It's still called AI War: Fleet Command.
I'm running a Let's Play. This is the epilogue.
Right, so let’s reflect on this one.
What went well?
‘Cheap and disposable’ worked really nicely for my early game, and I was lucky to get Neinzul Railpods to play straight into that. Railpods kind of fell off near the end there, but, well, they’re Mk1 or 2 vs. Mk5. I’ve never used Mobile Space Docks to actually build ships in the field, so having the chance to churn out crud on the fly turned out to be really fun. I keep them around to repair my fleet mid-combat, since they’ll do repairs after 5s without taking damage and there’s plenty of opponent ships or structures that deal damage every 6s or longer.
I think I did a pretty good job of scouting. There were a lot of Data Centers to bust, and having forward knowledge about the Raid Array kept me safe.
I could’ve taken Mk3 Railpods instead of Mk3 Cutlasses. Additionally, if I wanted MORE chaff, I could’ve used a hacker to change the advanced research result from Tachyon Microfiters to Laser Gatlings, which are also a) small b) numerous c) cheap. I think the Cutlasses did okay, though.
Securing the Mk5 Teleport Raiders and the Mk5 Needler/Missile Turrets proved ABSOLUTELY CRUCIAL. Being able to allocate magnitudes more defense for the same cost as the Mk1 versions allowed me to effectively repel most forces thrown at me. This freed up science points that would’ve gone to nicer turrets for nicer ships instead, and it paid off wonderfully.
If I was itching for science, there were two planets up near Perfidy with, uh, Spire libraries? Something like that? The structures looked like smooch marks, there was one in The Coming Calamity and another next to that. Holding them long enough would pay out 8,000 (!!!) science each, but give the AI 80 (!!!) progress points if they blew up before reaching that number. That’s a gamble I’m more likely to make in future games.
Broadly speaking, I think my choice of planets went exceptionally well. I took and held crucial chokepoints, which allowed me to funnel attackers in where I wanted to receive them. The Omeyocan coalition was, admittedly, a pretty weak conglomeration, but it did let me keep a conglomeration and crossed off all my core shield checkmarks. There were better A and D core shield options, but none that made sense for the B core shield.
The pattern of pausing the game to write about the game while thinking about how the game was doing gave me PLENTY of time to strategize. I mean, you get to pause freely and strategize anyways, but I did not end up with a lot of wasted time. This campaign ran a little over five hours on the clock, but I probably spent four times that with the game in mind. With AI Progress ticking up ever so often (or not, you get to configure it!), time saved made me stronger.
So what went poorly?
I invested into Mk2 missile ships and promptly let them all get eaten on the way to important planets during my homeworld assaults. It’s not a big science cost, but it’s not a good play either. Since there’s 96 in a tier, it would’ve been prudent to build one singular Transport and crammed them all in there so that I could continue to use them without a) dragging the rest of my fleet down to their speed and subsequently b) losing my youngling vultures to attrition.
Speaking of,
I didn’t use my Youngling Vultures well. They’re exceptionally cheap and disposable, but attrition ate them alive before I could get the rest of my fleet into position. They’re surprisingly punchy, except when they’re absent in a deepstrike firefight. Transports would help here again. Or not tying part of my fleet to missile ships. I largely do not like missile ships because they’re so damned slow!
It would’ve been advantageous to hack the schematic for Mk5 Teleport Leech ships. The Raiders lacked a bit of oomph where it mattered and having a second bunch of Mk5s to accompany them would’ve made a lot of a difference.
It would also have been advantageous to hack the schematic for Mk5 Sniper turrets or Spider turrets. I’d take one of these over the needler turrets any day. They were out there in the galaxy, and I ignored them.
I should’ve taken the AI Progress penalty and mended the Armored Golem. I’m not sure it would’ve saved Omeyocan against the one Raid Starship the AI threw at it, but it would’ve kept the lower coalition safe while I did more important work elsewhere.
If not the armored golem, I should’ve taken the energy cost and reinforced Omeyocan and its neighbor better. A little extra science towards forcefields Mk2 might’ve helped too (as would position turrets closer to the command center on Omeyocan, damn that raid starship).
AI Difficulty 7 is perhaps a little easy for me by this point. 7 is the point where the AI ‘comes online’ and your actions start having consequences, so it’s the right difficulty selection for when you’re ready for the ‘real thing’ as it were. Since I’ve done a couple laps around AI War, it feels pretty safe to say I need to bump this up to 7.6 or 8. Yes, difficulty does begin to get granular past 7. 9 is regarded as the threshold of what’s beatable and what’s not. You’re not supposed to be able to win against AI difficulty 10.
Jumping into Perfidy from Yriel was ultimately foolish- it was sandwiched on either side by core guard posts with heavy shields, and I didn’t manage to make sensible damage here besides destroying the mass driver. That could’ve been accomplished with a half-fleet of bombers or something and saved me a bunch of ship cost/reprisal rigamarole.
As far as this LP as a whole goes- yes, the entire thing happened outside of audience participation. This game is done by the time you’re reading post one. Audience participation would be swell, but having a completed playthrough is better. I wrote this because I love AI War and I wanted to tell its story. It’s my hope you’re reading this sleepy-eyed from bed while on your tablet, enjoying that someone else has made this wonderfully dense game more digestible for you, and possibly inspiring a few adventurous thoughts about giving it a spin yourself before you zonk out.
I got to write and show off my favorite toy. I’m so happy about this. My thanks go out to Arcen Games for embodying the indie gaming spirit, for trying at SO many genres, and for always, always having a Linux port of their games. For not shackling their game in stodgy DRM, for allowing me to inflict it in earnesty, in passion, on my poor victims-err, friends. For A Valley Without Wind, and Starward Rogue, for always being brave enough to do something *different*. At the time of writing, Heart of the Machine is recently released and it sounds like it’s a smash hit. I’m glad they’ve survived.
What’s next?
Like I’m going to stop playing AI War. I’m hooked.
< Prev | Next >






