30-07-2002 10:45									#1
	Nukes and Skulls Put an end to popups!

I noticed that the number of appearing skulls depends on the city size if a city is nuked. Big cities up to 8 skulls, small cities at least 2 skulls (haven't noticed less skulls yet). And I noticed that the skulls only appear on the 8 surrounding fields of the city - no skulls in a distance of 2 fields.

Has anybody found out the algorithm behind that yet?

	30-07-2002 18:45									#2
	Put an end to popups!

As an added query, is that just the chance of them appearing? I think of a city with many ocean squares around it. If the skulls are set to appear over water, they do not appear.

	30-07-2002 21:01									#3
	Support Apolyton, buy Alpha Centauri

quote: As an added query, is that just the chance of them appearing? I think of a city with many ocean squares around it. If the skulls are set to appear over water, they do not appear.

From general testing, and general knowledge of the game algorithms, the method for skull generations is:

1. Calculate relative location & store (as bit-mapped nonsigned integer).

2. Overlay with game situation (likely uses simple bit logic operators for outcome).

3. Store skull outcome, if not superceeded by map constraints (result can be determined by hex view of .SAV).


Note: Here are some things which prevent override the appearance of a skull:

1. Water.
2. Airfield.
3. Another city.
4. Existing skull (e.g, no "double skulls").


These can easily be tested. Here is one way (there are others, too):

1. Create a late game, deity, high-tech (say thru FT100), no Mass Transit (& no pollution control) city.

2. Create several huge cities. Use at least one "mirror" as a control for each variation.

3. Set up skulls, water, airfields, etc. etc.

4. Run the game. Observe the outcome, and compare to the control(s) if you have statistical doubts.



Example: You make 4 identical cities, all with land terrain. Two have all 20 squares clear of skulls, airfields, cities, water, etc. These are the control cities. Two others have, say, 18 of the 20 squares (tiles) with existing skulls. Save.

Now, run. 100% (maybe 99.9) of the time, you will see the control cities producing a skull. 10% of the time, your 2 test cities will produce.


Realated issue: Nukes may be a special case for destroyed improved terrain. I have not tested to see if water in the surrounding 8 tiles is accounted for before or after the determination of which tiles to recieve destruction. Knowing the general programming style of Civ, i suspect the answer is compute the tile to be affected, check the tile for improvements, and pick one to be destroyed. Water will have no improvements, so nothing will be destroyed in a water tile. Airfields are exempt from nuke destruction, because of the way Reynolds, et. al., implemented them (airfields are "cities" without an indexed city parameters... e.g., a city by map storages, but not matrix allocation, and cities are not destroyed by nukes).