0.16 and the future of the Grey Ghost

I was unable to work on anything game dev related for the past month and have lost momentum considerably, but it's quite natural given the events that unfold around here. I've made some slight changes to the prototype today and released the version 0.16, which would probably be the last one.

This prototype was always meant as a practice run and not intended to grow into a fully fledged game, but let's imagine for a second that it could've become something more. Original questions that I was going to answer by making the Grey Ghost prototype:

• Is it FUN to play? What other mechanics are needed to make it better? Can you prototype/add them fast to test?

It is fun to play, if carrier warfare is your jam. Otherwise, it's just "OK". Fog of war provides the thrill of not knowing where the enemy is and being scared of unexpected attacks, and there is joy in finding & striking him first out of the blue. Direct control over your task force gives more focused feel to the gameplay, but the opposite is also true - there's no time to look at the ammo/HP numbers when you're maneuvering and trying to aim/dodge attacks.

More knowledgeable players would probably welcome addition of more advanced logistics, presence of surface warfare forces, and an intelligence screen a-la WITP AE, but I was not able to implement them in a timely manner to test.

More action-friendly UI would also be great addition to allow for smoother, uninterrupted, fast-paced gameplay. I'm not a UI/UX specialist, but I can clearly see it as a crucial area that requires attention and thus needs to be looked at, probably with additional, smaller prototyping phase that just focuses on the UI/feedback aspect.

• How should the game code architecture look like? What design patterns should be used for various systems? Is it going to scale well performance wise and gameplay wise?

This isn't my first Unity project, and I managed to lay it out pretty decently overall, but the way I coded AI was really messy and would require a complete rewrite if such game was to be put into production. Well, it's a common practice to throw away prototypes altogether when you switch to production, but given that I'm a solo dev, there's a certain tradeoff here. Refactoring certain pieces of code would be less time consuming. So, a complete rewrite of AI logic, using either a Behavior Tree or State Machine, would be needed.

NavMeshPlus/2D NavMesh by h8man for Unity did not give me much trouble, aside some slight inherited bugs. I was satisfied with the results, but did not test the scalability, unfortunately. It would require additional, dedicated tests.

The structure of inheritance trees for map entities' classes and the interfaces for their modules turned out to be pretty simple, but would also require some decent refactoring. The speed of changes and new mechanics added during prototyping made me write some standalone classes and a bit of copy-pasting was done in the heat of the moment, which would have to be addressed. But it's fixable.

• What art style better suits the game?

My initial idea was to go all-out abstract on everything, including graphics. I sketched some weird looking unit icons and their "abilities" on paper, even though I intentionally chose to implement gameplay based on carrier warfare circa 1942. But the more mechanics were added, the harder it was to explain for the player how and why it worked the way it did. It was much easier to put in icons of ships and planes and, voila, it all makes sense now without much explaining.

The question of unit sprite projection (top-down vs side view) came up during sketching phase and I quickly realized that using only one would be unwise. Aircraft would have to be certainly drawn from above, but what about ships? The ships, on the other hand, and the player's carrier icon in particular, if it was going to be moved around using only WASD keys and no aiming is involved, a side view would work better. Ship silhouettes are better recognized in this manner, while top-down recognition and classification of units is more problematic. So a mixed approach was chosen: top-down for aircraft and terrain, and side-view for ships/bases.

I do prefer sci-fi setting much more game design wise, because it allows for more freedom and suspension of disbelief, while still being able to stand on the shoulders of existing history and using real world knowledge. But in this particular instance, 1942 Pacific War styling would be enough.

Here are additional questions that arose during prototyping:

• Should VPs granted for damage / destruction of units be delayed?

I ended up delaying VP awards because of my previous experience playing WITP AE and it felt more "right" to me personally. Fog of war means there's Battle Damage Assessment that needs to be made, and intelligence reports aren't instant. BUT. I feel like the delay made the game less "fluid", like it's just some arbitrary, artificial annoyance. I've decided to keep this feature, but more feedback and thoughts could change my opinion on this topic.

• Destruction FX for ships? How best to do it?

Again, with fog of war being a major factor, I don't know for certain! On one hand, player needs more feedback here, but.. should he actually get it? It's a follow up on a previous issue, really.

• How to deal with speeds/distances balance, and, more importantly, what to do when your "scan" UI that's suppose to reveal enemies does not fit the screen?

When the carrier is using its "long range scan" ability (aka sends scout planes to search for stuff), the player unit sends a circular wave that "scans" outwards around in 360 degrees. But this is an action game and the camera is zoomed in relatively close to make the player icon feel big. The difference in ship and aircraft speeds and their ranges means that for the "long range scan" to actually be long and meaningful, it needs to go out really far. And so when the player unit icon is quite big, the max scan range exceeds the visible area of the screen. At least in vertical aspect.

The easiest way to fix it would be to highlight targets with red arrows near the edge of the screen if something was found but wasn't actually shown on the screen.

Anyway..

It's time to wrap it up and move on.