Ok, the purpose of this thread is to provide suggestions and feedback on the new Core Shield Generators mechanic. Background on the mechanic:
One more note: as always, I want to make it clear that I'm not saying that I'll never change this, or that I won't consider other ideas. But, the suggestions that have come up so far are already things I've considered and discarded for good reasons, and/or which otherwise violate the core design principles of this thing I'm trying to accomplish.
To make things easier to discuss, here are those core design principles for this new mechanic:
1. It must require that players take a certain number of planets, at least probably 8 on the low side (on a 40/80 planet map), but more likely 10-12.
This is non-negotiable, as the ultra-low-planet-counts are something that I've been fighting for a long time and don't find to be a valid way to play this specific strategy game, because it's just too exploity and/or relies on the players grinding away at sacrifice of their personal time in exchange for the satisfaction of having pulled one over on the game ruleset.
I originally tried to combat that with the energy reactors having lowered efficiency per planets, etc, but the ZPG sort of bypasses that and players were able to work around it anyway. But fortunately, the energy reactors are at least required to be spread out (and thus more vulnerable), which was a secondary (but almost equally important) goal of the reactor efficiency thing. So that mechanic still has purpose and stays.
2. It must prevent the player from having any hope whatsoever -- I mean black and white 0% chance -- of winning the game before #1 has come to pass.
This is the one that people have the most issue with, I think. They don't like the feeling of options being taken away. But "please play the game" isn't really an ultimatum when it comes to game design, I don't think. Had it been this way from the start, people would never have complained, I feel quite sure. Who's complained that it takes 70 stars to reach bowser? What, you only want to do 50 before you quit playing the game? Why not just do 50 and stop if that's how you feel, honestly -- it's about the journey, not just about finishing as fast as possible.
And there are, of course, those who feel that this is just padding out the game. But I hardly think so: the majority of players already play in this style, anyway, where they take a certain number of planets. It's the natural way to play unless you're trying to turtle or you're trying to rush, both strategies of which the game actively -- and unapologetically -- tries to quash. "Hi, my name is Chris, and I'm a turtle. If the game let me, that's all I'd do." Seriously. The game is built to break that habit. And it's similarly built to break the rushing habit. Or at least it's supposed to be. This is a 4X at heart, not starcraft.
3. It must not restrict players to taking a specific arbitrary set of planets.
Now THAT'S a horrible reduction in choice. Having to take the two AI homeworlds is different and interesting, because they're special and built unlike other planets. They're the final boss. No problem. Having your roadmap laid out for you on what exactly other planets you have to take on the way there is incredibly lame.
That's why there are networks here. Sure, it makes you take 4 out of the 5 ARS planets, but again you really ought to be doing that anyhow. For the rest of the networks, it gives you a choice of at least two planets, and often quite more than that. It covers enough territory that you can choose your various four secondary-group planets from half or more of the galaxy. If you're telling me there aren't four planets in half the galaxy that you want... well, that comes back to being either a turtle or a rusher. The mechanic is here to help.
4. It must not be ridiculously over-complex to play with, or to code.
Clear enough, I'd imagine. We have limited time, and while this issue is important it is only important for a subset of players: since, again, most players would already be playing in such a way this would barely if at all impact, anyway.
5. Players shouldn't be able to just turtle up in one corner with all their planets next to one another.
This one was a late-add oversight. But it basically answers the question of "why not just make players capture x number of planets to bring down the shields." The answer is that, as hinted at above, this is meant to prevent players from turtling. If all the rule says is "your big fortress must be bigger before you raid the AI into oblivion," that means very little, right?
An example: in Chess, you can't just keep all your pieces on the back three rows, safely ensconced, while your queen rampages around the board. That wouldn't be much of a game. But, if players could just rebuild their queen every time they lost one, given enough time passing, that's what many might do. It becomes a war of attrition. And those are really lame. That's what many RTS games are, and I hate that. That's not strategy.
In the case of AI War, the fact that your pieces are not a finite resource is a necessary byproduct of the expectations of the genre, but it means it also comes saddled with the usual baggage of the genre. Players don't want to open themselves up for attack if they feel like they don't have to, and players don't want to feel like decisions they make are choices that are one-way doors. Those are uncomfortable. But that's where a lot of the true strategy comes from. If you make a mistake and then just hit the undo button, where's the strategy? Real strategy is about risk and reward, and it's true that minimizing risk is a huge factor there, but risk shouldn't be able to be eliminated all together. And that's what this turtle-raider strategy lets players do.
Further, to be clear:
It might sound like those three non-negotiable goals are just a way of saying that I won't consider ideas other than the one I implemented -- which is suspiciously tied to all of them, right? That's not the case. However, that's a pretty exclusive list of non-negotiable conditions, simply by nature of the problem I'm trying to solve and the larger mechanics of the game into which the problem fits, which means there are a pretty limited number of possible solutions. That happens sometimes in game design.
However, as the game evolves, new possibilities might open up. That happened with the hull types, with the shot mechanics, control nodes, and with various other things where I argued against the community and later changed my mind. At the time, I maintain I was right not to change things. Too much was resting on those particular supports to remove them or severely alter them. There wasn't a clear other path that didn't have serious flaws or shortcomings. But as the rest of the game evolved, things got to a point where the players were overwhelmingly right, and I had no reason to object. This is an evolving game.
That is all. 
Existing suggestions from previous thread:
Dreadknoght:1. Ether there be a toggle for the Core Shield Generators, for the people that doesn't find it that intriguing.
or
2. Make it so there is some kind of gain for destroying the Shield Generators, for the people who are a little sketchy on the idea (I.E. A few new unit types that you unlock (not necessarily powerful) when you destroy it).
HitmanN:Player captures X number of planets, AI homeworlds' shields go down. No CSG's or whatever needed. Simple, straightforward, doesn't force choosing specific planets, simple to code, etc, etc. It matches all the criteria.
Perhaps the AI decides to convert the invincibility into dynamic forces (reinforcements) instead, turning off the invincibility, and ramping up reinforcements. You'd get a nice boost of difficulty along with the gates to the AI home opening up.
The only thing this solution lacks is making ARS's a mandatory capture, but I still think that should be a choice anyway. For instance, if you want to take high-resource planets instead of an ARS planet, then that should be up to you to decide.
Kemeno:Here's an alternate mechanic that I *think* accomplishes much the same thing as the core shield generators, but keeps player options there.
Instead of forcing the player to destroy shield generators, make it so that the AI homeworlds cannot be attacked until the AIP reaches a certain level - maybe 200-500 or whatever. You have the freedom to do whatever you want to get the AIP up to that level (just throw some nukes around if you want), but if you try to go onto an AI homeworld before the AIP hits that level, the AI either A: launches *really* nasty waves at you (maybe these are special waves, or maybe the AI just gets some MASSIVE but temporary effective AIP boost until you get off it's homeworlds) or B: has shields so you just can't hit the homeworlds period. An in-game justification for the former method is easy: the AI essentially panics because it thought it'd reduced the humans to nothing, but suddenly they're on its homeworld? Better send in the special forces to quell the rebellion... but after the AIP hits a certain point, the AI perceieves the human threat as 'nominal' so it doesn't do this. Or maybe the AI home planets are invulnerable because they have a delicate home shield grid reinforced through its gate network or something - increasing AIP usually means you're taking planets, so you're destroying gates. Hopefully someone else can think of something better - the mechanic is the interesting part :-)
Perhaps 2-3 (maybe more) AI "Threat Assessors" are seeded throughout the galaxy, which reduce the AIP needed to assail the homeworlds by some amount - so you still have the freedom to play a relatively low (or lower) AIP game if you want to. But you have to hold these things in order to reduce the AIP needed, and if the AI ambushes you and you lose control of the planet, you could find yourself facing some very nasty AI waves. or in the best case simply be unable to assail the AI homeworld until you recapture it. Oh, and in order to get the AIP reduction, the Threat Assessor needs an active link into the AI warpgate network (needs to be adjacent to an AI planet with a warpgate), so the planet with it WILL be threatened by the AI.
Optionally, it could be mandatory to take and hold one or more of these threat assessors to hit the homeworld.
Here's an example of how this might work:
You need 500 AIP before you can attack the AI homeworld. Each "Threat Assessor" reduces the AIP needed to hit a homeworld by 50 (could be anywhere between 50-100, depending on the number of these things we have). You want to hit the AI early, but not too early, so you take one of them. You now have to spread your forces out to defend the Threat Assessor while you hit the AI Homeworld.
Superking (not from previous thread, but also very relevant):When I browsed over it I assumed cool, dispersed targets across the galaxy that weaken the AI in various ways. In reality its just a compulsory duckhunt that must be fully completed in any order before taking a homeworld with no element of player decision... there is so much more that could be done with this!
eg.
each generator could spawn in two or more places (two halves), one on the AI homeworld and one or more elsewhere in the galaxy. The ones elsewhere in the galaxy provides invincibility to the AI homeworld half, but if destroyed also scraps the AI homeworld & other halves; the AI homeworld half provides a planet wide effect such as
Regeneration - Has a large but finite bank of HP that must otherwise be depleted)
Speed Booster - until destroyed, AI units in system move at 2x speed
Hive Beacon - when the planet is alerted, all hybrid hives receive an immediate waypoint back to the AI homeworld
Zenith Portal - as long as the planet is alerted, the Zenith Portal receives a steady stream of BP. When the BP reach a high enough number, it produces a random (freed) AI golem and resets
Gaurd Station - as the zenith portal, but instead produces random MK V AI gaurdians
Counterattack Node - AI ships become enraged when player units enter the system
AI Command station invincibility
Reclamation - functions like a stationary, very long ranged leech starship (eg, firing large numbers of weak projectiles that tag player ships)
with such effects, an attack on the AI homeworld can realistically be attemped without destroying all of the other halves, but hunting them down first will make it easier. this could lead to situations where the player has to decide whether to press on, or delay and hunt for more CSG halves. this is all throw away ideas but it's just one of numerous more interesting applications of a treasure hunt mechanic Smiley
Winter Born:An alternative would be:
The impregnable FF on the AI homeworld GP's go away after your "planet cap score" reaches or exceeds the highest AI dificulty in your game (If playing with a 4 and 5 AI then 5, if a 9 and 10 then 10, etc)
cap any bloody planet you want
Planets adjacent to your starting planet count 1/4 of a point each {disincentive to turtle)
Planets not adjacent to any other friendly planet 1 point
Planets adjacent to another friendly not your home planet 1/2 point
for example a chain might look like this
HW - 1/4 - 1/2 - 1/2 - neutral or hostile - 1 - neutral or hostile - neutral or hostile - 1/2 - 1/2 - neutral or hostile
3 1/4 points earned 5 planets capped if playing AI7s then another 3 3/4 points required to open the gates so to speak.
********************
Ok, hopefully that covers all existing suggestions. If I missed any from the old thread, please let me know.
Edit - added suggestion from Winter Born that I overlooked - thanks for letting me know!
Poll