Some research into Civilization splitting

[peterval]

A recent game at Chieftain level provided an opportunity to research the requirements for splitting civs. Just in case this is new for some players perhaps I could summarise the results.

The year was circa 1760 AD, there were 6 civs including me. Germany had 5 cities and was backward, Rome had 8 cities, population 47, intermediate science. I was America, 5 cities, population 46, all the sciences.

There were some German units lurking around Rome so I saved, despatched a carrier full of bombers and destroyed the Roman defences. Next move the Germans marched into Rome. No split. I returned to the save and repeated the destruction and in the same move became the German civ by saving again and editing the file. This time as soon as I, the Germans, entered Rome the Roman civilization split into the English and Romans. So far everything as expected.

Now I returned to the first save, repeated the destruction, and took Rome for America. Although I was 5 cities and the Romans 8, no split. Slightly unexpected.

This time I moved back to 1750 AD because I knew the Babylonians would invade me. Sure enough three Babylonian divisions landed outside Atlanta, but while previously I had just swatted them away, this time I annoyed them and allowed them to take Atlanta. Now I was down to half the size of the Romans but when I took Rome again, still no split. I tried again and this time destroyed all my army, navy and airforce except one Armour, allowed one move to let the program register the new situation and took Rome again. No split.

However I had retained 39 caravans ready for a quick rebuild and I wondered whether the program had taken their defensive ability into account. So this time I destroyed everything except one Armour outside Rome, and with one band of settlers blocking the railway between Atlanta and Washington I waited one move an took Rome again. Hurrah. The Roman civ split 4 to 3 and the English were back on the scene. I had a sprained wrist but it was worth it.

The conclusions I draw are these: 1. In Civ1 AI civs do not split each other (although I have engineered such a split in Civ2). 2. Splits can be produced late in the game even when the human civ has a substantial scientific lead. 3. It is not enough to have fewer cities and a smaller population than the target civ; it appears that the sum of the human attack and defence value must be substantially weaker also. 4. If you reduce the size of your army it seems advisable to wait one move for the program to register it.

I would be glad to hear whether anyone has any different experiences to relate in connection with this fascinating aspect of Civ1.

I have wings for I can surf.
The hardest thing is closing down.
Peterval.

 

[Zwelgje]

Some research into Civilization splitting
Some research on Civ splitting??? Right! It must have taken you hours and hours to do this.
Very cool results though. It's always nice to try out some stuff to find out how the game actually works (and maybe to drive the computer a little bit insane.... )

 

[KiwiExpress]

damn, that mustve taken some time!

Just thought id mention that several times (over the last few years of playing that is) AI civs have split each other in my games. I cant remember for the life of me which AI was stronger and so on, or who had dissapeared...all i remember from one game was

1. The Zulus had captured Berlin
2. And the Germans had split into loyal (German) and rebel(English) factions

So i dont really know what to make of that! But it has happened more than once.

 

[peterval]

Thanks. It was a labour of love.
This is not a brilliant time for publishing research into Civ1, just when everyone is overdosing on Civ3, but here are a couple more findings on civ splitting.
There seems to be no random element in the occurrence of the event. Given the right conditions, (the human has weaker armed forces, fewer towns, smaller population than the target AI; and the target AI has enough towns - 4 or 5 I think) the split always happens. The proportions of the split seem random, except that I have always seen the original civ larger than or equal to the new civ. If there is only one colour free then it is 50-50 which of the two possible civs will appear. I have not studied cases where more than one colour is available.
An interesting point is that there seems no limit to the number of times the split can take place. In my current game at Prince level I have split the aggressive Zulus three times. On the first occasion 12 cities including the capital split 8 Zulu -3 Chinese. On the second occasion 10 cities split 6 Zulu - 3 Russian (I repeated the position and got 3 Roman instead). On the third occasion 9 cities split 4 Zulu - 4 French (I repeated this position and got 4 German). On each occasion there was only one colour available. As can be seen from the figures the splits did not deter Zulu aggression. On each occasion I allowed them to retake their previous capital but of course they had established a new one.

 

[peterval]

One final piece of data from the same game. A slight alteration from the previous conclusions.
Playing as the English I split the Indian civilization in 1934 AD. Before the split the figures were Indian: 8 cities, 71 population, attack values 79, defence values 74. English: 6 cities, 55 population, attack values 89, defence values 54.
But although the English attack values were greater, the Indians possessed a much larger number of units (musketeers mainly). As attack values are not as significant as I thought, this of course makes it easier to achieve splits.
Very important however was that I had to disband 48 caravans before the split would occur. Whether it is simply the number of units that is significant I do not know.
Very important also I had to wait a full intervening move before the program took account of the disbanding.
This time there were two free colours, white and light blue. I repeated the split twice and each time the Romans (white) appeared.
One more interesting point. This was the first split I have seen where the new civ was larger than the remnants of the old. Each time the Romans were given 4 cities and the Indians were left with 3.

 

[peterval]

I thought I would add to this old thread partly because people have been asking about the subject and partly because I was amazed by an occurrence a short time ago.
At King level there were only 2 AI civs remaining because Russia had been destroying everybody. I had only three cities because I was experimenting and 99% of techs. Russia was very advanced, and very powerful with 33 cities. Zululand was rubbish with 8 cities.
I was amusing myself detecting and sinking one laden Russian tanker after another. If they ever established a beachhead I was finished. Then I was astonished to see a message that the Zululand was swept by civil war and had split into 4 Zulu and 3 German. Sure enough when I checked, Russia had taken the Zulu capital.
In 10 years of playing Civ1 this was a first for me on two counts. I had never previously seen one AI civ split another (in Civ1); and I had never previously seen a strong civ split a much weaker one. Maybe it was a quirk in the program deciding I needed another opponent. Whatever, this sim is always surprising me.

 

[PritomD]

wow good research. ive only had 1 split...i greece took over germany capital, it became loyal germans, and revel chinese.