Join the Stellaris 3.3 Grand AI Tournament now! :)

[Dubanian]

It's good that my Kodra federation starts close to Cometborn. 3 vs 1 during early game is an advantage. If lucky, Kodra can gain some systems from Cometborn, and able to expand. They will be likely blocked by their buddies at the beginning.

 

[Dragatus]

Zark Swarm seems to have the strongest fleets even when it doesn't win, so it's a powerful contender for eventual domination. Cometborn Consumers are also one to watch, they're just excellently designed for the AI. But ultimately having the best tradition progress might be the deciding factor.

Observation: Kodranites got boxed in again.

 

[Thoughtful Gaming LP]

Zark Swarm seems to have the strongest fleets even when it doesn't win, so it's a powerful contender for eventual domination. Cometborn Consumers are also one to watch, they're just excellently designed for the AI. But ultimately having the best tradition progress might be the deciding factor.

Observation: Kodranites got boxed in again.

That is the weakness of the Common Ground / Hegemonial starts. Almost always, at least one of the empires is quickly boxed in and is severely hampered. We'll see.. it's an interesting setup with the Ravenous Hives close to action from the start. Something they should profit from.

 

[Dubanian]

Set hyperlane density to max will fix this weakness. We can always tweak Game Settings to bring balance to the Galaxy.

 

[Thoughtful Gaming LP]

Set hyperlane density to max will fix this weakness. We can always tweak Game Settings to bring balance to the Galaxy.

It's a good idea to fix the problem with the three empires blocking each other.

Maybe next tournament. Hyperlane blockage is what keeps Common Ground / Hegemonial "only" resilient. With a lot of hyperlanes (how many should we go for is one question), it is an advantage to the 3 empire block, which could make them stronger than intended.

I've mostly chosen default settings so we have a common ground to start from and so I don't have to establish a whole ruleset that is balanced, because I don't think I know what balanced is in the game. So I want to avoid this for the tournament. I have used hyperlane density when I had played a Common Ground game myself because it was just about having fun single player.

As people have rightly told me, I also see more and more that Common Ground and Hegemonial is a bit problematic as there's suddenly other empires that noone has designed, just like in Lost Colony or Scion.

In a partly competitive environment I want to take myself back as much as I can so I can point to a ruleset not designed by me that might favor some empires and others not so much. If I change it, there is probably a discussion, and rightly so. I'm just a little scared of and by balance discussions, that's all

Also: Final galaxy is runnnnnnnninnnnnnngggg still.

 

[Dubanian]

It's your tournament. You're doing a good job to keep us entertained.

 

[Thoughtful Gaming LP]

It's your tournament. You're doing a good job to keep us entertained.

Thank You! And here come the Grand Finals!
I compressed them as hard as I could! But it ran for a lot of hours

Finalists! Good luck everyone and have fun watching

1. Citizen Compact of Marass - „Friendly Bird Warrior Oligarchs“ Avian Xenophile Fan. Militarist Life-Seeded Citizen Stratocracy G4 6287 (10 WP), G4L 41523 (20 WP)
2. Grimer Conglomerate - „Friendly Anglerfish Traders“ Aquatic Xenophile Fan. Materialist Corporate Trade League G5 SUBLIMED (4 WP), G3L 40543 (20 WP)
3. Cornarian Interplanetary Combine - Friendly Fox Warriors (Military Dicatorship, Xenophile Militarist Materialist) G2 906 (8 WP), G2L 35321 (20 WP)
4. Agarthan Dominion (Clone Army Fanatic Materialist Oligarchic Technocracy) 5443 (10 WP), G1L 61892 (20 WP)
5. Cometborn Consumers - „Meteorite Menace“ Lithoid Calamitous Birth Ravenous Hive G8 18124 (20 WP), G4W 594623 (20 WP) Shattered the Galaxy
6. Kodranite Corporation (Common Ground Lithoid Fan Spiritualist Death Cult Corporation) G1 10170 (14 WP), G4W SUBLIMED „Last Survivor“ (18 WP)
7. Lernean Swarm - „Molluscoid Abominations“ Molluscoid Prosperous Unification Hive Mind Ravenous Hive G7 18986 (20 WP), G3W 566992 (20 WP) Shattered the Galaxy
8. The Legion - „Rockbreaking Machine Intelligence“ Machine Resource Consolidation Gestalt Machine Intelligence Industrial Production Core G6 11706 (18 WP), G3W 13387 (18 WP)
9. Zark Swarm - War-Geared Alien Menace (Very Strong Ravenous Hive Mind) G2 17433 (20 WP), G2W 598318 (20 WP) Shattered the Galaxy
10. The Galactic Stock Market of Worlds - „Corporate Hedonist Ants“ Arthropoid Xenophile Militarist Spiritualist Dragon Origin Megacorporation G4 8308 (12 WP), G2W 13092 (18 WP)
11. Mim Super Special Empire Directorate (Machine Intelligence with pumped up Pop Production and Maintenance / Unity) G1 18040 (20 WP), G1W 56711 (20 WP)
12. Sathweer Sacred Theocracy - „Slimy Water Dragon Worshippers“ Aquatic Egalitarian Xenophobe Spiritualist Here Be Dragons Oligarchic Theocratic Oligarchy G7 8136 (18 WP), G1W 28823 (18 WP)

Reference Video to look empire details up

Tournament Playlist
https://www.youtube.com/playlist?list=PLwfViAw_vAvJlACDzciBzytTWf3o86HDJ

 

[Koopatin]

Spoiler: it happened--- really.. do not read before you watch

the last galaxy was the galaxy it finally happened. a federation formed early enough to defend somewhat against the zark, the northern lernean swarm was ganged on by two neighbours and the Zark (again) and the cometborn did weaken each other enough for the other empires to jump on them. in the early game it looked like a devouring dominated game again. the Zark did declare on the cometborn and the federation used that time to counter attack. weird, in other games AI with better numbers did not use similar opportunities. did you increase aggressiveness? ^^

in any case, gz to the mim ^^ can't wait for AI tournament 2023, the Zark will be part of it again

 

[Cry_Havok]

Huh. Well, gestalts are OP at ensign, it seems like for all the attention that the swarms got, a normal machine empire is just as good, while not having total war. I wonder why Mim just decided to stop short of Galactic Conquest, it certainly could have

One thing I have noticed is that I do not see such when I play at scaling admiral. I think difficulty buffs help normal empires more, since it takes away some of the stress of the more complicated economy.

 

[Dubanian]

My Kodra federation fought poorly against Cometborn and was wiped out too soon.
I suggest to shorten the duration for 2nd round and finale from 500 and 1000 years down to 400 and 600 years.
Congrats to Mim.

 

[Dragatus]

I think the sheer number of genocidal actually helped to keep them down. Instead of having one swarm repeatedly take bites out of normal empires until it was too strong to stop, we saw two swarms at each other's throat. And the thing about genocidals and their total war CB is that it works the other way too. Others can declare total war on them too, which Mim took advantage off.

Maintenance Protocols appears to be a top tier civic for machines because it really helps a lot with their unity and allows them to get tradition at a respectable rate.

 

[Cry_Havok]

A potential idea for future tournaments is to run each galaxy at ensign and grand admiral, at least at first round, and if the top empires are basically the same you could stop, but I really think that regular empires benefit a lot from difficulty bonuses, more than gestalts

 

[Foxosaur]

Well I guess that is the end of this but well done Thoughtful, it was a really interesting thing to get involved in and thank you for organising it all. There was a lot of fierce competition and it was awesome seeing so many varied submissions.

Congrats to the machine intelligence ^^

 

[Oculument]

Swarms not so OP after all. One swarm per galaxy is optimal for them, multiple swarms create competition so are not good. Final was only run once, so who knows what the position and terrain randomness contributed.

 

[Cry_Havok]

Swarms not so OP after all. One swarm per galaxy is optimal for them, multiple swarms create competition so are not good. Final was only run once, so who knows what the position and terrain randomness contributed.

The Semis also had multiple swarms, around 50% seems to be the critical mass when swarms start to have problems

 

[Dragatus]

Swarms start to have problems when they start close enough to start fighting before they've grown too strong for the normies to handle them. Because they spend their strength fighting an equal they stop growing as much and the normies can team up and actually fight them.

Lernean Swarm also did poorly for some reason and had a smaller fleet than either neighbor alone, which lead to it getting jumped. But I have no idea why that could be, there are no obvious problems in the empire's design. My best guess is they were unlucky in some way. Like maybe they didn't get any weapon upgrades for some time.

 

[Thoughtful Gaming LP]

Well I guess that is the end of this but well done Thoughtful, it was a really interesting thing to get involved in and thank you for organising it all. There was a lot of fierce competition and it was awesome seeing so many varied submissions.

Congrats to the machine intelligence ^^

Well said, thank you a lot It was great to have everyone on board, we had a fantastic event and in the end even a surprise winner! Couldn't have worked better!

Thank you everyone for participating. We will come back to this somewhen in the Overlord patch when the first small patches are done.

Now I'm going to take a little breather, read the comments again, reflect on all the new ideas and how it all went, and take in all that we've had together.

All the best, until next time and Happy Gaming, Your
Immanuel Can

 

[Koopatin]

just wanted to let you know, (if you ever lost sight of the event.) it is done, and we have some interesting results. how does it compare to your mentioned perspectives?

 

[Foxosaur]

I’ll throw out the little I observed: -

• There are not enough early alliances, even in mid-game. They come far too late, especially if there is a genocidal threat.
• Empires should see opportunity in war when the enemy is already distracted (as in, they should declare a war, if the opportunity is there).
• The daily hull / armor regen for genocidal civics is very over powered. Maybe halving it or similar is more realistic. And according to @Dragatus, they do not factor into fleet power calculations so the AI does not take into account their ‘extra’ mobility in the game.
• Empires that get star base influence reductions don’t seem to take advantage of being able to grab them at nearly half the price ( Fanatic Xenophobe ) and are never any larger in number of systems than other empires.

 

[Thoughtful Gaming LP]

I’ll throw out the little I observed: -

• There are not enough early alliances, even in mid-game. They come far too late, especially if there is a genocidal threat.
• Empires should see opportunity in war when the enemy is already distracted.
• The daily hull / armor regen for genocidal civics is very over powered. Maybe halving it or similar is more realistic. And according to @Dragatus, they do not factor into fleet power calculations so the AI does not take into account their ‘extra’ mobility in the game.
• Empires that get star base influence reductions don’t seem to take advantage of being able to grab them at nearly half the price ( Fanatic Xenophobe ) and are never any larger in number of systems than other empires.

Very good! I'll add that just in general, the AI plays some Origins really badly. This has probably not been noticed but could be easily corrected, just three examples that I've seen happen:

1. Clone Army AI often fails to build an Ancient Clone Vat on their first and / or second colonies, leading to population die-out and the loss of the early colony. Only Clone Army AI with huge mineral production and the -15% building cost civic usually manages to do it in time regularily.
2. Whenever the AI finds a Digsite, it prioritizes this digsite over all other exploration with science ships, meaning if it starts with "On the Shoulders of Giants" Origin, the AI will not survey systems and so not be able to expand until it has a second science ship that will then start to explore and survey systems. The first science ship will be glued to the home system digsite. So the AI will only expand realllllyyy late. In general, finding early digsites in their systems screws the AI up.
3. Whenever it expands, AI will disallow early colonist jobs on new colonies, leading to unemployment, negative amenities and very low stability (~15%) there and often leading to devastating early crime / ethics / rebellion events for empires that do not have high bonus stability, happiness, or a way to offset it via species rights.

Because all of that happens so early, it has a big effect on AI performance with such bad starts.

Hoping that is interesting to @Offe as well

All the best, until soon, Your
Immanuel Can