To me at least, the AI retaliation feels more punishing than it should be. Taking a planet with what seemed to me to be relatively light losses invariably provoked retaliation strikes, and for a stretch it seemed that every wave that came at me was a retaliation wave. (Again, while taking pretty much just relatively minor losses). (This was in a 4 HW game with ultra-low caps.)
I think the key thing there is "4 HW game" as that makes for "normal" waves that are quite substantial.
Salvage-reprisal waves are just normal-sized waves with the salvage tacked on as extra strength.
I'd wanted to have the salvage-reprisal waves be
just the salvaged stuff but that would be very underwhelming in one of two ways:
1) Either the timing is kept the same (launches 3 minutes after you start taking losses, as long as there's at least 100 strength worth of salvage to send), and a lot of those reprisal waves are going to be really, really, really stinking weak
Like "why did it even bother?" weak.
2) Or it saves up salvage until it has at least as much strength as the last wave. But that could take several player attacks before enough is built up, and in order to help with the refleet syndrome this response has to happen pretty fast (otherwise you're waiting a long time for the thing that's supposed to help with waiting)
...
On the other hand, you only really get refleet syndrome if you just had a major wipe. And if your major wipes don't amount to a single AI wave that might say something right there.
So perhaps I could change the salvage-reprisal thing to only use the salvage, and just save it up until it's at least as big as the last wave.
It might need to be tweaked to get a higher percent of salvage, etc, but it'd be worth trying.
- I had a ton of enclave starships spewing out streams of drones. Lots and lots of them - hundreds at a time in a big battle. And their metal cost isn't that low. Are all those drones giving salvage?
They count, but they should be really cheap. But their cost wasn't really relevant before and while I did do a sweep through all the object types in the game to catch the egregiously-off metal costs (H/Ks were rather "cheap" before, etc) for units that might conceivably be salvaged, I may have simply missed these.
But just checking in-game, the drones had a substantially lower metal cost than normal younglings (40 vs 200, in that particular save's cap setting). That doesn't sound too high.
- Do "disposable" ships give salvage? Stuff like:
* Martyrs
* Shield Bearers
* Rams
* Mini-Rams
* Autobombs
* Nanoswarms
* Railpods
* Scapegoats
Absolutely
For the "dock cannon" ones this is actually something of a counter-balance, which I've felt for a while they've needed. But I still use dock-cannons shamelessly and while I do notice the reprisals they aren't anything like a true deterrent. On 7.6, anyhow.
If so, that makes me not want to use any of them.
I submit that this is not a rational nor a wise tactical decision
I can still clean mkIV planets with nothing but docks cranking tigers on the other side of the wormhole, and the reprisals get interesting but pose no threat to even modest static defenses.
On the spirecraft, we'll see if their costs cause problems, but I'm guessing not.
- Do things like Lightning Torpedos, Spire Blades, Hive Wasps, etc give salvage? Their price shows as relatively cheap, but we're talking hundreds of them per battle.
Yep, they give salvage. Those related ship types are very, very useful. The fact that the enemy (if fought in its territory) gets some kind of benefit from you throwing chunks of metal at it... quite intentional
But they're not going to amount to much compared to any significant normal-ship casualties.
- Do scouts provide salvage?
Nope, specifically excluded to avoid scouting causing reprisals.
- Does self-attrition death provide salvage?
Yep.
- small and cheap, swarmer-type ships suddenly seem even less useful This is my biggest concern. Tougher, less-likely-to-die ships won't hand nearly so much salvage to the AI to use against you.
As I've said before, even throwing swarmers into the grinder
as hard as I could (mkI, II, and III tigers vs a mkIV world with a Plasma Eye that was killing
dozens of them per second on low caps) didn't cause a problem modest amounts of turrets couldn't solve. Granted, that was below diff 8, but a serious defense could have handled far more. And that was actually a 2-human-homeworld game (multiplayer), so we were getting the double waves and such.
And more importantly, those swarmer attacks were still incredibly useful. I would very probably have taken far worse (and far more costly) casualties taking that planet via traditional means.
- Late game, low mark units become less useful. Late game, mark I and II units that were useful cannon fodder and extra firepower are now converted by the AI into its own mark 3-4-5 ships. Is it worth even building them now?
If you lose a mkI unit on an AI planet, that doesn't give them the same amount of salvage as a mkIII or whatever. It's all done by metal cost.
If the AI's tech level is up to three, sure, the reprisal waves will be mkIII. But they're still getting the same amount of reprisal
strength as they would have at mkI or mkII. So if it's buying bombers for the wave it's getting way fewer from salvage at mkIII than if it was still at mkI.
And normally, I might send a bunch of "chaff" ships along with them to increase the raid starship's survivability, but that's not desirable now.
I disagree, for two main reasons:
1) The reprisal salvage from just a few starships is not going to wreck you or even challenge your turrets. In fact you'd probably get more metal out of it than you put in (due to it being base waves + salvage).
2) Even if it did, it's very likely that judicious use of chaff ships could significantly reduce the overall amount of metal you hand the AI in the exchange.
- Smaller concern: Late game, I sometimes need to send a fleet to "prune back" the threat that builds up on neighboring AI planets. While this helps manage threat, it also majorly helps game performance once the ship counts start getting really high. Now it feels like this is handing the AI bonus resources for doing something that is more a chore than a benefit.
I doubt that the actual amount the AI would gain in that would be a problem to you. The trade would still be very much in your favor come next major-attack (or whenever the scales would tip toward that threat actually trying to attack you).
Anyway, thank you for the feedback on the new mechanic. Iteration is always necessary with such a thing. But I do think much of your concern is founded on theoretical extrapolation combined with a somewhat skewed experiment (4HWs having much larger waves).
That said, it does seem that not having every salvage-reprisal wave be "a full set of normal waves, PLUS salvage" is probably best. If that makes it too weak then that can be adjusted in another way.
