So this is the new AI War, eh?

[Pumpkin]

Hi folks, old and new!

I've just finished a 9h30 "quick and easy vanilla game", and I though I would walk by the forum on my way back. Overall, I'm quite happy with where the game currently is. Let's discuss the details, shall we?

Graphism

The game looks nice. It lacks some polish, sure, but nothing awry for an early access. On a personal note, I'm not seduced by the 3D. It's fancy, but I'm not one to seek that in my games. I like Dwarf Fortress; AI War Classic's 2D style was fine for me. The gameplay is spatially bidimentional and the level at which position matters is between galactic and planetary. When you see the units, you can't meaningfully interact with the gameplay (get meaningful information or give orders). The nice little 3D models of the ships and the gameplay are in two completely separated segments of the game. You can't have both at the same time. I don't know if I'm clear, here. Anyway, AIW Classic had the same problem, but at least it wasn't a central point of the game. Here, I feel like Blue's hard work is hidden far away from the gameplay. Sad.

Squads

Which brings me to the squads topic. At "gameplay zoom", squads have no relevance. As I worried during the kickstarter campaign, "units" could be one ship, it wouldn't change anything. And I believe it even adds unnecessary noise: there are ships, squads of N ships, but what is build and controlled is a squad, but the damage and health and stuff displayed are for one ship (Right? Heck, I'm not even sure!), but the unit-cap is for squads, but starships are squads of one... Ugh.

The argument I remember from during the kickstarter was "It would allow us to have hundred of thousands of ships without overloading the game engine". Yep, but if a squad behave like a lone ship, then it's just cosmetic. I'm ok with cosmetic. But here, I feel it's noise for no gain. Anyway, change it or not, it doesn't affect gameplay (save that little ship/squad noise) and I couldn't care less.

Mark

New tech panel is great. (New hacking panel too.) Having all units concerned by a research level-up together (essentially getting rid of low-mark units, compared to AIWC) is a neat improvement. I wouldn't have believed it on paper, but I like that change.

Bonuses and the Fleetball problem

I remember reading something about that in the Pivoting AI War 2 article, and I felt it. Currently, it's okay to just pack your fleet in one blob and steamroll your targets. In AIWC, what hugely encouraged you to do so was the bonus/hull-type mechanic. I loved that, while other games may give some +50% bonus to this or that, AIW went up to making x4-x6 bonuses a common thing! You really had to choose your units to assault one big target (usually a guard post).

Here, there are many little weird bonuses. Albedo? Mass? Eh... Also personal shields? I really loved the old bonus/hull system. It wasn't perfect, but at least it was clear. (Or was it?) I had grand ideas for that (see). I'm ok with getting rid of these admittedly weird hull/ammo/bonus, but they need to be replaced with something. My main critic of the new bonuses is twofold. They are more cryptic. They are not loud enough.

I wanted to merge the old armor/armor-piercing in that hull/ammo revamp for clarity and coherence. I feel like personal shield is adding even more noise.

I advocate for a central, unified system for damage bonuses.

Alternatively, speed, range, endurance and damage are huge tools for differentiating ships. In my vanilla game, the raptors were rarely in the fleetball: their huge speed put them aside, making them much more adequate for raid missions. Similarly, very-long or infinite range units are automatically set aside of the fleetball to not risk them in close combat.

There are ways to avoid the fleetball problem. A hull/ammo system is only one.

Parting words

Anyway, I hope I'll find time to try to mod my grand hull/ammo scheme.
Also, I'm interested to head Chris on that topic. Is it still a work-in-progress part or are you more willing to refine the current system rather than changing it completely?

Don't let my criticisms muddy the whole picture. I love that game. I'm glad it became closer to the original over the course of its alpha period.

[RocketAssistedPuffin]

Hello! Neat to meet one of the people from when Classic was fairly active. All before my time. Glad to hear you're happy on it (I'm sort of a volunteer so...I guess I can say things like that).

Graphism

On this topic I am a bit...eh, on parts of it. The models are amazing, but indeed you can't really see them and play at the same time. I believe that's been a topic that's been prodded a fair bit already with...not really any progress. At least the really large units are visible (Devourer being the most obvious).

I do really like the planets and backgrounds, though. And the shots, which are clearer than dark red on black background most of the time. Dyson Sphere is especially nice (if you haven't seen it, it replaces the planet this time around instead of being a unit to the side).

Squads

The stats are entirely for the squad, the "unit". The ship count is just cosmetic. On the topic above of graphics, the fact fleetships are...squads made of tiny ships does make them really hard to ever see, I've found. I added some 3D formation layouts yesterday to units that didn't have any, but it is admittedly hard to ever notice - a lot of the time I thought things were overlapping horrendously, only to find they're not - just perspective.

And a personal "Aargh" is the unit stacking, for gameplay mostly but what you say here makes me think:

"It would allow us to have hundred of thousands of ships without overloading the game engine"

Stacking, er...kind of reduces the AIs ship count by a lot, so even massive forces become...tiny.

Bonuses and the Fleetball problem

The comment about weird bonuses, yeah I showed someone a picture of the stats and they were very confused at things like albedo, the scaling, and why these were used. There is...in balancing, a fair bit of flexibility from it, but also some severe rigidity (example being Concussion units affect things with <=50 armour stat. Grenades affect everything <=30 armour stat. So the Concussions do the same job as Grenades, but...more. And what with being able to go up to Mark 7 (!), you don't really need to get extra ship types to do the same job anymore. Which is a particularly annoying thing personally).

On the topic of fleetballing...yeah, not much reason to do anything but that - other than Beam Cannons. I don't have any real clue myself on how to fix it.

I'll have a read through the thing you linked. As was mentioned before, the game is crazy moddable (even a coding idiot like me managed to do some C# without causing an inferno).

[Bummeri]

I think that AI force-fields also discouraged fleet-balling. It was often advantageous to just take the bombers to kill the force field and then bring the rest of the ships in for cleanup. This could also be preceded by first taking out some nasty bits from under the force-field with raiders etc.

Half of the battle for capping planets was the puzzle you had to solve of how to get to the important bits without losing too much of your fleet.

Also logistics was an important bit. How the transports and such should be used so that you could bypass nasty-unimportant planets. This also discouraged fleet-balling since you had limited options for how to do the movements. Planet hopping was also quite dangerous, so it was wise to only bring parts of the fleet you needed for the raid/task. This way you did not loose your whole fleet and get a nasty retribution right away if you had failed something.

My main critic of the new bonuses is twofold. They are more cryptic. They are not loud enough.

I think you hit the nail on the head with this analysis. They need to be made clear so that you understand them and their significance intuitively from the game without ever going through any sort of tutorial.  And like you said, they should be loud! They should be a game defining element. This is the most important game mechanic that leads to the tactical challenges that form the puzzle of how battles should be fought.

Which brings me to the squads topic. At "gameplay zoom", squads have no relevance.

I think the greatest value that the squad system brings is that we could differentiate between precision weaponry and AOE better. If we were to have a squad consisting of 20 very small ships, some weapons regardless of their damage could only take out one part of the squad. In the other end of the spectrum we could have weapons with massive AOE and massive damage, that could devastate multiple squads at a time. This sort of granularity could be intuitively conveyed using the squad system and some status effect for weapons. This sort of granularity would create yet more depth to the tactical aspect of the battles.

[RocketAssistedPuffin]

Half of the battle for capping planets was the puzzle you had to solve of how to get to the important bits without losing too much of your fleet.

This was something I had tried to...implement, but I think I've failed on that part...it was honestly something I did really like about the first game.

Also logistics was an important bit. How the transports and such should be used so that you could bypass nasty-unimportant planets. This also discouraged fleet-balling since you had limited options for how to do the movements. Planet hopping was also quite dangerous, so it was wise to only bring parts of the fleet you needed for the raid/task. This way you did not loose your whole fleet and get a nasty retribution right away if you had failed something.

Is a fair point. Transports...are more likely to be a thing now, I think. At least with those, I'd feel better about upping the defenses of AI worlds, so you aren't quite as stuck behind MK4-5 (MK-5 being a normal planet is possibly part of the problem, instead of just Homeworld Guards).

In terms of the bonuses, I wonder if the old hull system would be...better in the end for damage purposes, but the current stats could stay for things like tractors, gravity, etc (which is the real interesting part of it to me - units being immune to an Etherjet tractor, but not to that of a Guardian, things like that which I've sadly not been able to super implement). As a later note addition, an odd quirk of the current system is when changing stats for...use with things like tractors and gravity, you also start affecting the things that damage the ship, so it can be a little awkward.

Re-reading an older post from Chris (first reply here https://forums.arcengames.com/private-alpha-discussion/questions-about-and-feedback-on-game-going-forward-(era-of-discovery)/ ), it...seems to be the hulls themselves were fine, it was just all the immunity tags, etc. Which the stats would still fix...even with hull types.

I like it in theory, but this isn't my game nor my decision.

[Pumpkin]

Hello! Neat to meet one of the people from when Classic was fairly active. All before my time. Glad to hear you're happy on it (I'm sort of a volunteer so...I guess I can say things like that).

Nice to see the new volunteers are as nice as the old ones.

Half of the battle for capping planets was the puzzle you had to solve of how to get to the important bits without losing too much of your fleet.

Exactly! And that puzzle was a lot of fun! And that's what I'm missing the most, currently.

it...seems to be the hulls themselves were fine, it was just all the immunity tags, etc. Which the stats would still fix...even with hull types.

Yep. I pondered that quite a bit. My idea was to tie immunities to ammo.

Regarding tractors and stuff, that sound interesting, but I'm worried it would be quite unfun to mentally decrypt. I would prefer "Raiders are immune to tractors. Neat." rather than "Okay, ships A and B are immune to tractor turrets because stats X > Y, but only B are immune to tractor guardians because X < Y2."

Yeah, I miss the old system, to be honest. It was far from perfect, but I wanted to improve it, not change it. Eh, I guess I just need to allocate some time for modding that stuff I wrote ages ago. Quite daunting, though.