AI on AI- earliest elimination record you ever experienced?

[Keler]

Ok, not much to discuss here but I just wanted to share anyway.
I have been running some AI vs AI autoplays using this mod and I saw the shocking early turn 58 elimination.
My question is be it in your test runs or single human plays, what is the earliest elimination have you seen an AI destroying another AI?
I just can't believe but this is the record here.
Here is the World builder file in case I made something wrong.

For those who watch Sulla's AI Survivor games, they can recognize this map probably.

 

[Saxo Grammaticus]

Nothing that low--very impressive! The most fragile AI I remember from last season was Wang Kon in Game Four, who I anecdotally recall folding a few times under T100.

 

[antimony]

Wang Kon is often one of the easiest deity AIs to attack. He builds few troops, and may beeline Code of Laws for his favorite civic, locking himself out of the option of whipping more.

 

[Saxo Grammaticus]

That's a good point--I hadn't thought of that!

 

[ManiaMuse89]

Is that turn 80 kill of Zara by Alex still the quickest in Sulla's AI survivor history?

Tbh, I think you are going to struggle to see much earlier than a turn 58 elimination. I have done a fair few AI autoplays with Sulla's settings (including aggressive AI) and I can't remember many games that have got close to that elimination date.

On standard settings the map is still being settled at that point and won't declare wars until they have finished the expansion phase. So most leaders are only going to declare a very early war if:

- They research bronze working and wheel (or start with wheel) very early on and have copper in their capital or 2nd/3rd city (bonus points for settling on it).
- They go into 'dagger' mode because they sense that their neighbour is very weak relatively in power score.
- Or alternatively, they run out of expansion room themselves very quickly because they are boxed in.
- They have a high probability of plotting a war.

For the super early war to be effective:

- The aggressor needs to declare the war almost immediately after going into 'We have enough on our hands right now' mode to catch the victim off guard.
- The victim needs to have done a poor job of early expansion and teching. Ideally be stuck on 3-4 cities maximum.
- The first city needs to get sniped instantly with minimal losses for the aggressor's stack.
- The aggressor needs to have high attack bravery so they throw their units at the defences regardless of the odds (attacks won't always succeed at lower survival odds but when they do succeed it means that they can steamroll to the next city quickly).
- Ideally all of the victim's cities are on flat ground.
- The victim does not have access to metal or access is cut off very quickly.

I think your best bet would be to have a map full of Shakas. I have seen him declare wars immediately after connecting copper many times and the Impi can be really effective if used correctly. He will also throw the dice at low odds which can occassionaly work out for him.

As for the victim, you need a leader who is:

- High peaceweight
- Has a low build unit probability
- Has a low build wonder probability (so no cultural defence) or tries to build wonders but get beaten to them by other industrious leaders/leaders with access to stone etc.
- Non-creative (so less cultural defence)
- Doesn't have a military tech flavour (so they might avoid archery or bronze working)

Not sure who would be the best 'victim' in that situation. Maybe a leader like Lincoln? He was pretty pathetic when Hannibal rolled over him in season 7 of AI survivor and doesn't have many things going for him to help him defend.

 

[Keler]

Here is all the eliminations I have seen BC date from my excel. Turn 115 is 1 AD. Although it would make sense to look at once I share who starts where in games.
Aparently you are right, here is something very similar in Alt Histories of Sulla.

https://www.sullla.com/Civ4/civ4survivor5-7A.html a turn 60 elimination.

And thanks for mentioning how a protective leader Wang Kon manages to die so fast sometimes. He made to list alright

 

[Eauxps I. Fourgott]

Is that turn 80 kill of Zara by Alex still the quickest in Sulla's AI survivor history?

Yes, it is. Also, to answer Keler's original question, I've never seen anything ahead of that T80 date, which is why I didn't bother responding before

I wonder if we'll ever see the Raging Barbs score a kill on a wildcard map? (I've seen rare cases of non-Raging barb kills on Noble maps, but I don't know if it's feasible on Deity.) If that were to happen it would represent our best shot of beating Keler's record.

 

[Thrasybulos]

In my Season 4, Playoffs-2 replays, there was a Gandhi T72 elimination. I don't think I've seen anything faster.

But try beating a T495 elimination, on Deity !

 

[Eauxps I. Fourgott]

In my Season 4, Playoffs-2 replays, there was a Gandhi T72 elimination. I don't think I've seen anything faster.

But try beating a T495 elimination, on Deity !

That's right, you said you saw a Time victory at one point? How in the world did that happen? I'd thought it was impossible for this setup.

 

[Thrasybulos]

Elimination of the tech leaders by backward civs, followed by constant and inconclusive warfare among the survivors.

I've only seen two time victories so far (the other instance was an AH of the season 5 opening round game which saw Bismarck illustrate his "special"), but do happen, ever so rarely.

 

[ManiaMuse89]

Elimination of the tech leaders by backward civs, followed by constant and inconclusive warfare among the survivors.

I've only seen two time victories so far (the other instance was an AH of the season 5 opening round game which saw Bismarck illustrate his "special"), but do happen, ever so rarely.

Season 4 was rocky climate and season 5 was cold climate so maybe that had in impact?

Latest finish dates that I have seen on temperate maps have been low turn 400s. Usually messy games with relatively equal leaders towards the end, slow tech rates and usually some dumb shenanigans. Think I had one where a launched spaceship got sniped during a late game war and then another leader went on to win a very lategame cultural victory.

 

[Thrasybulos]

Season 4 was rocky climate and season 5 was cold climate so maybe that had in impact?

Good idea, but I don't think so : as you can see, finish dates were standard fare on that map, the longest game (apart from that time victory) being a T379 spaceship win.
It was just a freak happenstance.

 

[Thrasybulos]

Not record-breaking, but just had this when doing AH for S5-G4:

T83 is still pretty damn early.
(Hannibal was apparently in a rush: he went on to win a T254 Domination after that).

 

[Plains-Cow]

The Zara-Alex rivalry in AI Survivor is my favorite rivalry.

 

[Thrasybulos]

Just equalled that date, with JC murdering Viccy on T58:

Spoiler :

 

[Keler]

With raging barbarians on and archery tech given, I have seen two times barbarians destroying an AI starting on S2 Wildcard map Hannibal spot.

 

[Thrasybulos]

Yes, I've had that happen a coupla times too when running AH, but I wasn't recording kill dates. Plus, barbs don't really count.
Those were annoying though, because I needed to update my control formulas when that happened.