I would give the planets each some command point(maximum fleet size) then let the player set the exact amount of ships in every fleet and just let them be build.

But apart form it, I wonder, why is there such "resistance" against any proportional/percentage system? Isn't it more fun than setting equal units for each fleet? Or maybe the whole concept of "let the player tell what kind of fleet he wants and the game will try to build one for him" is flawed? Or maybe it's too confusing?

I'm asking because I want to establish if it is only me who likes this concept and normal players would not find it appealing at all? Or if there are only objections to details (how exactly it should work) but the overall concept is very appealing to you?

"The disadvantage for automation has always been if a human could do it better, they will perceive an error."

Only if the automation is an option, not if automation is the only way to do it.

Do you remember a game on Amiga called Lemmings? These blue "peoples" were beyond total idiots. And the player's job was to not make them (all) die horribly and in a completely stupid way (like falling down the cliff they obviously see). It didn't feel wrong or an error. The imperfection of the automation was the part of the game (actually the core of that game :D).

You could not micromanage there, you could not manually order these stupid lemming to turn left or right, you could not take over a single lemming (each order was a limited and scarce commodity). The automation there was not an option, it was the only way to interact with the game.

"so the player has even less control unless" and "to get it exactly right"

Yes, less control and not exactly right/perfect/accurate. That's the core idea. The player gives orders/priorites and the AI executes it, but never perfectly accurate to what the player wanted exactly.

Yes, the player moves premade packs of ships, not single ships and yes, he can't send a single ship on a recon mission. Althrough, the different fleets (red, blue, yellow) can be composed differently so he could make very small/weak blue fleet and use the blue flotillas as cheap recon formations.

That's the core idea, so, what you think? Will it be fun? Is it worth pursuing?

Make it so, that at some point, let's say the player needs to put his ships in defensive positions,
he will actually want more of flotilla's from yellow to stack them on his planets,
he wants some flotilla's of grey because they're fast and puts them on the planets that could be surprise-attacked,
while he can see the attacks on other planets coming and can move grey flotillas to help defend(but not yellow, since they have tiny engines and are too slow)

The player says that the Yellow fleet is defensive force and should have heavy defence, the Grey fleet should be fast and have the best survival pods since heavy casualties are expected due to light armour, the Red fleet is the invasion force and should have best weapons, elite crew, contingent of Space Marines and pink toilet paper because that's the favourite colour of the admiral :)

(and then the Admiral of the Red fleets gets constantly frustrated becasue the pink toilet paper is soo low on the priority queue that factories produce only some stupid shields and warp drives and never deliever any pink toilet paper :D)

And all these are "said" by the player. "Ordered" to be made. Without all this horribly tedious "let's move the newly produced Fury of The Emperor #416 battleship to fleet A, but wait fleet Z just lost a flagship so maybe move it there instead" and repeat it for every single ship every single turn :)

Same strategic decision, but almost no micromanagement. Also less perfect execution and less direct control over which small ship is where at the moment.

I'm looking for mechanic and interface that would allow something like this.

The game would be defined by more intricate details. It's my opinion that if those details didn't all come together, it would still be possible to make a fun game.

Hm? It could be a fun game even if details don't fully work together? I haven't looked at it this way... but you might be right...

This automation thing an indirect control over fleet building seems quite fun (at least to me ) and even if all together not working seamlessly it might still be a fine game...

If I make a comparison here, it'd have to be that the game Lemmings was a physics puzzle, the simulation was gravity which was meant to be circumvented by building and digging. What we might might have is an economy puzzle, the simulation being everything the player can't change directly but would attempt to through all decisions being made.

LOL, I haven't thought about Lemmings as physics & gravity puzzle, but now that you mentioned it it actually was like that

As for what we discuss here being a puzzle, I only partially agree. I mean, Lemmings was pure puzzle, you have X parachutes and a map designed to require all/most of these. What I mean is a strategy with many ways of victory and random scenarios, not a hand crafted puzzle. More randomized and unpredictable I guess.

Tie the different fleets to something that seperates them.

I was thinking about it more and yeah, allowing the player making *distinctive* fleets is important.

I got also this idea, not sure if it's a goood one. How about adding a "marker" to a flotilla? Like you have this Red fleet (with all priorities, compoisition of ships, etc), but in addition you can mark that 2nd Red flotilla is supposed to be "Recon" while 3rd and 4th "Combat". So, in addition to the Fleet setup the flotilla would be altered by the "designation/subtype/role/specialization" of the Flotilla.

So the Red Fleet would have fast ships (setup) and when you designate (per flotilla) the 2nd Red flotilla as "Recon", so it would get even more fast ships (and lighter armanent) since it's supposed to fill a role of fast recon formation (of an already fast Red Fleet).

Not sure about it through...