion somewhere in the game. Does the number of clicks equal the number that is displayed in the factory queue? (Yes) Does the displayed number equal the number of units that will leave the factory? (Depends) What number should be displayed in army content overviews (not yet available; it should tell the number of actual units ("vehicles")). So one army says it has 20 Brutes and Archers, so how do I order another 20 Brutes? And how 20 Archers? -- "5" and "not possible, click 7 times, get 21".
I believe this will always confuse, one way or another, because there's different ways to count certain types of units. Now I don't claim to know what's best in the end for this game; it needs to work, and make players happy. But I can't shake the feeling that this one introduces more confusion than it is helpful, with a distance. I can't say how most players will feel about it, but at least one other person seems to agree with me (see first reply in this thread)...
I believe that!
So these T3 abilities will change stats of units in the same army as well? Or will they be "functional"/"physical"? (More drones, larger beam, etc.)
Heck, I'd prefer that. But it would probably not be a good idea, for the average player's sake. I mean, if you're worried that 4-second build times will overwhelm/alienate people, then this certainly would too. Since the essential part is to avoid ambiguity, selection (and deselection) would strictly require Shift to be pressed, and a simple left click wouldn't do anything -- that'd be weird.
Suggestions:
Looking forward to it!
I actually hate middle mouse panning as well (because pressing the scrollwheel is awkward), that's why I've mapped the middle mouse button to other buttons on my mouse.
I thought about Spacebar+Move first too, but Spacebar is the map, currently. With that, or with Alt+Move, we'd have:
Quadrim in stardrive you don't micro manage a every single ship in a fleet just as in atos,your fleet can be tailoered to do exactly what you want when you want,so I'll use aots unit examples,if you have attached to a meta unit and an opposing meta unit has anti air instead of the bombers attack the meta unit as a whole you could have bombers attack anti air before any other unit in that meta unit,so you makeup a template for whatever way you want them to behave,I'm probably giving the idea no justice
Brutes are great in packs of 4.
I see no problems other than I want to be able to tell units to follow each other.
Panning The Camera
Shift is a commonly used key. What if holding shift and moving the mouse to any margin allowed panning? If I'm not holding down the shift key I'm limited to my current camera view as I am today. But if I press the Shift key and hit a screen margin I begin to pan.
I'm trying to remember what the control scheme was that I thought worked well for panning and I don't recall. I play so few RTSs (again, cuz I suck at them). But this sounds almost intuitive. Critiques?
Can you find a good online video that demonstrates this?
yeah but who uses THAT?
I think it comes in handy for quickly seeing what's happening (hot spots) on the entire battlefield. I wouldn't be sad if this stayed, even with full strategic zoom.
No problems? What about all the problems mentioned so far in this thread? (Not just by me.)
Since armies should definitely report the number of units that are part of them (the actual number, not the subgroups), but this number isn't the one that's used in factories, you have a complication/confusion: You have an army with 15 Brutes and 15 Archers, another one with 25 Brutes and 25 Archers; how do you make the first as strong as the second?
If no such per-type-packing existed, the solution would be simple: Order 10 Brutes and 10 Archers.
With the current packing, you need to memorize what the pack size is per type, and then start doing math. So you can order 2 or 3 Brute packs, resulting in 23 or 27 Brutes, and then you can order 3 or 4 Archer packs, resulting in 24 or 27 Archers.
Now, I don't know what you call this, but I call it confusing, complicated and possibly frustrating to use.
I think Ashes won't support that. And it doesn't need to, since "units following each other" is done by combining them into an army.
Isn't this what Victory Points accomplish?
There are many great features available to you once you register, including:
Sign in or Create Account
