Discussion in 'Civ3 - Hall of Fame Discussion' started by killercane, Apr 29, 2005.
To quote the above post.
"Without Longevity the city's cycle is:
40 turns at size 10 to fill the bin and grow
20 turns at size 11 to empty the 1/2 full bin and shrink
So without Longevity we get 1 point for 20 turns out of 60, or 0.33 points/turn on average.
With Longevity the cycle is:
40 turns at size 10 to fill the bin and grow to size 12
7 turns at size 12 to empty the 1/2 full bin and shrink
1 turn at size 11 to shrink again
So with Longevity we get 2 points per turn for 7 turns plus 1 for 1 turn, = 15 points per 51 turns, or 0.29 points/turn on average."
There is a fundemental error in this analysis. Without Longevity if a +1 food size 10 city takes 40 turns to fill the bin and grow it should take 40 turns to empty the bin before it shrinks back to a size 10 city for a .5 points/turn on average. On the other hand with Longevity if a +1 food size 10 city takes the same 40 turns to fill the bin and grow to a size 12 city it should take 14 turns to shrink to a size 11 city and another 1 turn to shrink to a size 10 city then the cycle starts over again. 14 times 2 plus 1 equals 29 so 29/55 equals .5272 points/turn on average and that is higher than the .5 without Longevity.
When a city grows it doesn't give you a full food bin. So there's no fundamental error. The cycle starts at size 10 with an empty food bin (because you've just starved down to 10), takes 40 turns to get 40 food, and grows to 11/12 with 20 food in the bin, so 20/7+1 turns to starve back to 10, not 40/14+1. The original numbers are correct, no longevity gives 20 points in 60 turns, longevity gives 15 in 48.
Nuts. You are right. The comparision is .33 without Longevity versus .31 with Longevity. At .02 differential I'm not sure it is worth deleting Longevity for the 1 point lost every 50 turns for any city that has a +1 food surplus.
But if a city loses one point with Longevity every 50 turns, and you assume 250 turns, thats five points per city. Huge milk maps can easily have 200 cities, that's then a thousand points. On Sid, that's 8000 points!
(I admittedly haven't studied the Longevity math in depth since I first understood, so someone correct me if I'm wrong.)
8000 divided by 540 turns
or about 150 score
However huge milk maps will often have 400, not 200 cities..
Oops. The disclaimer proved wise.
More points than Kuningas had over Moonsinger. Well, for one update anyway.
My Emperor game had 284 cities at the end and even though Longevity was built in 1505 AD I was using it to generate population at least until 1850 to move cities. Only about half of the 284 cities were +1 food cities so that is 142 cities for 150 turns. On Emperor thats 426 times 5 or 2,130 points. 2,130 divided by 540 equals 4 score. BTW 8000 divided by 540 turns equals 15 not 150.
I wanted to knock the dust off this thread, and even though I've been playing a bit of Civ 4 lately I had another go at a histo attempt. My last one was abandoned when I was just going to get mid 50K (turns last forever, if I am going to go through that the goal is 60K).
Ive played literally hundreds of maps with the Iro and Maya, which have differing strategies. If I generate an ideal Maya map, the expansion should continue until way into the ADs. With an Iro map, I want a juicy ironless neighbor or two to start a golden age at about 250 BC.
So anyways here is the start, Maya:
And after settling:
Lots of food, lots of wood to chop, and enough shields for 4 4-turners, the last one getting online at 1500BC. This led to my best QSC ever, 34 cities and 29 workers; now thats the power of the Maya! Hopefully this makes up for the low domination limit (4200). I am alone on my own continent, and I have built the Great Library to keep lux up (only one native lux) and Bach's.
boss ,the link in #2 is invalid, and I use google search for "civ3acad_maxscore" and can't find the exact match.
how can I go there?
Looks like the III War Academy got moved down a subdirectory:
The maximizing score article is by SirPleb at the bottom of the game mechanics section.
Finally I read the Longetivity calculation...
Very nice, SirPleb..
On another issue, I hate the Maya getting their GA too early if Pyramids are built.. it's the only reason I milk with the Iroquois.
You have a better strategy than early Pyramids, and a GA with Republic when you are 1/3 domination?..
Are you sure it's a BAD thing to get an early GA? Especially if you just got the Pyramids, effectively doubling your food as well as shields. That doubles your Setter/Worker build rate at an important part of the game.
I still think 20 turns of a GA would better be spent at 1000BC-1500BC with over 20 cities (still expanding) and in republic, than with just 1 or 2 cities benefiting from it sooner.
Having both built the Pyramids and having not built them, I like to not build them and use the GA in late MA when Im fighting my first war. I find you usually get to Mil Trad with the Maya at that time (500 AD)and thats the first time you can actually start increasing your pop/land dramatically. I dont know if pop boost from 1500 BC to 750 AD (steam power, rails, and electricity for irrigation everywhere) is worth the tradeoff in land score you get on the first 100 turns. Maybe if I had enough settler factory spots, and a huge land area to cover, the Pyramid shield cost would work better. The GA is nice but minimal in the long run, and better used to slaughter the hundreds of units a deity AI throws at you.
With the Iro I hope to grab ToA and the Pyramids when I start my conquest, preferably around 200 BC with accumulated chariots to upgrade.
On another note, what about razing cities? Im getting into the conquest phase here, and keeping the AI cities, which usually have markets, aquas, and granaries and are OCP spaced, seems the better option if you can win a war in ten turns, eliminating the AI. You retain the (soon to be happy) pop, and just have to deal with the culture flippage until you eliminate the AI. At first I was always razing cities, but Ive gone away from that and position cavalry between every few cities to take them back. I think this works a bit better, but havent heard it discussed much.
Personally, unless there's a specific Great Wonder I want, I raze. It's not only tremendous fun, but it's free workers, which usually means not having any worker upkeep during the milk phase (more money for research and maybe cash rushing). As you noted, there's no flips, so troops can extend much faster.
Granted, other AI might go in to fill the gaps, but that just extends their military out that much more and makes the next target weaker.
And to be honest, the AI city positioning sucks...
The AI is a bit weedy on city placement at times, but for the end game a lot of times they put them optimally. Take this picture for instance. Uskudar, Sinop, Antalaya, Konya, and Bursa are all pretty well placed, with varying degrees of infrastructure in place. When I get this war done (should be less than 4 turns, theyre on their last gasp of breath), I get a better point return on stopping resistance with my army as they move back down towards Persia, than growing them and then building the infrastructure to keep them happy from sizes 6-12. Only Istanbul needs to be moved. I really cant wait to trap all those persian units in the north and eradicate their army. There are about 100 headed up there .
Istanbul should be one tile north, Kanya one tile northwest, Uskudar one tile northeast, 113 one tile north, etc, etc.
You may want to download the final.savs of any Moonsinger, SirPleb or Kuningas milk run and study their milk phase city layouts. You'll quickly see that many of the cities in your screenshot are in horrible milk position because the incorporate too many coastal tiles.
*The only exception might be Sinop, because a cultural border expansion might snag some "free" sea tiles (as they don't count under the domination limit), and even that's rare/tricky to pull off.*
What's the Mapstat breakdown on your map's tile counts? You should be hitting nothing but grassland/plains/desert/flood plain before you think about any coastal tiles...
The pros and cons of razing really do come down to personal preference, but I'm sorry, those are bad milk positions. On the plus side, you've got plenty of time left in that game! Raze or not, you should take them all out and then worry about repositioning.
Separate names with a comma.