Bombardment of settlements in C3C 1.22

[justanick]

That is indeed a bit confusing. My orginal testing from 2013 does not explicitly refer to the case of simulated walls(but by implication it does). It seems i added it here as what i considered common wisdom. So now i am a confused. When exactly do simulated walls reappear after "destruction"? Also i like to point out that i did refer to C3C 1.22. Other versions may differ, but C3C 1.22 is the one version that matters unless you explicitly refer to say PtW or an even lesser version.

 

[Jivilov]

@ Lanzelot: Yes these were towns of a Great Wall-owning Civ (Portugal in this case). Whether or not that refutes justanick's rules is uncertain; my OS is Windows 10 which might have something to do with the discrepancy.

@ justanick: My version is Civ3 Complete 1.22 playing Conquests (Epic Games from 4000 BC) with Random setup for everything except map size (Standard).

 

[Lanzelot]

my OS is Windows 10 which might have something to do with the discrepancy.

Definitely not: the executable code is still the same old binary from 2004, bit by bit. And if it can be executed by Windows 10, then it must be executed the same way as it always was. The only differences during execution can be in Windows operating system libraries that are loaded by the process at runtime (graphics libraries, sound libraries, memory management, network libraries in case of multiplayer, hard-disc controller etc. etc.)
But the "rules of the game" are defined in the Civ3 executable itself (and in the .ini/.biq files that are installed with it), not in any Microsoft operating system libraries...!

We know that there have been significant rule changes regarding the bombardment rules between PtW and C3C. (And perhaps also between the different patches 1.0, 1.15, 1.17 and 1.22.) If we disregard walls for the moment, the behavior was like this: in Vanilla/PtW artillery was first destroying buildings, and only if there were no more buildings would then damaging units. In C3C first units are damaged, and once all units are red-lined, buildings are destroyed.

But since 1.22 (which came out in 2004), Firaxis has no longer changed the code, so the rules must be fixed since then.

 

[Lanzelot]

It seems i added it here as what i considered common wisdom.

And it matched with what I remembered from my own observations as well, So I'm very confused, too.

Perhaps the following could be an explanation: the "rule" as you stated it above, is not a fixed rule, but a probability distribution!
So instead of something like

City walls are hit first. This is even true, if the walls are only simulated by the world wonder "great wall".
If there is are no city walls or the settlement has already reached city level, then land units are hit.
If neither city walls nor land units are left to be hit, then citizens or buildings are hit, apparently with same probability.

it is something like

• If city walls, then city walls are hit with 95% probability, land units, buildings and citizens with 1.33% probability
• If no city walls, then land units are hit with 90% probability, buildings and citizens with 5% probability

(And of course we also have the "miss probability" that needs to be taken into account.)

So to someone conducting a small test series, it may look like a "fixed" rule, if he is "lucky" and always lands in the 95% or 90% bucket.
But now with more people conducting tests on different machines, we also get alternative results. (Note: the random number generator is an operating system function... So here we may indeed get different results on different OS versions, different hardware etc.)

 

[justanick]

If it really is a case of low probability instead of a hard rule, then i would probably have detected it back then. In either case it should be easy to test the hypothesis with the 2 files from 2013.

https://www.civforum.de/showthread.php?86307-Bombardierung-von-St%E4dten-in-C3C-1-22

 

[jarred!]

Just to make sure: all these tests (tmn, jarred!, Jivilov) were conducted with the Great Wall?
So that would mean, the Great Wall (or the simulated walls) is messing up the rules outlined by justanick in post #1 completely!?

My tests were conducted on the town with the Great Wall, and on a town with simulated walls due to another town in the empire having built the Great Wall. Buildings were destroyed in both cases, and units were injured once all buildings were destroyed (I only tested the latter on a town with simulated walls and not on the Great Walls town.).

When exactly do simulated walls reappear after "destruction"?

It seems the walls are immediately restored, at least they are restored before the town with a simulated walls is bombarded or attacked again.

And the .biq attached to the OP of the link does not test this directly, as it's the modern era, the Great Wall is obsolete, and the AI town we attack does not have a walls.

 

[tmn]

Before the target civilization has reached the industrial age this is not the case and world wonders can be destroyed like any other building.

I was unable to reproduce that without the Great Wall bug.

I've also tried to check for the Great Wall bug past Metallurgy as reported here, but couldn't reproduce it either.

 

[Jivilov]

FWIW in a current game England has the Great Wall and after destroying cultural buildings (temples, libraries, coliseums and in one case a Forbidden Palace) in Elizabeth's towns Pachacuti's Trebs started hitting defending units. Dunno if there were other types of improvements--probably were, the towns were part of London's first or second core. Interesting, no?