Mostly Game Design Related Feedback
Hi! I finally got around to writing proper feedback. Overall I'm really enjoying the game, but I think there's still some work to do before it can become really popular like LoL did back in the day (wow it's already been 6 years since that game came out of nowhere! Also keep in mind how they started exploding with really pretty minimal features, having a good base will go a long way). My apologies if I'm pushing for a bit too much moba-like features, but I don't think it's reasonable to request too major feature changes even if it is still alpha. Here's some thoughts on what I think is missing from the GoA experience:
Reward Mechanics
The concept of reward and punishment should be familiar to most people. Too much reward and the game gets boring, too much punishment and the game gets frustrating. I think GoA is currently leaning a bit toward the latter.
The main reward mechanics in the game are as follows:
- Shard Collection
- XP Collection
Shard Collection: Having to pick up shards is a tedious task that may decrease the amount of satisfaction compared to automatic collection. They say that you have to give a dog its reward a few seconds after the action or it won't be able to correlate the reward with the action. This is much the same with humans too. Take for example the concept of delayed gratification - it's difficult to resist the temptation of a short term reward instead of a bigger long term reward. You might think this is far fetched (we collect shards from the ground at the nearest opportunity anyway, right?), but there's a few cases where I feel like this becomes a pretty large problem: destroying structures. Not only is it difficult to pick up all the shards from destroyed structures, the shards will also split unevenly between players, even though getting the Titan was something everyone contributed to!
--> Solution: You could simply remove pickups altogether and opt for a traditional instant reward system. I heard a developer mention in chat that there were readability issues with this, but honestly I don't see how this would be a problem. You could play an animation of shards flying toward your shard count in the UI and show how many shards you have collected above it. Alternatively you could show the amount of shards at the shard source location, just a blue label that fades upwards and says "+50 Blue" should be enough (basically how LoL does it as I recall). As for structure kills, I think they should always reward the whole team.
XP Collection: As I'm sure you are aware atm XP is super underwhelming, it only contributes to hero level which doesn't really feel like it makes any kind of difference in the vast majority of games.
I would like XP to be the reward mechanic that we are kind of missing right now - permanent rewards. You could argue that upgrades are also permanent rewards, but they don't necessarily feel that way because it is preferred to spend shards on replenishing your army and upgrades are just a luxury. The player should feel like with every good move they are stronger in relation to other players, this is the strongest kind of reward I can think of (and part of the reason why LoL and DotA are vastly more popular than HotS). It's hard to put into words why, but the individual satisfaction of becoming an unstoppable monster creates a hungering thirst to play more and play better. It's addictive.
This can be achieved simply by making XP more significant. Scale hero stats up with level (btw just in general it feels a little silly how weak heroes are atm imo), perhaps give upgrades you can buy with XP (talent trees, anyone?), maybe add more hero abilities or the ability to upgrade them if you want to go that route. One wild idea is to make upgrades cost XP instead of shards.
Social Rewards
The notorious e-peen is an important tool for motivating players, no matter what you personally think of it. Ranking players both in-game and out-of-game are great ways to motivate people to spend more time playing the game.
Credit players for good performance in-game: We already have decent after-game scoreboard (still a lot more interesting statistics that can be added there too, though!), but it is also very important to have one in-game! I don't have too much to say about this really, a table of stats like damage dealt, damage tanked, shards destroyed and shards collected will go a long way.
Ranking players in-client: This one is probably obvious, I'm looking forward to ranked ladder games! However there are also other ways of rewarding players in-client - avatars, achievements and the like - they should not be underestimated. Perhaps it's best to keep the focus on in-game features right now though
Self Expression
People often think about cosmetics such as character skins when they think about self expression in games, but it can be so much more than that. Any set of choices of equal viability comes down to self expression. Sometimes it can be gameplay oriented such as finding your favorite gun in an FPS game or a hero that fits your playstyle in a moba, other times it can be purely cosmetic such as character skins, badges, avatars etc. but it can also be somewhere in between - a really cool form of self expression is base building in RTS games, it technically doesn't make a large difference if you had rigid base building like BFME1, but most people much preferred being able to design a base of their own in BFME2.
There's some amount of self expression already in the game - choosing your hero and army composition - but I feel like it's fairly limited atm. While hero/army compositions make a difference ingame, they are not really decisions that you make while playing. Here's some ideas
- A larger (and more unique) selection of upgrades (thinking of them a bit like you would think about items in a moba)
- Building placement (or units that create structures). Engineer is an example of a good unit, it gives you a lot of choice of where and when to put down your turrets and whether to defend them or not, Purifiers, Lancers and flowers all reinforce self expression as well!
Unpredictability
Imagine Chess, the game is completely deterministic - it would be theoretically possible to calculate the best possible move 50 turns into the game. This creates intense pressure on the players to find the exact right solution and as a result Chess feels very stressful for a lot of players. In the opposite corner we have Hearthstone, where the game is so unpredictable that a master player can lose to a beginner. I feel like Starcraft 2 has found a very nice sweet-spot. FoW creates a good amount of unpredictability - a player can guess at what their opponent is doing, but it's possible to hide things you don't want your opponent to see, or even fake tech switches.
In GoA we have FoW and it does create a certain amount of interesting gameplay opportunities in terms of positioning, however I feel like it's not quite at the sweet spot:
- Positioning is fairly predictable since there's a very limited number of things players can be doing at any given moment
- Opponent tech options are very limited: Firstly you see the opponent tech so from the beginning you are aware of their possible choices, secondly the upgrades - which are hidden until finished and thus could enable some really interesting stuff - really don't play a major role in the game
Ideas:
- Create more choice in map objectives, maybe a certain camp gives more yellows and other more blues, perhaps a camp could give a buff
- Make the map larger and add sneaky paths to enable camp stealing
- Hide opponent 'decks'
- Buff upgrades in general (and try to diversify them a bit .____. )
General Feedback
The smaller details...
- Need an audio cue for when research queue is about to finish
- Map size is a bit small, this makes AoE abilities extremely dangerous
- Carriers should automatically stop dropping units when you load something in them
P.S: I'm open for discussion on TS/Discord if you want help with brainstorming. I really wish for this game to succeed! 
P.P.S: I did not proof-read this t.t sorry if I'm a little incoherent
Comments
Hey, awesome feedback, as usual I agree with a lot of the observations made, but I also think that there are some misrepresentations of certain aspects of the game in this post. I think this is a result of a big issue of Atlas. There are things that you do in this game that are very significant, but they do not feel significant. So for example, I think unit composition choice is a huge part of the game, but it is something that is glossed over as a strength because it's so ingrained in strategic norms. The same goes with positioning, because of the map people think there are less options, but don't realise that with a larger map, you might be even relatively more constrained because of how good the mobile units will become. I actually think positioning is very important, especially at higher levels when people will exploit your bad positioning.
Anyway, what I really wanted to talk about in this reply was heroes. I think it is a mistake to interpret Heroes in GoA as heroes in MOBAs. It would be better to interpret them like a hybridization of the classes in Hearthstone and the Planeswalker Heroes in MTG (example: http://www.planeswalkerslibrary.com/images/som/006.jpg). It's not a unit that is supposed to transcend the battlefield and dominate everything. It's more like a stabilizer and a constant 'power' for the player. In terms of how they affect the game, they are more like Hearthstone classes, but similar to the MTG planeswalker for the constant small utility and periodically big utility aspect. With this in mind, I think Heroes in GoA are in a really good spot at the moment, and I'd dislike if the game became more hero-centric than it already is.
I think a wider unit pool and hero pool is going to increase the amount of self-expression as well as the build diversity so I'm really looking forward to the game down the line.
This is a really intriguing idea to me. What if you could see opponent hero selection, but not what cards they have in play. This is akin to, say, Hearthstone where you can see opposing class but not decklist. Of course you'll pretty easily see eventually what is in opposing decks, but it could create interesting situations if you're not sure, for instance, which T3 unit(s) the opposing Green player took, until those T3 units actually come out. Maybe this would result in more potential for mindgames, strategic depth, interesting meta, even cheese?
I agree with hero levels feeling underwelming, but I dont agree making heroes stronger since i would like it to be more about units (more rts, less moba).
How about givining units levels, like stars in red alert and command an conquer games. 3 levels, big units level up slowly, small leveling up fast?
I aggre with increasing mapsize, I heard they already tried it, and that interaction was to sparse. But this may change as players get better.
A bigger map would also open up for towers defending parts of it, making killing towers actually good. Towers dont have to be in 3 lines like now, they could be a bit scattered.
Something along skill tree or hidden decks would be nice to increase diversity, i feel like each race should have more units, i dont feel i have to sacrifice annything by choosing 5, i get all i need, but i guess this will come in time.
God, now I'm imagining Aquadillos and Bramblethorns that are allowed to level up
I think the fact that unit choice is the sole aspect of complexity is really good for the accessibility of the game. It's supposed to be a very arcade'ey game I think. While the idea of stacking different measures of complexity and uniqueness on top of each other appeals to me as a strategy gamer, it's not very good for a broader audience. Some new players already struggle to pick up on certain game mechanics in the game already, like unit specific upgrades.
There are some arguably "interesting" interactions regarding picking up shards, mainly when your opponents can defend those shards and prevent you from picking them.
What I found really confusing is how the shards are shared. At first I thought that they were shared evenly among the team (HotS style), as I was playing with friends and they said they were getting shards when I picked them. Then I read somewhere in the forum that that's only true for people close to the shards (akin to LoL/DotA XP). What a bummer. I'm predicting carry-support dynamics where some player farms but intentionally doesn't pick the shards so a teammate can pick them and tech unusually fast. And I'm also predicting teams in solo-queue where one player is vastly better than the others and is getting most of the shards by solo-farming while the rest of the team is resource starved and having a small amount of units (and fun).
Would you prefer warcraft-like heroes? Your hero dies after a "small" mistake and you pretty much lose the game. That kind of devalues the rest of the units and makes unit composition less interesting and meaningful.
Think of units are your consumable items. You commit to a set of choices pre-game, and then in-game you have to decide what units to make, when to make them and what to do with them. Also, think of upgrades as your permanent items (like the generic stat-bags of MOBAs). Your build-path (when you build something or upgrade something) is another form of self expression.
It looks like you are trying to shoehorn forms of self expression from other genres into RTS without realizing what's already there.
Agree, more specialist units to force some army splitting, map control and multitasking would be awesome.
Those are my bigger concerns. Map objectives feel too straight forward because they don't offer many tough choices. Late game you either have AoE or big tanky units that aren't vulnerable to AoE. There's not that much reactive AoE-dodging micro, but instead you preemptively position units to be less vulnerable to AoE or build more big tanky units.
Totally agree. The best strategy games are those where basic mechanics are kept simple but interactions are deep, interesting and hard to master. It's better to have a deep game (interactions, mechanics or systems that are interesting and very deep, as that's quality) than to have a broad game (many different interactions, mechanics or systems, as that's just quantity with no guarantee of quality).