itch.io

With charge attacks, I don't feel they should go off automatically once ready. That's not to say it should remain indefinitely until you're happy to let it land, but I think there should be a timing involved where once it's fully charged, depending on the initiated skill, give it a set window of time before it goes off automatically A fighter for instance that tends to have a lot of strength may be able to hold a charge the longest (ex: up to 10 seconds) but not indefinitely, while a rogue or bandit may hold a charge but only briefly (ex: up to 3 seconds). And a mystic (haven't touched mystic yet) might have a time window in between since they're not physically all powerful, but they are a master of their craft (if they have a charge attack, maybe 5-6 seconds?). 

With an auto-released charge-attack there's still an aspect of timing - it's just easier because now the thing the player is timing for is when the full-charge comes out, rather than for when an enemy comes within range or interspersed with dodging/parrying. There just isn't that much of a benefit of having a high burst-damage attack require that timing when it's realistically only going to reach a small part of the entire arena.

In contrast, for skills instead of the dedicated "heavy attack" button, that's an entirely different matter - those I think SHOULD have some abilities that have the option to hold, because things like the Deadly Dash fighter skillhave elements of optimizing their DPS by maximizing how many targets you hit with it. Makes a lot more sense for maintained charge attacks at range as well for similar reasons, combined with the requirement to aim.

I just recently picked this game up and from what I can tell with the change notes that were given (was trying to get a sense of when the demo stopped) a lot of fairly drastic changes have occurred to make it more tolerable for solo-play. Just from what I experienced, however, (got a bit stonewalled by the 12-18 level catacombs due to Fire Slimes having unrelenting DoT) I think I can get an idea of why the steps were taken, and at least from my perspective of how other games deal with 1v+ encounters or genres, there's more steps that could be taken to further a vision where that can be balanced with team-play setups. Given the game's selling point is already a bit of an exotic cross-genre approach that's combining elements of action-RPGs like Zelda or Soulslike, and MMOs with enemy-aggro and detection mechanics, mission/quest setups, and intent for teams - I think it's acceptable to consider influences from a few other sources.

• Beat-em-Up games: Power in Ambiguity, and Target Slots - So one thing I notice about Atlyss is that most every enemy, other than the miniboss and boss, unambiguously commit to pursuing the player and unleashing an attack from their moveset when they're ready - some of it dependent on their distance from the player, but the ultimate outcome ends up approaching the same: either a stack of several enemies repeatedly hitting the player, or a stack of several enemies repeatedly getting stunned by the player at the same time due to a generous parry window. They snowball into a mass around their target and this ends up exacerbating some issues that a lot of the later-demo designs have (the models are so big the player can't tell what's going on!) So what good examples of this genre of game will do (streets of rage, final fight, double dragon, god hand, sleeping dogs...) is that each player will have a set of target slots that can be occupied by enemies - basically locations to the left and right (or all around the target in 3d games). Basically this conveys not only a framework to tell the enemy which player they're going after (hey good tie-in with potential MMO aggro mechanics?) but it also can coordinate what everyone else in those slots is or should be doing - if we wanted to make it so that parrying had a shorter window again, this could be used to help space out the attacks instead of having them occur in rapid succession. And what happens if enemies enter or exit that list of slots? If there's no room or if they've moved out of position? They orbit, posture, loiter, wait their turn to get back in there. There's a really good video that I found that puts into words how/why this works, great 15-minute watch that can get gears turning for implementing clean 1vX battle dynamics.
• Metroidvanias: Foe Variety - This honestly could just as well be covered with the above genre, but I say, heck why not this instead? It's more noticeable here because most entries of this genre aren't going to have nearly as many entities fighting you all at once, which then makes the composition of what's being fought more noticeable. Examples of this can be seen in games like Symphony of the Night or Bloodstained where specific room layouts will challenge the player based on the combination of two tactically diverse enemies - maybe some Skeleton Musketeers stationed behind an Armor Lord, forcing Alucard to time his jumps, approach from above and cross them up. In Atlyss where the player's deliberately given options for both melee and ranged combat, this kind of variety in enemy loadouts should be heavily encouraged! ESPECIALLY if team-play is a priority, because this now would mean that, hey, if perhaps an encounter that used to be two of those death guys, was instead just one of them, and then some kind of golem turret dude - then a duo of players, maybe one a Fighter and the other a Bandit, could split targets between the two.
• General: Input Latency - Maybe partially due to how the combat behaves at the moment, but the matter of the fact is that it's currently  highly-reaction-based. Parrying is such an important aspect to master that being ready to use it on a moment's notice is probably the most important part of winning any fight. As a result my first and utmost focus in this situation, to make the fight take shorter, is to have the attack button held down, and this has a couple of issues concerning some of the cooler toys players get:
  
  • Charge Attacks are inherently laggy, get interrupted by blocking, and sometimes their attack animation's outcome (as I found out with katar charge attacks...) create a bigger opening on you than if you hadn't just continued attacking normally. It almost NEEDS a parry-stun lead-in to work. The fact it can be held indefinitely (?) also means that maximizing its DPS versus just standard slashes is really rough to time and that sour-charges represent a massive loss on that front, as well as for your stamina. I'd just make this work like how DS3 onwards did heavy attacks - you can charge them, and they go off automatically once ready.
  • Skills are deliberately greyed out while using normal attacks and their use cases usually aren't consistent from skill to skill - felt this the most with the Fighter Loadout, with Stomp being a charge that would automatically channel once charged, but required you to be on the ground to even initiate, then Lethal Slice being a charge that isn't auto-activated on full-charge and can be used in-air... So in order to use any skill I would need to deliberately stop using my attacks (which often were passively contributing to building up stun on my enemies) and take massive risks of being hit for what might be a bigger burst of damage across a larger spread of oncoming enemies. Frankly this should be taking pages from MMOs that have auto-attacking - your fancy skills interrupt your basic attacks, not the other way around.

Anyways hope any of that rambling made sense and/or was tactful at all. I think aside from such issues everything else about the game I've seen seems really neat.

Glad you're continuing work on this, and appreciate the executive decisions you're taking with how Action Economy works here.

now for the Pedantic Rule Stickler things as Guy Who Player Way Too Much Of This Game, i mean this all in a purely constructive manner:

• QUICK/FULL TECH: You can take the same option for a tech action multiple times within the same action (PG 71, FULL TECH, last sentence) - or in other words, you can use a Full Tech to do the same single Quick Tech option twice (i.e. you can Lock-On twice in the same turn without Overcharging. You can even Puppet Systems twice within a single turn.) This is also working under the same looseness of how this implementation of how Skirmish/Barrage don't strictly exist (which by the way I personally support. Though it might also introduce some questions as to how to implement talents such as Skirmisher that rely on this distinction...)
• FRIENDLY INVADE: You are allowed to use Invade options on friendly targets without either having to roll to hit, OR actually inflict Heat to your allies (Core Rulebook, PG 70). This gets particularly important if you choose Goblin and use Puppet Systems - yeah, it's not only an Overwatch activator on enemies, but it's an ally mobility-tool as well.
• LOCK-ON: Consuming lock-on is an option instead of automatic - a player can choose to not consume lock-on, which can be important if the weapon they use to attack is inherently accurate and would benefit less (since accuracy stacking doesn't help that much).
• ATTACK CLASSIFICATION: some actions classified as "attacks" in this demo, such as LOCK-ON, are not actual attacks. It's a silly distinction, I agree, but it somehow gets important when considering how what is and is not an "attack" interacts with frame traits such as Kobold's Mimic Carapace or systems like Swallowtail II Retractable Profile.
• SWALLOWTAIL: still lacks Integrated Cloak trait?
• STABILIZING: while I do appreciate the fact that just doing all the options at once is nice, it does mean that just turtling on the Holdout Point repeatedly becomes a much more viable tactic when you can just repeatedly full-heal and un-expose all at once.
• MANUAL BOOSTING: This is something weird that gets ESPECIALLY important when considering Nelson: doing a Boost action without using your basic movement, because Momentum is a thing. As it is right now, you have to move far enough to deplete your standard movement range to use the Boost quick action. Meanwhile using a Boost to immediately activate Nelson Momentum is a Strat. Also outside of Nelson stuff, it is the way to work around some of the more minute traits of how Engagement works with size (Page 62)

Especially considering how convoluted having to manually keep track of how the mechanics of some frames work (particularly: Duskwing, Zheng, and Hydra) I absolutely endorse and look forward to how this can streamline the process.