itch.io

Legitimately very well done. A self-contained V-slice with attention paid to little details of how things operate around you. Designwise, the fact that you can't ever get to the ground floor despite countless stairs between the first and second strongly evokes "the nameless feeling" that the game is going for.

I had actually been thinking of building something exactly like this for the past few weeks - a giant, BLAME!-inspired megastructure of a procedural airport. Audibly said "no way" when one of my idle Google searches brought me here.

> and if that target is already being attacked they look for the next closest one.

There are certainly scenarios where a target saturation algorithm makes sense (I once wrote one to hunt down a spread-out group of fragile targets during a hackathon, for instance), but I think that's actively counterproductive here.  Focusing down one target is the optimal strategy for line battles.

> Putting all defense on one ship and using it as a forward tank is a valid strategy but is risky as losing a ship is punished hard. So maybe spreading ships in a line is actually safer because it distributes damage better as enemies will attack different ships depending on their position. That is where positioning becomes important. 

In practice, I'd say that around 95 percent of the damage goes onto the front ship pretty consistently, so long as there's a meaningful gap. For positioning to be of interest to the player beyond that, I think there has to be a bit more complexity.

I like spaceship games, so I gave it a shot. Some bugs/issues:

 - The achievement unlock system definitely provides progression, but there's a huge gap between maybe getting a 100 credit item every round and maybe getting a 600 credit item every round. Also, the achievements and what they unlock should be visible with a button placed somewhere prominent on the main menu.

 - Energy weapons, even with the buff from their dedicated ship, seem useless. You want PD, autocannons on the mediums, railguns on the larges. That seems like the only worthwhile fit. 

 - Related, missiles seem useless too. They appear to explode before hitting the target, doing very little damage. Enemy missiles are lethal, though.

- If you lose a ship while still getting started, it's usually just game over. Getting to keep the weapons/equipment of a dead ship, and maybe sharing its EXP with your other ships at a positive-sum ratio, might make that less the case.

 - In general, you've got one ship taking all the damage unless enemies arbitrarily decide to focus down something else. Difficulty is low enough that you can afford to stick an armor+repair on everything else, but putting all defense on your middle ship and clustering everything else just slightly back from it is meta. Positioning should be a more major part of the game - maybe enemies should attack the weakest target in range rather than the first one they see.

 - Related, friendly ship AI is all over the place. I've had a single ship shooting at three different full-health enemies at the same time.

 - Am I missing something, or are crit buffs useless? 5% critical damage on a 5% critical chance is a 1.0025 multiplier to your total damage, if that 5% is applied to base damage. Still lackluster if it's applied to crit damage directly and it's 400% base. A 10% fire rate boost, assuming you spend a quarter of the time reloading, is a 1.075 multiplier to your total damage, which is thirty times better. Likewise, 10% faster reloads are three times worse than that fire rate buff.

 - One try, I started up the game, beat wave 1, and then wave 2 was a big ship that kited me to death with no opportunity to shoot back. No opportunity to get a weapon with longer range, nor any warning to upgrade range beforehand.

Beyond that, some more ambitious stuff:

 - Autobattlers need cool battle effects, so dedicating some effort to tuning the VFX and crunch of each weapon so that they feel different could be worthwhile.

- Variety is a major gap, both in player ships and enemies. Having different ship classes that do different things could help. As it is, you're just building a better stat brick.

 - Enemy wave variation in particular could use a look, for replayability's sake.  Maybe have a map or something, and enemies in different 'areas' have different doctrines. One round, the player chooses between fighting the enemies with lots of fighters and the enemies with long-ranged guns, and then, a few rounds later, has to choose between a subdivision of the fighter enemies where fighters drop close-ranged bombs, or one where they upgrade into fewer but stronger gunships with rocket pods.

 - Likewise, interplay. Right now, you have three ships that all work kind of the same. Upgrades are all mostly independent from each other, and amount to making some number higher. The core gameplay of this genre is planning out complex builds, since there's no direct player control or anything like that, so having upgrade X have some major impact on upgrades Y and Z that the player plans around is vital.

As an example of variety, strategy, and interplay, suppose the first upgrade branch let you choose between "destroyer" and "command ship". Command ships have one less medium turret, but allow the player to set 'orders', like telling friendly ships with low health to retreat behind the nearest friendly (if unobstructed) until they recover, or to always target the lowest health/fastest/slowest/deadliest enemy with XYZ weapon type.