FFH2 Performance Variables

Goodgimp

Warlord
Joined
Jan 20, 2006
Messages
261
Location
Salt Lake City, UT
I just wanted to ask a quick question about what factors in most heavily to performance of the late-game in Civ4/FFH2. What has the greatest impact on performance?

- Size of map?
- Number of AI?
- Other factors?

Obviously I know Python is a real slowdown, but that's not a controlled variable during game setup.

Also, given the release date of Civ4 I'd imagine it doesn't take advantage of multicore systems?
 
Mostly by the number of units I've found. This usually correlates with the number of AI but not always.

You're right regarding multicores. Your single fastest core is whats important, but at least you can alt-tab safely with multi.

If you want to play Huge, 30 AI Marathon games you'll get bored. If you want to play 10AIs on a Large Normal, then assuming you're not slower than my 2.4GHz, you'll rarely be waiting more than 10-15 seconds in the late game.
 
I found that putting all graphics setting on their lowest massively improves the number of crashes you get. Don't select the buttons that turn special effects and animaions off, though, or some unit graphics will spazz.
 
I found that putting all graphics setting on their lowest massively improves the number of crashes you get. Don't select the buttons that turn special effects and animaions off, though, or some unit graphics will spazz.

IIRC turning off special effects will screw with Tar Demons and Water Elementals and turning off animations will frack with just about every unit
 
I have a GTX295 for video so I'm not worried too much about performance during the turn, I'm more looking at the long (well, long for me because I'm impatient) wait times between turns in the late game.
 
More units existant = More delay between turns
Larger map = More possible units
More Civs = Faster approach of the unit 'limit' (delay between turns increases at a greater rate)
Fewer Civs (below map default) = more space for barbarians = more barbarians units = possible early onset of delay (worst on larger maps with fewer civs)
Barbarian World option = barbarian cities are on the map at the start = many extra units in play from the beginning of the game (increases the starting turn delay, worse on larger maps)
Raging Barbarians option = many more barbarians can spawn = increase in delay once barbarians start spawning (worse with larger empty spaces and on larger maps)
No Barbarians option = eliminates delay from barbarians (but allows faster expansion of civs and so may speed the onset of non-barbarian-related delays)

To reduce the eventual delay between turns, play on a smaller map. To delay the onset of turn delays, reduce the number of civs in the game, but not below the recommended number unless you are playing with the No Barbarians option. If you find the delay between turns to be a problem in the early game then you should definitely play on a smaller map, and should turn off Barbarian World or Raging Barbarians if you have them on.

You may also see a slight performance increase (over the course of the game) by decreasing the amount of land on your map. Every additional tile of land is another place the AI can stick a cottage, and that commerce translates to many additional units it can support. Most mapscripts have a Sea Level option (or the equivalent) that controls how much water there is on the map; more water means less land. Just note that the AI can't handle dispersed maps, like Islands, Archipelago, or even Continents, because of its current refusal to build and use naval units.
 
I think the correct terminology is each tile of land is where another farm can be placed ... as your commerce center, a farm can allow the AI to support many more units.

.... I just played a good-sized sp game where I was the Borg-Rhoanna ... using 0% science and nothing but mercenaries (with Iron from Galdur), to stomp over everything. I used my worker crews (plus irrigation from Construction) to set up every city with a commerce-farm machine (using Aristofarms). So every city started out making money, so I had lots of money. I started out getting about 150 gold per turn, soon I had 300 gold per turn, and during the mid-late game I was getting 400-600 gold per turn.

During the wrap-up phase I was getting 800-1000 gold per turn. And all of that gold equated to Mercenaries.

I ended with 270 mercenaries, and statistics showed that 249 had died over the years/campaigns. That means I had hired 519 mercs over the course of the game.

I kind of felt like the Unobtanium corporation on Pandora ... only winning by ALOT.
I mean ... by the end of the game, it was less about winning and more about being obsessed with making more and more money (per turn).
 
I've found that amount of units has a huge impact, besides other things of course.
Especially if you play with challenge options like increasing difficulty, where the AI gets more and more bonus over time, you'll see tons of units and SoDs once you reach the higher difficulties, which of course increase the delay between turns.
 
There are 3 AI civs that cause a lot more slowdown than the others: Amurites, Calabim, Luchuirp.

A huge map with none of them is faster than a standard map with all 3 of them.
 
Back
Top Bottom