Fair enough. Honestly the fleet concept from Gal Civ III automatically PREVENTS the gigantic stacks of doom we saw in Civilization series.  A fleet can only get up to a certain size. This is a HUGE advantage of Gal Civ III over Civ IV.  

But the basic concept of "concentrate your forces in safe locations," and then launch them together as a team remains the same.  

Personally I think the next Civilization series should incorporate Gal Civ's use of "fleets."  The reason Civ V is considered not as good as Civ IV by many players is because it lacks Civ IV's strategic depth.  And a lot of that strategic depth came from trying to handle those gigantic stacks of doom.  Civ V required the AI to be tactical, which is normally an impossible task.  

The main difference between galciv3 and civ's units is,
that in civ no signle army/fleet was able to move 8 fields into your territory to pick of single units running around.

In Civ territory did often cost more than 1 move, and you weren't able to use roads on enemy territory.

In GalCiv thats no problem. I can even move into enemy territory _before_ a war with no systems to prevent it.

I can even move next to your so-called "safe" shipyard. Nothing is safe in GalCiv, the AI can't cope with that.

Moving anything with attack > 0 into my area of control should equal a declaration of war, except there's open boarded treaties,

and why this isnt in the game I can't fathom.

Erm, I believe it is in the game (or something similar, anyway) - you have to get the UP to agree to it, though.

AI rally points should be set at their homeworlds.  Without the concept of borders, that's the safest place to place ships.  

^ this +10

If you look at the game map and think about what the tiles mean in terms of real distances, you will see that something similar to this is already in place. It doesn't reduce the number of actions you can take per turn, but it's abundantly clear that the tiles are greatly distorted by the presence of large bodies such as stars. A planet can be as close to a world in a completely different system as it is to another world in its own system (and sometimes it can be closer to that planet in some other system than it is to a planet found within the same system as it is), as far as the number of tiles between them goes. We can also establish an upper bound for one dimension of at least some of the tiles in the Sol system as we know the maximum orbital radii for each of the planets which can be found within the Sol system; the tile normally between Earth and Sol, for example, has at least one dimension which cannot exceed ~1 AU. We also know that there are no stars closer than ~4 ly (~253000 AU), meaning that we can establish a lower bound on the average tile dimension between Sol and the nearest star to it. Last I checked, Sol and the nearest star to it were separated by something far less than ~253000 tiles in game, which means that 1 move action taken "close" to Sol (or Earth, or some other large body) represents a significantly lower average speed than 1 move action taken "far" from Sol (or Earth, or some other large body), to the point that it's likely that the in-system speeds of our FTL drives is actually less than the speed of light.

Beyond that, we're talking about a system which permits a vessel to travel at speeds which are in excess of c (or which are effectively in excess of c for the purpose of getting between two points in real or naturally-warped space even if the vessel does not actually exceed c within an artificially-warped bubble of space around it or in some space which is in some way not 'real' space, e.g. hyperspace). Justifying a reduction in actions per turn because real-world craft using real-world technology to escape gravity wells is not exactly the most well-reasoned justification out there, especially since it's quite obvious that there are already significant effects upon the speed of the object relative to unwarped (or naturally-warped or real) space.

Why exactly nebulae and asteroid fields are special and have additional movement and sensor range costs, I have no idea. Presumably it has something to do with the way the drives and sensors function and may be a way of representing the size of a nebula more 'realistically' without having the nebulae sprawl across enormous sections of the map (nebula tiles take two move actions to enter rather than one; one might reason from this that they are twice the 'size' of a normal tile in a similar position) or it could be a way of representing the effect of very low density high-mass bodies upon the GC hyperdrive at great distances from dense high-mass bodies (nebula tiles are typically, though not always, relatively distant from tiles containing stars and planets, and therefore would normally be expected to be quite large; ships would therefore normally be moving quite fast in such tiles, with the nebula effect cutting the apparent realspace speed in half; the ship is still probably moving much, much more quickly than it is in tiles in close proximity to planets and stars, but it's also moving much less quickly than it would be expected to in 'empty space' due to the relatively high mass density in the nebula as opposed to in 'empty space').

the reason closed borders is not in the game is because no one wants Haiti being to throw out the us navy. You should have to be able to throw me out to throw me out. If you don't. Have the force then you will have to deal with it. Closed borders is a bad mechanic for civilization and endless space.

Which starbase effects need modification?

Culture starbase effects? Influence doesn't make any real sense anyways, so who cares.

Economic starbase effects? The only sensible way for a starbase factory or resort to function is for there to be cargo/passenger ship traffic between the starbase and the worlds under its effects; increases in absolute tile dimensions are cancelled out by the corresponding increases in ship speed. Labs need to be sufficiently close that you have at minimum a sufficiently low-latency communications link between the starbase labs and the planetary labs to enable effective collaboration, though you likely have that to most if not all points within your empire anyways; the more sensible limiter on lab range is once again the need for ships to go to and from the starbase, this time carrying test samples and other resources for the labs instead of raw materials for the factories and refineries (and refined or finished products back to the planets), and so you once again have a situation where the variation in absolute tile dimension is cancelled by the variation in ship speed. Markets can either go with the communications link or with actual ship traffic; once again, if it's ship traffic you have a situation where the variations in actual tile dimensions are cancelled by the variations in ship speeds. I don't see a sensible way for a trade route value bonus to work except if the station works as a handling facility, which means ship traffic. Nothing here is difficult to square with the effects of mass upon tile dimensions.

Mining stations? The only rational way for a mining station to function is to send out mining ships to the resources within range and have them gather resources and haul them back, or alternatively for it to function as a handling point which both provides the supplies for and collects the resources from smaller mining outposts at the actual resource site. The variations in speed and tile dimension cancel out, so there's no issues to resolve here.

The modules for investigating relics and ascension crystals? Probably the same deal as mining stations; you're either sending probes and low-duration expeditions to or supplying research outposts at the relics, in which case it's ship traffic, or you're establishing an on-station lab which analyzes some kind of sensor readings (if so, then likely with special-purpose sensors rather than the ones used for detecting ships; after all, you can use the station to study a relic without it being within 'normal' sensor range).

Military starbases? The effects of military starbases are not sensible, at least not by the implied causes to which the effects are attributed (i.e. station-based effect generators). You used 'inertial dampeners' or 'mass-lightening technology' to allow your mass drivers to be accelerated to much greater velocities than they should be capable of moving at? Great, but when you let that mass come back, guess what? You (should, if it were to make sense) lose the extra speed, which means you may as well have just used a lighter projectile in the first place. No, the attack and defense boosts offered by starbases make far more sense as a "we can expend munitions and resources more freely because we can be resupplied more rapidly from the nearby starbase" than anything else, and resupply depends on the ability of some kind of cargo craft to traverse the distance between the station and the ship(s) under its effects. If you do want to go with the implied causes of the station bonuses, then the justification for the station bonuses can be summed up as "magic," and as far as I'm aware "magic" does not need to follow real-world logic, obey any process which can be explained by science, or operate according to the limitations imposed by reality.

As far as the speed modifiers of a military starbase, I would ask where adding a large amount of mass (or adding something which has the same apparent effect as adding a large amount of mass) makes a bigger difference - in a region with a lot of mass already present, or in a region where very little mass is present? I would expect it to matter most where little mass is present. If you model the warping effect as a right triangle with x being the distance between two points in 'real' space, d being the apparent difference in elevation in naturally-warped space as seen by the hyperdrive, and h being the increase in apparent difference in elevation in artificially-warped space (i.e. space affected by a Gravity Field Generator on a Military Starbase), where does a constant h have the greatest effect upon the apparent distance between the two points as seen by the hyperdrive? Why, where d is small, i.e. where the 'surface' of space is 'flat' - in other words, in places where there isn't already a large amount of mass present. At larger values of x, the station's effects are simply being obscured by the preexisting effects of the stars and planets already in the region. If space is a sheet of rubber and a gravity field generator is a marble, which changes the sheet's deformation more? Adding 1 marble to an existing collection of marbles on the sheet, or adding 1 marble to a sheet which does not have any marbles upon it?

This; while I do get the whole idea of the tiles being different sizes etc (although I find the idea that we can excuse some mechanics working through 'magic' but need to stick rigidly to physics for others a little distasteful), game play should always trump lore when the two are in conflict. And if we need something to slow down late-game ships just to provide some reaction time to the defender (which I think we do, tbh), then that needs to be implemented somehow; gravity wells for stars would do the trick and could be excused with a bit of physics that's only a tiny bit more hand-wavingly vague than relativistic tiles.

It's called "a fleet stationed at the starbase."

Very little of what I posted sticks 'rigidly to physics.' It does keep the game internally consistent, however. Your solution is to increase the cost of taking an action close to a star on account of its gravity well distorting the space. The game already accounts for the distortion due to the gravity well; this can be seen simply by thinking about what the tiles mean in terms of real distances.

Besides which, you are proposing to increase the cost of movement through tiles close to stars as a solution to late-game ships being too fast. What you are proposing is effectively 'nebulae, but around stars.' Nebulae are mostly only an obstacle to slow ships; a ship with a large amount of movement per turn more or less doesn't care that each nebula tile takes 2 actions rather than 1 to traverse. Let's say the 'gravity well' affects 5 rings of tiles and that your planet/starbase/whatever is in the second ring from the center. I will therefore require 5 move actions to get from beyond the outer edge of the 'gravity well' to my target. To which kind of vessel does this make the most difference - a late-game ship with 'dozens and dozens' of actions per turn, or an early game ship with maybe 10? How much of a difference does this make in the amount of warning I get? Sensor range of R, minimum warning time of floor((R + 1 - M) / M) turns for standard-cost tiles or floor((R + 1 + N - M) / M) when there are N tiles that each cost 1 additional action point. How many turns does it take to traverse D tiles? (D + N) / M, where N is the number of tiles which cost an additional action to traverse. Where does the extra movement cost make the bigger difference? When M, the number of actions per turn of the oncoming hostile target, is small. In other words, it makes the biggest difference to the early-game ships. You are suggesting this as a solution for late-game, fast ships, when it is upon those ships that this has the least impact. You are apparently fine with the current situation for early-game, slow ships, upon which this change would have the greatest impact.

Given that you are fine with the current situation for slow, early game ships, and given that the thing you primarily wish to adjust is the effective speed of fast, late-game ships, don't you think it's just a little bit backwards for your solution to hit the slow, early-game ships hardest?

If late-game ships can take too many actions per turn and early-game vessels are reasonable, then the appropriate solution is to reduce the effectiveness of the late-game engines, not change the tile movement costs so that early-game vessels are penalized greatly while late-game vessels are penalized to a rather lesser degree.

That's why what I'm proposing is a % reduction of movement allowance in the gravity well, rather than an increase in movement cost. And the movement is taken from unused allowance first.

If you have a ship that moves 4 spaces per turn and it enters a gravity well, then it's movement is reduced to 3, and remains at 3 until it exits. You have lost 1 movement per turn. If you have 20 moves and you enter the gravity well, then your maximum speed is reduced to 15 moves. And if you had 10/20 moves left when you entered the well, you now only have 5 moves left. 

This would benefit the game in that it limits the range that a fleet can rush-in from, and also makes star systems into an actual terrain map where a path that crosses many systems is much less efficient than a path that only crosses a few. That's both more useful and more interesting than just nerfing late-game engines.

I don't know whether to cry or sob. This is not the type of news I was hoping to hear from someone about a month after release. Given all the other fronts where the game needs improvement it feels like a major rework is needed. It is really too bad.