Ics?

[yin26]

Dan's ability or lack thereof isn't the issue here guys

The larger issue, of course, is if the aggressive AI is seemingly pushed back by culture alone, the threat wasn't much to begin with. However, as I recall he later said he was going broke from being such a pacifist (having to pay people off not to attack him), so it's good to see that the econ can kill you if you just try to build and build and build.

Still, we won't know till we play.

[korn469]

ok one last post for the night

let me explain the reason behind why food limiting growth instead of shields limiting growth is so huge

if shields limit growth as it does in civ2, then when a builder focuses shields on building they are falling behind in terms of growth of a player who uses ICS...when food limits growth, then the builder can still build and not sacrifice growth so building has less of an oppertunity cost associated with it, you sacrifice little growth for infrastructure compared to sacrificing major growth for infrastructure in civ2

yin

yes a weak AI could easily ruin a SP only game...think there is enough interest on starting an AI discussion thread?

[yin26]

I think there is great interest but nothing to discuss yet. One of the only things that would really get me to buy Civ3 is if I knew I could totally revamp the AI somehow and present Poly with some truly killer SP play. I would do that primarily by studying the strategies of our top players here and making specific AI profiles as counters. Not exactly a perfect solution, I know, but it you are an ICSer, you could d/l the Anti-ICSer profile and find a challenge. Same with the anti-builder profile, etc. Put enough of them in a game and set them up to not go to war against each other very easily, and there could really be something for SP across the board.

I am sure Firaxis has given that their best shot, but the AI can really only be programmed at the level of people playing the game, and as the strats evolve, so will the AI scripts have to change...and that could be a fun project to do later on.

[Skanky Burns]

Originally posted by yin26
One of the only things that would really get me to buy Civ3 is if I knew I could totally revamp the AI somehow and present Poly with some truly killer SP play.

Id have to agree with you Yin. An open-source, evolving AI would always be an improvement upon an AI written before release, and never changing.

But back to ICS:
Korn:
I agree with you too
Like i posted:

The only difference is that the perfectionist expansion will peter-out much faster, as they will start concentrating more on improvements (or pumping out military units, for the warlords), and will fill up available land-space much faster.

...while the ICSer will keep expanding indefinately.

Metamorph:
You misunderstood the thing about culture (ill admit, it wasnt too clear ).
Military units in a city prevent the enemies culture from stealing your city. Think of the Berlin wall after WW2. Heaps of military units on that wall, not to stop an invasion, but to stop their own people from defecting! Military units in your cities reduce the chance of your own cities from joining other empires.

About cash: ill have to test that out to see what happens (ive never played ICS before...) If its true, then money probs wont stop ICS.

About raising cities: Damn, i forgot about this completely!! This changes quite a bit of my *proposed* ICS strategy. I can imagine a large 'builders' city near the empire's frontier... impending cultural takeover. (I still think builders will have an advantage in culture) Rather than waiting for your cities to start leaving your empire, you launch an attack on the city, and burn it to the ground. The enemies culture is now no-where near your precious cities. Sure, the other players will hate you for it, but if you're an ICSer, they will hate you anyway

EDIT: Damned tags

[Father Beast]

OK Korn, If there really isn't a luxury slider, then "We Love" days won't make that big a difference, since they will be danged hard to get.

One further limitation on ICS, is that the ICSer typically builds his cities close together, making it hard to grab more of the map. Of course he could spread them out, and fill in later, but I don't know how viable that is...

Because the slow start applies to EVERYONE, the ICSer may well have an advantage on large maps, where he has space to get going. On small maps, you might as well forget it.

All in all, ICS will be a LOT more balanced with other strategies, and Deity may be a lot tougher without the easy win of ICS.

the disadvantage may be that the OCC will be much more powerful.

[Skanky Burns]

Originally posted by Father Beast
the disadvantage may be that the OCC will be much more powerful.

Maybe, but you'll have to figure in resources... at the very least you'll need some to build defensive units, and apparently for building wonders as well... Plus for some small wonders, you need 5 banks or 5 sam sites... which is a bit hard to do with only 1 city I crack me up

[Metamorph]

Wow, this thread is huge

Beast: "OK Korn, If there really isn't a luxury slider, then "We Love" days won't make that big a difference, since they will be danged hard to get."

WLTKD was always a silly idea anyway. Aside from being utterly unrealistic (as if any of these games ever approach realism ), it rewards the rich, since achieving this was cake for people with the appropriately broken wonders (Bach's, et al).

"One further limitation on ICS, is that the ICSer typically builds his cities close together, making it hard to grab more of the map. Of course he could spread them out, and fill in later, but I don't know how viable that is..."

Not sure I follow you here. ICSers build cities close together *because* they want to grab more of the map. Their intent is to have lots and lots of small cities harvesting the yummier tiles across the land. Further, they wish to do this at a more efficient rate, since the city tiles themselves are being harvested (and improved) for free.

In abstract, I don't really care where my cities are. Building them closer together gives me two advantages: they're easier to defend (and network); and it takes less turns to plop the settler down. Being able to build cities two turns after the settlers are produced (rather than wandering halfway across the continent for 15 turns) is a huge, significant influence upon the exponential curve of ICS growth rate.

"Because the slow start applies to EVERYONE, the ICSer may well have an advantage on large maps, where he has space to get going. On small maps, you might as well forget it."

Perhaps. But that all depends, I suppose, on how small is 'small'. ICS is an economic exploit; the less land available, the less production is available for exploitation. As the land shrinks, ICS becomes less and less appealing; but how low do you go?

Further: is this even relevant? Land size is a variable; but ICS is forever. I don't want to be doomed to an eternal existence in teeny tiny maps simply in order to prevent sleazing.

"All in all, ICS will be a LOT more balanced with other strategies, and Deity may be a lot tougher without the easy win of ICS."

I certainly hope so. What are the difficulty effects in Civ3, by the way? Is that known yet? In previous games, the difficulty settings were rather lame. I for one would like to see a much more interesting difficulty system, where a wide variety of parameters could be custom set per game by the player(s). This, in turn, would be evaluated on some sort of scale, to ultimately determine the overall effective 'difficulty level' of the scenario selected (a la Tropico).

"the disadvantage may be that the OCC will be much more powerful."

OCC?

- Metamorph

[korn469]

here is an example of how ICS isn't as overpowering in Civ3 as it was in civ2

If you are the egyptians then you are a religious and industrious civ...if things work like i expect then a 40 shield temple will cost the egyptians 32 shields (-10% shields for industrious, -10% shields on temple for being religious) and a settler will cost 36 shields...that is 68 shields altogether, the egyptians will also gain a +1 shield bonus to their city square, so if the egyptians builds a temple then a settler it will take them 68-40-25-25=-22 or 25 turns to build a settler with 22 shields left over...the same amount as time as what it would take any other civ to build, so the egyptians can build a temple, settler, and still have enough shields left over to build a scout or a military unit or to use the extra shields to generate gold (from what i heard capitalization works like it does in SMAC {ie you have that ability from the start of the game} and not as it does in civ2)

OCC=one city challenge

[Metamorph]

korn469: "here is an example of how ICS isn't as overpowering in Civ3 as it was in civ2"

I'm following your math -- I think -- but all this does is seem to indicate (to me, anyway) that ICS may be strengthened, not weakened, by an overall slower model. If everybody is slowed in the beginning, then everybody is slowed in the beginning. There's no reason to necessarily presume that an ICS player is somehow *more* slowed.

A non-ICSing player would not pop out a settler on turn 25. He or she would wait, build some military units, explore the area, grow the city to size 4 or 5 or 6, maybe throw in a worker somewhere along the way, make a bunch of cheap, low-tech city improvements... and *then*, when he's good and ready, build that oh-so-expensive 'first' settler.

Meanwhile, the ICSer has two cities working simultaneously. His population is growing twice as fast. Still 'slower' than Civ2; but still possessing more potential than his opponent. Then he eventually splits his cities into four. And then eight. And then sixteen.

And oh yes, it will again take far, far longer than it would have in the Civ2 environment. I know that; I understand that; I embrace that. But is it long enough? Is it enough of a restriction that the ICS curve won't be able to skyrocket in time before his 'normal' neighbors come a-knockin'?

Perhaps. But there are so many other factors. How does time flow in Civ3? What are the costs of techs? What are their effects? How does production micromanagement really work? What other uber-bonuses exist on tiles? What other free bonuses exist for cities? For nations?

What is the population growth rate model in Civ3? Is it still 10 food for size 2, 15 for 3? If so, then we're transforming a 10-turn city split into a 25-turn city split. ICS is 'slowed' by a factor of 150%. It will take 2.5 times as long for ICS to explode in insane power. 2.5 times as long, that is, when compared to Civ2 standards; not necessarily when compared to his normal playing neighbor, who is *also* impacted to some degree by the new standards for expansion.

And then there's the granary factor. Why should an ICSer waste time with silly temples, when he can build granaries instead? Doing so could potentially cut many turns off of the rate at which he can produce settlers. He's got the extra shields, after all; may as well make it worth his while. This is in contrast to Civ2, where population would usually grow *faster* than the ICSer could produce settlers! I recall often being excited when I opened a hut and found gold (rather than the 'best' find, a settler) because it meant that I could rush-buy three or four settlers in various cities that had already hit more than sufficient population but were lacking in shields.

I still feel as I originally did, korn. The 2-pop rule is a band-aid. It's a huge, tremendous, violent band-aid though, and its ramifications still need to be carefully weighed and tested. We wanted to try it ourselves once, long ago; looks like we're going to get the opportunity now.

- Metamorph

[Edward]

posted by yin26
I would do that primarily by studying the strategies of our top players here and making specific AI profiles as counters.

Better yet, have the AI adopt those strategies.

Of course the big problem for Civ3's AI is the fact that it's a new game.

If Firaxis had decided to make Civ3 be "Civ2 with a couple of rules tweaks but mostly just an improved AI", that would be great. Just adopt some effective Civ2 strategies. However, Civ3 has a bunch of new factors: culture, armies, resources, etc. We still don't know what the best strategies will be. A human who has played Civ3 for a year or so will easily beat any Civ3 AI developed today because the best strategies just aren't known yet. And past experience has shown that even a software production house that knows good strategies generally can't create a tough AI. (Not a slam on software companies, just a realistic look at how hard it is to make smart AIs. Also, I think other game factors are higher in the companies' minds.)

There's a host of things Firaxis could do to somewhat improve the AI. They could improve the AI's combat prowess if they 1) had the AI create a large (8+ unit) assault force before attacking and 2) spent some time creating algorithms that would let the AI attack one city with all the units at once. I'm not talking about having them move as one easy-to-kill stack. The AI needs to have 2-move units wait one square away from the city (on different squares so they aren't taken out as a stack) then move them all in (via flatland or roads) on the same turn. OR have the AI fortify a number of defenders on some rough terrain adjacent to the city, then move in the attackers under their cover, then attack with all the attackers at once. While this seems easy, the difficult part to code (barring naval assaults) is the movement of the troops to the city, especially when terrain movement costs and zones of control come into play. And what about that fortress you have in the way? Does that become a preliminary target, or does the AI need to make some "ignore ZOC" units as well to sneak the attackers by. So many details...

[Sava]

"but Metamorph, if we fix ICS then the game would be realistic and hence; not fun"

That's my unrealist personality breaking through

" no realism is stupid, we want to cheat, we want the game to be about super beings that repdrouuoduoere "

DOWN WITH YOU UNREALIST PERSONALITY....


ahhhh now I'm back

ICS is one of the many tools of the devil. I agree with others that the culture model will help hamper ICS as well as the whole transportation thing.

[Metamorph]

SoulAssassin: "but Metamorph, if we fix ICS then the game would be realistic and hence; not fun"

ROFL!!!!!!!

- Metamorph

[Sava]

hehe, sometimes I write long posts like you metamorph and I rant, but every so often I just snap and go crazy

[korn469]

Metamorph

If everybody is slowed in the beginning, then everybody is slowed in the beginning. There's no reason to necessarily presume that an ICS player is somehow *more* slowed.

you are drawing the wrong conclusion here, because the ICS player is slowed down to a greater extent than a non ICS player, and in my opinion this seems to bring ICS close to other strategies

like you said

This is in contrast to Civ2, where population would usually grow *faster* than the ICSer could produce settlers! I recall often being excited when I opened a hut and found gold (rather than the 'best' find, a settler) because it meant that I could rush-buy three or four settlers in various cities that had already hit more than sufficient population but were lacking in shields.

and this is true, but in civ3 this won't be the case in the early game, you'll have to wait for population to grow

A non-ICSing player would not pop out a settler on turn 25. He or she would wait, build some military units, explore the area, grow the city to size 4 or 5 or 6, maybe throw in a worker somewhere along the way, make a bunch of cheap, low-tech city improvements... and *then*, when he's good and ready, build that oh-so-expensive 'first' settler.

i really don't think that this is an accurate estimate of what a non ICS player will do first thing is like i showed you before, the egyptians can build a temple, a warchariot, and a settler all in the same time it takes an ICS player to build their first settler, in civ2 this wasn't possible, every shield not devoted to early settler production would be a waste, in civ3 this is not the case

after careful consideration i now think that the American Civ will be the best ICS civ and most likely their build orders will be

Capital City
pop growth
granary (54)
pop growth
settler (36)
total: 90-90=0

then they pump out two or three more quick settlers before becoming the Supercity at the heart of the empire


Secondary Cities
warrior (9)
pop growth
worker (18)
pop growth
granary (54)
pop growth
settler (36)

not exactly sure how the math works out on the second build order but it shouldn't slow it down by that much...and with a granary then ICS will be more effective plus it gives the player early defense, and a workers to build roads and other tile improvements

to me that is the best Civ3 ICS build order, using the americans who are the best Civ3 ICS civ

like i said though, the Egyptians

Egyptians

Capital City
pop growth
temple (32)
scout (possibly 2) or 2 warriors or capitalization (18)
pop growth
settler (36)
total: 86-90

the egyptians build their first settler on the exact same turn as the americans and have four shields left over, while they cannot pump out settlers as fast as the americans (or other expansionistic civs that have built granaries) most likely they will build a worker next to build a road in order to pay for their temple, and the egyptians are now well on their way to establishing a cultural stranglehold on the game since the earlier you build a culture enhancing structure the more benefits you get from it (the earlier you build a temple, the more culture it generates per turn as time goes by is what firaxis said)

Secondary Cities
warrior (9)
pop growth
worker (18)
pop growth
temple (32)
pop growth
settler (36)

the one point that you do have Metamorph is that this is just a bandaid solution, and the reason why that is true is because it seems that 2 pop settlers are an early game only solution...yes you are spending over twice the amount of food to build a settler now, but look at these figures

civ3

city X
2 pop=20 food
3 pop=30 food
settler
2 pop=20
3 pop=30
city X=size 3
total food=100

city X1
2 pop=20 food
3 pop=30 food
city X1=size 3
total food=50

so for 150 food you can take a size 1 city and turn it into 2 size 3 cities, while it takes the same amount of food to grow a size 14 city to size 15...so even discounting population booms from WLTKDs and adding in a 2 pop settler a player using (abusing?) the growth mechanics of Civ3 can gain 5 times the pop and can end up with 6 times as many squares worked

civ2

city X
2 pop=20 food
settler
2 pop=20
city X=size 2
total food=40

city X1
2 pop=20 food
settler
2 pop=20 food
city X1=size 2
total food=40

city X2
2 pop=20 food
city X2=size 2
total food=20

so in civ2 for 100 food you can gain five times the pop and eight times the squares worked for the same amount of food as what a city growing from size 9 to 10 will get, plus they now have 3x as much support

if i could change anything i'd would make the food boxes a fixed size (at least per epoch, like 20 food in the ancient era, 30 food in the middle ages etc) and then leave settlers at 2 pop points that would certainly fix ICS in my opinion...as it is, smaller cities still gets big rewards over larger, and that is the basis of ICS

[Father Beast]

Good gosh, I'm gone for a few days (road trip) and this discussion continues...

Metamorph, You have said most of the stuff I would have wanted to say. I agree that ICS will still be a force to be reckoned with.

On reflection, I think the 2 pop settler does little to deal with ICS, since it cripples the non ICSer, while only annoying the ICSer.

Korn, I am finding little difference between your 2 scenarios. They both look like ICS to me...

It seems to me that the basis of ICS is still available. if having more cities is desirable, then having lots more cities is very desirable. pretty soon you become the behemoth Metamorph often describes.

OK, some numbers, assume 2 food, 1 shield per square, need 20 food to get to size 2, 30 food to get to size 3.

ICSer:
capitol:surplus 2 food, 2 shields. build granary first.
turn 10: size2, 2f,3s.
turn 17: granary done, start settler.
turn 25: size 3, 2f,4s
turn 29: settler built, size 1, 1f,2s. build another
turn 31: capitol size 2, 2f,3s. city2 founded. build worker.
turn 39: capitol size 3, 2f,4s
turn 41: city2 size2, builds worker, size 1, build granary, pattern of capitol.
turn 42: Capitol builds settler, size1, 1f2s, build another
turn 44: settler builds city3, pattern of capitol. Capitol 2f,2s Worker connects city2 and capitol
turn 45: Capitol size 2, 2f,3s
turn 47: Worker connects city3 and capitol, starts to irrigate (Since it is obvious that the lone worker can keep up with connecting the cities, I will ignore him from now on)
turn 51: city2 size 2
turn 53: Capitol size 3
turn 54: city3 size 2
turn 56: Capitol builds settler, size1, build another
turn 58: city2 builds granary
turn 60: capitol size 2, city2 builds settler, build another. city4 founded. same old pattern.
turn 61: city3 builds granary
turn 62: city5 founded.

OK, I think that's enough for comparison.

the only difference in this other civ, is that he spaces his cities decently, in a non ICS way. it takes the settler 5 turns to build the next city.

Non ICS:
capitol:surplus 2 food, 2 shields. build granary first.
turn 10: size2, 2f,3s.
turn 17: granary done, start settler.
turn 25: size 3, 2f,4s
turn 29: settler built, size 1, 1f,2s. build another
turn 31: capitol size 2, 1f,3s.
turn 34: city2 founded, build worker. capitol 2f,3s.
turn 40: capitol size 3. 2f,4s
turn 44: city2 builds worker, starts connecting to capitol, city2 build on capitol pattern.
turn 45: capitol builds settlers, build another.
turn 46: capitol size 2, 1f,3s
turn 51: city3 founded, build granary then settler. (yawn) capitol 2f,3s
turn 54: city2 size 2
turn 56: capitol size3, worker finishes connecting city2 and capitol, starts on connecting capitol with city 3 (will finish by turn 68)
turn 58: capitol builds settlers, size 1, start again.
turn 61: capitol size 2, city2 builds granary, city3 size 2

By turn 61, the "ICSer" (they are really both settler producing madmen) has 3 cities size 2, and 2 brand new cities. once the granary is built, the city n=can pump out another settler every 12 turns. the "non ICSer" (only difference is in spacing of cities)has a slight disadvantage in producion time, but is mainly hampered by travel to founding time.
so the one who puts lots of cities close together has an advantage, even in just building settlers!!

OK, this doesn't take into account unhappiness, variance of terrain, etc. but as a basic model, it shows the ICSer's advantage simply by jamming his cities together.

and heaven help us if granaries provide culture. by the time an ICS city put out its first settler, it would have that one square around it.

I think the 2 pop settler doesn't really fix it, and isn't that big of a band aid really.

our only hope is for culture to have a serious impact.