Special Forces Insanity

[Mad Rubicant]

My immediate first action after the hack is to destroy the Mutherhung Command Station. I need knowledge.

13:04: The plan is to clear out the north branch. The reality is, I'm due for a wave any minute now.

13:08: I was right! 11000 AI 1 ships, 80 starships, 9500 AI 2 ships, 69 starships I'm just going to unlock Fort IIs...

13:10: Wave 1 hits

13:11: Wave 2 hits, somewhat underwhelmingly, since the first wave is past half dead. It was bombers too, so it could have done quite a bit of damage.

13:15: Waves repelled. Next up, northern branch.

13:19: Yezing goes down.

13:21: Parfallus, the planet behind it, falls.

13:32: Barkyaqmoat falls.

13:34: Kuticmel goes down, bringing my AIP up to 727, and into Tech Level 3.

13:37: Dire guardian spawning in 5 minutes, and 8500 K.  I buy Reprocessor III/IVs.

13:42: Dire Needler Guardian. Meh, not too scary. I unlock Fort IIIs and Metal Harvester IIIs. Now nothing can touch me!

13:49: Celeasi goes down. Then I kill the Dire Guardian Lair. My AIP is now a Boeing plane, at 747.

13:52: Haboatnei goes down.

13:56: 5 EMP Guardians detonate on Mutherhung. Ohhhh dear. Then the waves declare: 20,000 AI 2 ships, 87 starships. 23,000 AI 1 ships, 80 starships. 43000 ships headed straight for my homeworld. The only way this could be worse if there was a CPA right about now. I really really home that the new fortresses can hold this off. I think I'll also buy HBC IIIs.

13:59: 12 EMPS! 3 MINUTES OF PARALYSIS! AAAAAH

14:07: Somehow, the wave is repelled.

14:08: The Ion IV finishes. All that's left is the Ion V.
I think this is a good spot to take a break. Man, these waves are getting crazy. I may have to try and superterminal down to the floor.

[Mad Rubicant]

Well, I dun goofed and lost about an hour of play. To make a long story short, I tanked a Champion Exo, a 20k ship CPA, a 20k ship wave, and a 30k ship wave, all within the span of 20 minutes. I had to use a few warheads, too. I'm probably going to lose this game soon. My homeworld is half-trashed

14:47: The rebuilding begins.

14:52: The gravity and mine networks are back up. Now it's time to blow up another planet in the northern branch with my champion. I need the Logistics command stations

14:53: Inimucha falls.

15:02: Tchmurderpi goes down. Now there's only one planet left in the northern branch, and it's a rude gesture that got a CSG-D.

15:07: Only triangle ships and starships remain to be built. I also unlock Logisitics III stations. I hope that it'll keep me alive for the next wave.

15:10: I destroy Waybul. Now all the waves will be coming from one wormhole. I'm down to starships for rebuilding.

15:13: Waves declare. A total of 55k ships headed for me this time.

15:21: The carriers are up to my command station, and there's still 20k ships left on the planet. I pop a lightning warhead, to little effect. Next is an armored warhead, but it only takes out 2k. I think this might be it.


15:23: The AI has won. The Home Command Station on Murdoch falls, with 10k ships left on the planet. AIP was at 875.


Well, that was fun. I had two of the branches under my control by the end. If I'd had attritioners or a botnet (or both) on my home planet, I'm sure I could have won. As for the salvage nerf, while I'm glad it's not a fire hose, I'm sad that I'll never be able to make an Ion V. Maybe in a FS game.
On the AI side, I barely saw the Special Forces, since both AI homeworlds were in the same branch. The H/Ks weren't very threatening either. I'm glad the AI will dip into the SF for CPAs, at least. There were at least 50k+ running around by the end. But they were also waaay late to the party whenever I went after a high priority planet. Maybe they should patrol between planets marked for protection, so that they get there in time?
Lightning Torpedo Frigates are just amazing on defense. Also, are they using the same kind of attack as before the warhead change, where they hit everything in range for damage? Or do they hit for a maximum of 1500 total?

[keith.lamothe]

Congratulations on seeing that all the way through, it was impressive to get that far with the semi-recent changes to make reinforcements send their "excess" to waves (that's why it was getting so absrud, mainly: past 200 AIP it just does reinforcements like it was 200 and sends the rest to waves).  So it may be that one really does need to get the memo nowadays, or die horribly.

AI War hopes you have enjoyed this annihilation experience and that it may also serve your future being-annihilated needs.  Share and enjoy!

But they were also waaay late to the party whenever I went after a high priority planet. Maybe they should patrol between planets marked for protection, so that they get there in time?

Hmm, possibly, though that may lead to more situations where it's pathing through player planets, etc.  But having it shift the "rally point" more often wouldn't be a bad thing to consider when I get around to moving its logic to the AI thread (where it can use more CPU time).

Lightning Torpedo Frigates are just amazing on defense. Also, are they using the same kind of attack as before the warhead change, where they hit everything in range for damage? Or do they hit for a maximum of 1500 total?

I don't think LTF torpedoes were ever "strikes literally everything in range", but have a target cap like most other stuff.  But maybe I remember wrong.  Anyway, they were not affected by the warhead changes.

[Mad Rubicant]

Congratulations on seeing that all the way through, it was impressive to get that far with the semi-recent changes to make reinforcements send their "excess" to waves (that's why it was getting so absrud, mainly: past 200 AIP it just does reinforcements like it was 200 and sends the rest to waves).  So it may be that one really does need to get the memo nowadays, or die horribly.

AI War hopes you have enjoyed this annihilation experience and that it may also serve your future being-annihilated needs.  Share and enjoy!

Huh. I probably could have pulled this off before that change, then. Oh well. The AI was also waiting almost an hour between waves, so that was big multiplier, too.

Actually, if wave sizes are getting this much bigger at high AIP, what does it mean for Fallen Spire?

[keith.lamothe]

Huh. I probably could have pulled this off before that change, then. Oh well.

That particular change was primarily motivated by the fact that people were in fact winning this challenge, which should generally not be winnable

The AI was also waiting almost an hour between waves, so that was big multiplier, too.

It will wait longer if you've only got one planet it can send waves to; was that the case?

Actually, if wave sizes are getting this much bigger at high AIP, what does it mean for Fallen Spire?

It may prove problematic, I don't know.  In general the idea is that the FS stuff should be able to vaporize waves (but be more challenged by the exos) but that may not be the case.  Buffs for the spire city stuff are quite likely, and buffs for the capital ships are also possible.

[Peter Ebbesen]

Regarding what the redirection of reinforcements to waves does to Fallen Spire, in my limited experience so far the result is that the regular waves are considerably more dangerous than the exo-waves.

This might be different if I were playing with the Zenith trader or champion goodies and tanking normal waves on my homeworld with full Mil3 and Log3 station support rather than tanking them on another planet, as those extra goodies are made for killing humongous numbers of weaker units, but using just the normal stuff and Fallen Spire, the exowaves don't come anywhere close to being as dangerous as the regular waves at the two spire city level and 450 AIP I've reached so far.

Here's an example from when I tried logging wave strength last night to find out just why the Mad Bomber was throwing more than a thousand Electric Bombers at me as part of a 6k wave at rather lower AIP than Mad Rubricant ended up with. Around 400 or so IIRC.

normalWaveStrength:22852.76
AccumulatedStrengthForNextWave:22863.54
waveStrength += this.SurplusWaveStrengthDonatedFromReinforcements = 52531.76

In other words, the bonus added from reinforcements is considerably greater than the base amount granted by AIP, to the degree that I seriously considered halting further conquest until I'd polished off every single guardpost in the galaxy outside the core worlds, working on the assumption that this would lower the number of reinforcements directed to waves. Which would admittedly be a rather silly approach to war, but seemed to follow the logic of the game mechanics as I perceived them. "We don't dare conquer anything until we've destroyed everything".

[Mad Rubicant]

The AI was also waiting almost an hour between waves, so that was big multiplier, too.

It will wait longer if you've only got one planet it can send waves to; was that the case?

Yeah, my homeworld was the only planet with adjacent warp gates. Gotta get that 50% salvage, after all.

[keith.lamothe]

Regarding what the redirection of reinforcements to waves does to Fallen Spire, in my limited experience so far the result is that the regular waves are considerably more dangerous than the exo-waves.

Yea, I'm thinking maybe the contribution from "extra AIP" (> 200) to waves should be halved, or some such.  So it still gets to spend the strength, but less efficiently.  Maybe even 1/4, dunno.

normalWaveStrength:22852.76
AccumulatedStrengthForNextWave:22863.54
waveStrength += this.SurplusWaveStrengthDonatedFromReinforcements = 52531.76

In other words, the bonus added from reinforcements is considerably greater than the base amount granted by AIP

Haha, yea, that's a bit much.  Though eventually it's to be expected, as before the AI's reinforcements were capped to not benefit from AIP over a certain level it would spawn TONS of reinforcements.  Which didn't tend to cause a loss, but did tend to cause stalemates.  So the idea is for them to now try to cause a loss.  But on FS the increased response should be something that can be dealt with.

to the degree that I seriously considered halting further conquest until I'd polished off every single guardpost in the galaxy outside the core worlds, working on the assumption that this would lower the number of reinforcements directed to waves.

Actually that's one of those cases where you'd be better off not knowing outside-the-game info (like the redirection of reinforcements being why the waves go up so much) or else knowing more completely what's going on, but being in the middle leads to conclusions like that

When the AI gets bonus wave strength from > 200 AIP not being "used" on reinforcements, it does so at the "base rate" of what that extra AIP would have granted to reinforcements.  Specifically, that's before any bonuses due to guard post count or troop accelerators or whatever.  So destroying guard posts will not substantially help there.  Nor do I want it to, to avoid motivating precisely such behaviors.

[Peter Ebbesen]

to the degree that I seriously considered halting further conquest until I'd polished off every single guardpost in the galaxy outside the core worlds, working on the assumption that this would lower the number of reinforcements directed to waves.

Actually that's one of those cases where you'd be better off not knowing outside-the-game info (like the redirection of reinforcements being why the waves go up so much) or else knowing more completely what's going on, but being in the middle leads to conclusions like that

Hah, good one!

When the AI gets bonus wave strength from > 200 AIP not being "used" on reinforcements, it does so at the "base rate" of what that extra AIP would have granted to reinforcements.  Specifically, that's before any bonuses due to guard post count or troop accelerators or whatever.  So destroying guard posts will not substantially help there.  Nor do I want it to, to avoid motivating precisely such behaviors.

That's where the benefits of being both a mathematician and a games developer for most of my professional career come in. I'm way too logical to actually do all that much work without structured testing. If I'd decided to go that way based on the data, I'd have systematically polished off a handful of easily accessible systems and then set down to parse the data from the log to see whether it worked the way I assumed rather than do the honours on all the remaining systems first. In the event I decided not to do it as it wasn't in the spirit of your original challenge, but I iwas sorely tempted.