Au402 Dar5: 1000ad-1500ad

[nbarclay]

By the way, one thing that makes the trick I just described trickier in this game is that with probaly twenty or more surviving civs, techs don't devalue as quickly. That's an especially big deal on Emperor where AIs expect to get more from you than they give you in trades.

[jshelr]

We have a problem. I'll be interested if others agree.

This huge game with all civs in has the inverse of the earlier tech problem. I'm not sure inverse whoring is going to catch on big time as a term but here is an example.

After finishing US and changing to democracy, old Abe noted that there were several gpt trades about to expire and he decided to sell electricity around the wheel to replace that income. Electricity was a full 4-1/2 techs behind our leading position.

Abe received, conservatively but approximately,

1500 gpt
2000 gold net of gold paid during the trades
2 luxuries
nationalism
espionage

This was maybe 6 times the research cost of electricity. So, Abe blew 5,000 gold on a big hurry production splurge to celebrate.

The problem, IMO, is that the AI prices tech too dearly for this game setting. The human player can therefore undercut the AI reservation price and sell the tech so many times that the human civ makes a net profit.

To be clear, this is not an example showing astute gameplay. It simply shows the AI is not set up right for a 20+ civ world.

[Dominae]

I was wondering how long it would take someone to point this out as jshelr just did (it's something I guessed when I first heard the idea for this scenario).

Trading with the AI is exploitable. Most trades with the AI feel like exploits when I play, actually. The more AIs there are to trade with, the more the human player gains an absurd advantage in trading. As soon as the human player reaches tech parity, it's basically "game over" (assuming, of course, that the human is not at war and horribly outmatched in the field).

The most difficult games are those where the human has trading disadvantages (i.e. Total War, Isolation, etc.). In this game, unfortunately, most of the AIs have trading disadvantages as well because most of them are isolated for quite a while. Therefore the patient player should never fear a loss here if he or she just takes the time to trade into a superior position. I would not be surprised if many players win this game without ever expanding beyond the two home islands (the Gems one and the Wines one).


Dominae

[vmxa1]

I can only say that I am seeing that effect. I am able to keep large amounts of money, all of the lux by trading a tech when the deals expire.
I am trying to keep all alive to make the game get to as large of a size as it can to see the impact. So I am not taking any occupied lands and gifting stuff to the also rans.
I have seen civs that have zero cash pony up 80 gpt or more for a tech. Stronger ones will pay even more for it.
This has made them poorer and piled up tens of thousands in my pocket. Everyone is happy with markets and 8 lux at 10%.
Kind of like real life, every lives for today and not worry about tomorrow. I mean sending all you money and lux to one civ can not be helpful to your civ.

[Catt]

Trading can be almost exploitative, especially in this game.

A while back I started experimenting with largely peaceful games in an effort to better understand how to secure a comfortable win without offensive war. Getting a feel for trading ensures that all but the less desireable starts or circumstances means you have a shot at a win. The power of, and underprioritization by the AI of, Theory of Evolution and the structure of the Industrial Age tech paths means that if you can stay close enough to the leading AIs to get ToE, you can often seal the game. I still don't care much for Deity games as my strategies tend to channel into one pattern regardless of other factors, but even at Deity, ToE is often the turning point in the game.

In this AU, the tech costs are so high in comparison to standard map fare (a combination of the map size, the 80% water setting, and the number of civs) that a tech lead is way too powerful. I've noticed from the various DARs that a bunch of players are running 100% spending (science and entertainment) beginning in the early middle ages or even late ancient age, and still enjoying large surpluses per turn. How often do your "regular" games feature that situation?

I would not be surprised if many players win this game without ever expanding beyond the two home islands (the Gems one and the Wines one).

I think you're right on the nose with this. I played the huge map game (AU 208?) entirely peacefully, not even taking out the Babylonians, and a superior trading position was, all in all, pretty easy to secure and keep once you got the hang of trading down. Partly out of that peaceful experience, I resolved not to do the same thing in this game

Catt

[Catt]

Originally posted by jshelr
Catt: "When Bismark entered the Industrial Age, the Queen was pertubed to discover that somehow the German scientists had learned the secrets of Medicine in the blink of an eye without any effort at all."

This happened in my game as well, cutting a substantial tech lead down to nearly nothing in one turn as three science civs all went through the turn together and got different techs. Need to keep that risk in mind when trading for gold per turn with science civs that are right behind.

I've enjoyed the good fortune of having the science civs all get the same tech - Medicine. I was thinking of asking whether or not the new implementation of the truly random free tech for scientific civs in the latest patch was coded in such as way as to ensure that only the first scientific civ gets a random tech and all others get the same free tech -- I've not actually watched it closely enough to be sure in previous games or in the middle ages in this game -- but I gather that multiple different free techs among scientific civs are the norm?

Catt

[vmxa1]

Catt AU207 was the big planet. AU208 was total war for ever.

BTW does anyone know what the norm is for stealing tech (no spies)? Is it a sure thing using safety option?
I asked as I used it three times and got the tech each time. I sort of expected it was a long shot, but I had so much money. I normally never bother with stealing, I do plant spies, but don't use them.
I wanted to get all the stuff I had skipped before the ToE finished and they would not trade, so I stoled it.
Well I did it once to find out were the nukes were before they could use them in one game.
I am not intending to plant spies as with 20+ civs that is a lot of work and cash. Also I do not want to make waves. I am thinking of a UN vote as the way to end the game. A first for me with more than 3 civs left.
That is if it is turned on in this game.

[PeaSoup]

Originally posted by vmxa1
Catt AU207 was the big planet. AU208 was total war for ever.

BTW does anyone know what the norm is for stealing tech (no spies)? Is it a sure thing using safety option?

my diplomats have failed to steal a tech with the safely option. spending the extra cash for safely means even if you fail to steal something, your spies won't get caught (at least mine have never been caught on safely

[vmxa1]

That is what I thought, so it was just sheer luck.

[vmxa1]

Just looking at the system meter and it seems that Civ run at 50-52% of my CPU all the time and the storage is 280-288M. With all civs at war with someone and a few with more than one civ. The turn ran 9 minutes, but never used more than half the cpu and a quarter of the memory. In fact commit was 379M of 24xxM.
So it looks like anthing around 1.7 GH with 512MB should be able to run fine.

[nbarclay]

Originally posted by vmxa1
Just looking at the system meter and it seems that Civ run at 50-52% of my CPU all the time and the storage is 280-288M. With all civs at war with someone and a few with more than one civ. The turn ran 9 minutes, but never used more than half the cpu and a quarter of the memory. In fact commit was 379M of 24xxM.
So it looks like anthing around 1.7 GH with 512MB should be able to run fine.

I wonder whether the 50-52% CPU utilization figure is due to hyperthreading. On my old single-CPU system, Civ 3 was happy to eat up every available CPU cycle just sitting idle, and I had no reason to think it wouldn't do the same on a one million gigahertz system. The 50%-52% range matches what I see on my dual-CPU system, with one CPU maxed out and the other all but idle at any given time (although Windows shows both CPUs at 50% or just over because the system switches the load back and forth between them - which is probably good for minimizing the risk of overheating). If Windows interprets "one thread using all available power but no other threads contending for CPU power" as "one processor 100% loaded and the other idle" on a system with hyperthreading, that could account for a claimed 50%-52% processor utilization even with one thread using all the resources it can get a hold of. (If that's what's happening, Microsoft has a pretty serious bug in its processor utilization reporting algorithm where hyperthreading is concerned.)

Speaking of hyperthreading, if you can find out how to turn it off, I would suggest that you do so and see whether that affects Civ 3's habit of crashing. When I first put my new system together, I tried using two different speeds of Athlon. Other programs seemed to work fine, but PtW consistently locked up during my first turn. It's not inconceivable that whatever caused that might have a similar but more random effect on a system with hyperthreading. (Just a wild guess, but it seems like something worth trying.)

Nathan

[vmxa1]

I sort of thought that could be the deal, not sure. Normally all the games seem to drive cpu to 98-100 under XP. Now with the fake 2 cpu of hyperthread I suspected it was not reporting properly.
I will have to check out on the web, but I doubt I can disable it. I wish I had went with a 3GH instead to avoid any issues.

[nbarclay]

From what I've read, disabling hyperthreading is definitely possible. I'm not sure how easy it is, though. You might check www.tomshardware.com; I know they've done reviews where they tested with and without hyperthreading enabled, and they may mention how to do switch it. (Since I don't have a hyperthreaded CPU, I've never paid too much attention to exactly how it's done.)

[vmxa1]

Thanks I take a look.

[Catt]

When we last visited Elizabeth's empire, it was 1435 AD, China had been subdued and the invasion of America commenced. The English armies were bracing for an American counterattack. Most English forces were garrisoned in Buffalo, with only a single musket and two cavalry units holding St. Louis. Not surprisingly, the Americans went after St. Louis. A stack of 25 units, though largely outdated, could almost certainly take St. Louis, even if reinforced with additional units from Buffalo -- however, the Queen only had about 15 cavalry in America, in addition to 4 musketmen. Assuming perfect combat success for the English cavalry, the Americans would still have at least 10 units for the counterattack, with more no doubt making their way across the roaded but unrailed American countryside. Elizabeth could reposition troops each turn, bouncing American forces from target to target without an attack ever coming and all the while shipping reinfircements to America, but that didn't strike Elizabeth as within Queensbury Rules. Faced with the possible loss of the only English invasion force, and the almost certain loss of St. Louis, the Queen elected to make a 20-turn peace before resuming the hostilities.