I think military star bases should have 25% more HP and it's ZOC start at more than 5 hexes. Or maybe if a military star base is nearby and in a planet's ZOC, players must destroy the military star base first before they can attack the planet, invade with troops.

I believe that military starbases are mostly ignored by most players.  I have only used it once, and it was ignored by my enemy.  Stardock needs to nerf this mechanic to make it useful.  I'll note the following issues...

1. Military starbase enhancements come too late in the tech tree to be really useful.  By the time you get to them, battle fleets are more better.

2. They should have more attack & defense abilities than regular starbases.

3. They should have a larger ZOC than regular starbases, and they should come already stocked with perimeter scanners.

4. They need to be dangerous enough to enemy races that these races are forced to attack them (rather than simply ignoring them).

The Dev team needs to put their heads together to determine how to make military bases both useful and fun.  As things stand, the players ignore building them and, if they are built, the AI ignore them in the game.  Simply stated - this mechanic is broken.

I think thats actually a good idea.  Instead of parking your fleet at your planet for a last ditch effort of planetary defense, you can park it at the military starbase in the planet's ZOC.  Planetary Defenses which help fleets are only in Age of Ascension, but military starbase ship support comes in the Age of War.  That could help the AI focus its defenses when its getting overrun.

Military SBs are glorified garages and little else.  Mil SBs provide viable attack bonuses to battles within their ZOC but this is far less useful than it sounds in practice.

The single most impactful thing that renders Mil SBs pretty much pointless isn't how weak they are, it's how much of a non-entity they are when it comes to being an effective defence: attackers can just swoop right past them and ignore them.  Later on in the game when fleet moves run up into the many dozens (and low hundreds if you go balls to the wall on moves) fixed defences are of no use whatsoever because everything just swoops right past them.  Even the moves reduction modules do nothing to prevent an enemy fleet from flying right on by.

GC3 is broken in so far as its turn based system is concerned once moves go above ~8 because there are no interception mechanics to stop stuff blitzing in from beyond FOW in many cases, weaving through all of your defences, and slamming right into that high value target that they had their eye on.  There is no way to defend an empire against an attack force that can move so far and have its pick and choose of targets within its moves range.  (Most people don't see this happening because the AI isn't able to exploit this, but you've surely been able to take advantage of the AI with huge moves fleets, and as soon as you play against a Human the problems that GC3 has pile up very rapidly indeed.) 

Mil SBs could have a valuable role to play in fixing this fundamental problem by means of interception, i.e., as soon as an enemy fleet enters the ZOC of a Mil SB an interception battle is started or at the very least the enemy fleet loses the remainder of its moves (or some variation on that theme), but they don't ergo they're a waste of resources and effort (cf. constructor spam management).  A better solution would be to do what every other sensible TBS game does and give everything with a gun the ability to intercept enemy moves as soon as they come within a certain range, and Mil SBs could have the largest of such ranges thus giving them functional strategic value and perchance making them worthwhile.

Strongly agree that military starbases are nearly useless.  They become *much* more useful (and I don't think this is imbalancing, either), if they can be set to automatically engage enemy ships that enter their ZOC.   You can build a defensive line with them that way.  Otherwise (as has been noted), they can simply be ignored, which is ludicrous.

- They should be able to engage ships within their ZOC
- They should have improvements available that expand their ZOC
- They should be able to build up weapons and other components comparable to large or huge ships.
- They should be able to build up very heavy hitpoints and armor with additional constructor modules

It's always bothered me that there are only a few scanner upgrade levels for starbases, but I can build ships with epic scan ranges of 50+.   I would expect starbase listening posts to make more sense than putting ships on station on my empire's frontier. 

And if you happen to enter a ZOC that is overlapped by multiple military starbases, you get to engage multiple starbases -- it should be possible, if you care to expend the resources, to build monster fixed defenses.

Why not minefields around a SB to slow down a enemy fleet ?

Yes, force fields are science fiction and the good old minefields are obselote in space, but perhaps usable on the planet itself ?..or are there also a force field possible what prevent enemies to invade.  

As far as forcefields capable of preventing invasions go, this is possible in settings which allow fully-shielded planets and shields which can block the passage of or destroy large objects. The Star Wars EU has examples of such, the Star Wars movies debatably have examples of such and provide circumstantial evidence for the Star Wars setting to be capable of such, and GCIII may also have such in the Elerium Defense Shields and perhaps the Planetary Defense Domes.

As far as the minefields go, I have no particular objection to minefields of the kinds used in the real world, but as such would best be represented as a planetary defense or resistance bonus and could not sensibly slow hostile fleets, I don't see how this relates to the previous suggestion of slowing hostile fleets with space mines.

As for the implicit suggestion of making it impossible for planets to be invaded as long as there is a sufficiently upgraded military starbase within range, I am completely opposed to the idea. Make invasions more costly? Fine. Make it likely that the cost of the invasion will be so high that invasion is impractical without taking down the starbase? Fine. Make invasion quite literally impossible without destroying the starbase, though? No. GCIII has issues with the availability of invasion immunity already, and making it more available won't help, though at least a station-based invasion immunity is vulnerable to attack whereas improvement-based invasion immunity is not. If you do not have enough ships in the area to control the space around your planets, your planets should be vulnerable to invasion. Space stations of any type should not change that, no matter how heavily upgraded they are.

No worries; everyone forgets things from time to time, especially when the thing in question is part of something which sees little use.

a work around might be to create a whole new starbase ring and place range extension and slowdown modules only in that ring
if you want a tarpit starbase you need to have a tarpit ring which prevents you from having military/econ rings and all this ring does is give you range and slowdowns 

Let's consider for a moment just how large a volume of space each tile represents. According to this post by Draginol (Frogboy), the minimum dimension of a tile is ~0.02AU or ~10 light-seconds. If we're very, very generous, GCIII ship combat ranges might be roughly 0.4 light-seconds (beams, which include lasers, take about 0.2 seconds to propagate from point of origin to target when fired at maximum unboosted range, and the maximum possible missile range is roughly double the unboosted beam range; you could also argue for engagement ranges more like tens of meters to a couple kilometers based upon the sizes of the ships and the apparent distances between them as seen in the battle viewer, and put the laser component in the pile of science fiction 'lasers' which aren't true lasers). If you're very, very lucky and happen to be mining the minimally-sized tile, you might be looking at ~600 mines per sheet across the tile, with several sheets of mines required to effectively block travel through one face of the tile for any significant length of time (moving parallel, or nearly parallel, to the minefield is not going to be hampered). If you want to block all travel through the tile, you're probably looking at something more like 16000 mines just for the smallest possible tiles in GCIII. If you assume that the average tile size around a star with a system similar to the Terran home system is ~1 AU (and in my opinion the average tile dimension is likely far greater), you're looking at needing ~2500 times as many mines per sheet (~1.5 million mines per sheet; roughly 100 million mines per tile with 25 sheets per field using two or three mutually perpendicular fields to block passage through the tile, or ~2 billion mines per tile if you want to fill the tile). Once you get out into interstellar space, the tile dimensions probably jump up to something on the order of a lightyear rather than an AU, at which point you're looking at something on the order of 1e15 (one quadrillion) mines per sheet across the tile. Mining a full tile, even with generous assumptions about mine spacing, requires an enormous investment of resources and a lot of time. This is quite simply not practical, especially if you want to apply it to every tile in a starbase's area of effect (without AoE upgrades, a single starbase covers 90 tiles, not including the tile in which the starbase is located; the first AoE upgrade increases this to 168 tiles, and adding the second increases this to 270 tiles).

Then there's the issue of whether or not the mines are sufficiently difficult to clear to actually provide a practical impediment to a ship's progress. Unless you have near-perfect containment, anything carrying a significant amount of antimatter is going to be emitting fair amounts of high-energy radiation, which should be relatively easily detected and strongly indicative of the presence of something unnatural. Anything carrying fissile material is likely to be radiating due to the ongoing radioactive decay in the material. Anything with an active station-keeping system (which would likely be necessary for mines in free space rather than in orbit around a body) is going to be quite visible and easily distinguished from natural space debris. Anything with an active power supply is likely to be fairly detectable and readily distinguished from natural space debris. If the mines are readily detectable, that removes the single greatest obstacle to clearing them; now they have to fall back on being difficult to target or present in such great numbers that clearing them is very time-consuming.

There's also the issue of whether or not ships travel with inactive defenses, whether or not a mine is capable of penetrating those defenses if active to a degree sufficient to be worth the cost of the minefield, whether or not the time between the mine detecting a target and detonating is long enough for the target to detect and respond to the mine's activation (you're proposing a mine which orients itself towards the target; it's going to be quite visible for at least a short time before it detonates), how many mines a ship must trigger before the damage sustained becomes problematic or causes the loss of the ship, and so on.

There are scenarios where space mines can make sense, but Galactic Civilizations III is not in general a setting in which space mines make sense; at best, you might do something like set put up an orbital minefield around a planet to deny certain orbitals or impede passage through certain orbitals, but this is going to have detrimental effects on civilian traffic to and from the planet and is unlikely to be something you can set up rapidly enough that you'd only have the minefield while the planet is under threat of invasion, and that minefield would be in the same tile as the planet, not something impeding passage through an adjacent tile.

Very much so.  Why can't a mil base have hanger and send ships out to help?   Sure seems like this would be a tricky thing to do (not that I would know )  are you going to release it as a mod?

The msb is coordinating and financing local squadrons and also detecting enemy presence early enough to send escorts. 

Or at least that's one way to justify it. It's all abstract.