Forts in times of distributed defenses

[Fleet Unity]

In the overall situation there's another important factor to consider: the AI's attacks can be made more suited to defeating distributed-defense.  For instance, having the core of a CPA (perhaps randomly determining what %) that comes from the strategic reserve be launched as a carrier-ball on a beeline for a human home command station; basically stopping at nothing but without the exo speed-boost stuff, they'd just individually be "galaxy-wide-chasing" the station and any ships they deployed could either inherit that individual flag or just behave normally.  Such an arrangement would be far more resistant to attrition.

But the player can then just warhead the carrier-ball and call it a day.  Granted, that's more AIP, but (Deep Thought voice) "who will that inconvenience?"

So currently I see three main things, any one of which wouldn't necessarily be a huge problem, conspiring to make the game too easy for players who really know what they're doing:
1) The expanded selection of per-planet defenses are too good for their costs (i.e. OP).
2) Warheads are overwhelmingly powerful, able to annihilate out-of-hand all but the most "sky-darkening" of AI attack fleets.
3) AIP could stand to hurt more.

Here is the problem I see with this, some players like per planet and some do not. The problem is there is no way to please everyone. Every one plays AI War differently and has their own style of play. Some see this as making the game to easy or not, I do not see how you can balance this out because some will like it and others will not. Some will think it would make you to overpowered  or not I guess its in how you decide to play the game or your skill level at the game. The only way to please everyone would be to let you decide by an option to make the turrets per planet or galaxy wide with advantages and disadvantages to both. I know this would not be easy or maybe even possible to do or not. The issue I see here is that some see it as being overpowered and some do not because everyone plays differently and is a different skill levels.

Edit: Also I like the per planet stuff as you can defend more planets if you need to, also I never usually put a full cap on every planet you would need a lot of energy to do that, and it just good to have a choice if you want to or not.

[TechSY730]

In the overall situation there's another important factor to consider: the AI's attacks can be made more suited to defeating distributed-defense.  For instance, having the core of a CPA (perhaps randomly determining what %) that comes from the strategic reserve be launched as a carrier-ball on a beeline for a human home command station; basically stopping at nothing but without the exo speed-boost stuff, they'd just individually be "galaxy-wide-chasing" the station and any ships they deployed could either inherit that individual flag or just behave normally.  Such an arrangement would be far more resistant to attrition.

But the player can then just warhead the carrier-ball and call it a day.  Granted, that's more AIP, but (Deep Thought voice) "who will that inconvenience?"

So currently I see three main things, any one of which wouldn't necessarily be a huge problem, conspiring to make the game too easy for players who really know what they're doing:
1) The expanded selection of per-planet defenses are too good for their costs (i.e. OP).
2) Warheads are overwhelmingly powerful, able to annihilate out-of-hand all but the most "sky-darkening" of AI attack fleets.
3) AIP could stand to hurt more.

I'm a little sad that this small to medium design "hiccup" made it into 8.000. I know you had a deadline and all, but still.

Also, yea, per-planet caps for turrets seem to be good in singleplayer, but become downright near OP (if not actually OP) in multi-player, especially when core turret fabs start getting captured.

Randomally giving CPAs an "exo-like" "mission" is a cool idea on its own, but I am not sure if it would solve the current "problem" (the current situation being too easy).

I don't use warheads very often, so I can't speak to their current balance.

Good point about AIP though. When the planetary turret cap changes went in, no corresponding wave size to AIP relation or AI defensive "strength" (especially over time) was really made to compensate for that, potentially making the game a bit too easy in the early-mid game.

[TechSY730]

I'm torn. :\

I like the per-planet stuff because it means that I don't need a Perfect Wall on my front lines and can actually defend some other systems if things slip through.

But I dislike the long-slog of rebuilding (a major complaint of mine for quite a while now: I have not "finished" a game since before the transports got wormhole attrition) so I don't engage in massive distributed defense layers.  Which of course means that I'm constantly losing against both the single-chokepoint counters and the defense-in-depth counters resulting in abysmally suboptimal play.

I absolutely despise using my fleet on defense for the same reason: if my fleet is defensive, then I'm not on the offense with it and the rebuild times push my fleet into only defensive capabilities: by the time I've rebuilt, the AI's rebuilt, and we clash in M.A.(fleet)D.  Rinse and repeat ad nausium.

So basically, having the defense in depth option is a great improvement (single chokepoints all the time can easily lead to stalemates), but it is expensive which can mean long rebuild times (yes, it should be expensive, that is the balance tradeoff, but that doesn't change the annoyance of having to rebuild them)?

You can use fleet to aid defense which is a bit more cost efficient, but as you noted, then the amount of time you are on offense gets cut into, also hindering your pace.

TBH, I don't really see a way out of this.
Sure, there are some ideas to make it less painful:
Maybe salvage needs a buff (for both sides)?
Maybe build times need to be looked at (especially for static defenses)?
Maybe AIP to wave size relation needs to become a bit steeper in light of the recent turret changes?
Maybe AI waves need to come more often (but still have the same size per wave)?
Maybe tweaking what is per-planet and galactic caps (like maybe we got it backwards, the expensive defenses should be per-planet, and the cheap ones galactic), and maybe rolling back part of the Mk. V turret cap buff could help too (some comments in recent AARs seem to agree that Mk. V turrets are a bit OP at the moment, even by Mk. V reward standards)?

But none of those really solve the issue, just make it less likely to matter. TBH, this sort of problem (either stalemate or really long rebuild times for both sides) almost seems like an unfortunate consequence of the core design of AI war (nonsymmetric PvE, with a high average HP to average weapon damage ratio, and a flow based economy). Then again, almost any sort of design will have some sort of unfortunate, fundamental shortcoming, which is usually balanced by fortunate, fundamental advantages (like I would argue is most certainly the case for AI War)

[Draco18s]

So basically, having the defense in depth option is a great improvement (single chokepoints all the time can easily lead to stalemates), but it is expensive which can mean long rebuild times (yes, it should be expensive, that is the balance tradeoff, but that doesn't change the annoyance of having to rebuild them)?

Correct, and it's not so much the annoyance, but the fact that until it's done I have to be less risky--or in the worst cases, bottom out the economy so I can't rebuild anything before its needed on the defensive.

TBH, I don't really see a way out of this.

Me either, hence the problem and my indecision.

[TechSY730]

(Last post for now )

So it seems like the radical shift in human defensives (per-planet turret caps) haven't had all of their "consequences" "discovered" yet, the "metagame" is still "reacting" to it, and thus more subtle design shortcomings with how the rest of the game interacts with it are only now coming to light?

Hmm, seems like a typical sort of thing to happen with a major shift.

Heck, the biggest one (much bigger than this turret and defenses thing) was the ship vs ship bonuses to ship vs hull type bonuses + shields to armor changes from 3.0 to 4.0, which took 2 additional MAJOR versions (aka, not really until 6.0) until the "fallout" from that change really got dealt with and the game was reasonably balanced and "fun" again, and people are still making some pretty good arguments that more balance work is needed or even that the current systems' flaws are big enough to make another major (but not quite as major as the original change) revision to that system.

[Hearteater]

I'm seriously considering doing the "AIP, what's that?" challenge on 10/10 thanks to per turret caps. Spider Turrets on every planet...it just isn't even fair. It is so beyond broken. Actually, I wonder if I can full cap 120 planets.  With no Exos it is possible. Is it cheating to use a Snake map. NOTHING would ever get to my home system. I want to earn "Control 110" on a 10/10 game. This seems too hilarious not to try. I might be able to start the game up tonight.

[keith.lamothe]

The result of that game would be helpful info, though it would be somewhat moreso on a non-snake map (that's been a "special case" for defensive balance for a long time).  Sounds like it would be pretty entertaining as well, though potentially the fun-level would diminish into the mid-late game.

[Hearteater]

Ok, what map type would you consider "fair" for this test? I'll use that and see how far I can go.

[keith.lamothe]

Pretty much anything other than snake would be fine.  Spokes is perhaps a good example of one that allows a lot of chokepointing but isn't "just chokepoints, all the way down".  Concentric is another that comes to mind.

Something like Crosshatch or Honeycomb would be a good example of the other end of the spectrum, which would require "defense in breadth" as well as defense in depth.

[Nodor]

The hole in the defense in depth strategy is turret energy costs.  A full cap of mark 1-4 Spider and Sniper Turrets requires 268,000 energy.   That means 2.5 Matter converters per planet.. before you place other defenses - not exactly cheap.

With the Nomad planets, you need some defenses on most planets as they disrupt whipping boy mechanics.  So we need to have some available distributed defenses or we won't have nomads activated.  That does not mean I want to rebuild even more turrets on every planet.

I will say that I notice the waves in the latest version.  I regularly need my fleet available for defense/cleanup.  (cleanup for the reprisal waves when my fleet bites the dust and I lose my front line planets).

[Hearteater]

I don't need Marks I-III. They all cost 700 energy (on Normal/Normal settings), but Mark IVs do 800 engine damage. Since I'm defending in depth I'll get all the ships eventually, I don't need to kill all the engines on the first world (in fact I don't want to do that). So I only spend 33.6k power per planet for Spiders. Since I have so many planets I don't mind burning the K on turrets I won't be using.

[Peter Ebbesen]

For the record, I hate the new per planet turrets.  The results are terrible.

Playing the 7/7 ignore AIP test game, I find myself setting up chains of planets for reprisal waves/CPAs to plow through.  My defenses are designed to attrition the "nasty teeth" and do so effectively.  I plan to lose about 6-8 planets during each CPA and it takes 45 minutes or so to rebuild them.. but aside from grinding forward, I'm not feeling like the waves pose any threat.. even if they take out a planet or 2,  and this is possible with level 1 turrets and just mini-forts at AIP in excess of 500.

Pardon me, but this part - I cannot comment on the multiplayer part - reads to me as an argument in favour of per-planet caps, not an argument against them.

After all, what you describe certainly sounds like something that can also be done with per-galaxy cap turrets, rebuilding the full set of level 1 turrets as each individual planet in the chain is falling and the reprisal wave/CPA and moves on to the next.

Doing it with per-galaxy caps requires you to save up metal in advance in addition to the substantial metal income you'll have from harvesters and salvage due to paying for the construction of the n planets worth of level 1 turrets over a shorter period of time than you do in the per-planet example, where you might wait until all have fallen, which argues using logistics rather than military stations.

Thus the issue you describe doesn't seem to be so much a question about a boring strategy of doing something that is possible because of per-planet caps and wouldn't be possible with per-galaxy caps, but instead it seems a variant of another strategy, both of which would be considered tedious.

Per-galaxy:
+ Since the turrets for the n planets are not active at the same time, the energy overhead is only for one set of turrets, allowing more energy to be spent elsewhere.
+ Less tedious rebuilding when reconquering the chain as you only put the level 1 defences into place at the chains end.
- Need to prepare a certain metal stockpile before a large reprisal wave is triggered or CPA unleashed. (As you control the first and the second comes with a warning well in advance, this is something that should be trivial for you to do, but it is an added consideration.)
- More tedious micromanagement rebuilding along the wave of advance
- The level 1 turrets are not available to defend at your whipping boys while doing it. If you feel your whipping boys with their level 2+ turrets aren't capable of doing the job, you'll need to operate with a longer chain of planets to win more time and have level 1 turrets at the whipping boys when needed. (This seems a low percentage scenario given we are talking about hundreds of AIP by which time higher level turrets should be unlocked in large amounts.)

Per-planet:
+ You can get all the tedious rebuilding ordered in one go after dealing with each wave.
- This requires a substantially larger constant energy investment, since your actual energy expenditures more closely reflects the defence you actually put up. (EDIT: I should perhaps mention that from a game-balance perspective I actually consider this a positive rather than the negative it is for the player, though of course nobody prevents you from circumventing this issue by only building the turrets as needed and effectively turn the per-planet approach into the per-galaxy approach)
++ Fortunately, you can build the higher level turrets per-planet too, which means you don't need to set up as long a chain, which while it may not save you all that much metal or number of clicks to rebuild, means that the whole state of affairs gets resolved in less time.


This might not have been possible to do effectively do with level 1 turrets with per-galaxy caps before salvage was added to the game, but on the other hand, at that time you might very well have performed a MUCH more powerful distributed chain defence than anything you can do with level 1 turrets by using whatever mark V turret controllers you had unlocked. (And in a game running hundreds of AIP, surely that would be at least a few.)

So either way, I have a really hard time seing this sort of tedious defence as being something introduced by per-planet caps on the unlockable turrets.


EDIT2: I realize I made an implicit assumption above, namely that if you are playing a conquest game with hundreds of AIP, you've unlocked mark III engineers to be able to rapidly build/repair wherever you want on very short notice. It seems a safe assumption to me, but I guess it is possible that other players do things differently.

[Hearteater]

It was not possible before salvage. It might be possible now, especially if you just used Spider Turrets. But Spider Turrets have always been really easy to abuse thanks to their sniper range. Too bad no one suggested shortening their range .

But you do bring up a good point. Speed rebuilding of fixed structures under combat is probably too strong in general. In particular, Force Fields and Turrets. Actually, I kinda feel building in general is too fast. But that might be because I play at a slower pace (15+ hours/game). I don't even unlock Eng IIIs because I don't need the speed, and I unlock IIs only because I run out of Mark I cap.

[TechSY730]

Yea, I know the rule where forcefields, under construction or not, takes 1HP damage if a forcefield that is also covering it takes damage, thus stopping construction and repair assistance rule was put into place to stop such tedious FF micro (and "fix" a survivability "exploit" in the process), but it can be trivially defeated by building forcefield outside of range of other forcefields and moving them into range when complete. In fact, this new micro is even worse, at it requires an additional micro step.

I like the suggestion to give forcefields a "timeout" after moving to give protection. This would severely reduce the effectiveness of this "exploit".
It has already been proven to work well for mobile multi-repairers (forts, mobile space docks, etc.). A timeout of 1 minute may be too long for forcefields though. Maybe 30 seconds?

[Draco18s]

Nope, wouldn't help.  The most efficient way to do the micro you're referring to would not change with your suggestion.

If you instead overlap your forcefields like a venn diagram with the command station in the center (spot 7):


Then you don't have to move any of them to repair any of them.  Or move them small amounts.  As soon as one isn't protecting the command station, it ceases to be covering it, touching the others, and thus repairable.  The repair of a half-dozen engineers is fast enough to bring it back up to 90% before it takes damage (even the 1 point from the forcefield network rule).