Announcement

**solo** · May 17, 2001, 22:09

Great game, samson! I've already printed it out and look forward to reading how you did it! Let me be the first to say that this early landing the game is now the one to beat!

**Scouse Gits** · May 18, 2001, 06:41

First Class Work - go to the head of the class

------------------
Scouse Git[1] -- git1@scousers.net

"Staring at your screen in horror and disbelief when you open a saved game is one of the fun things of a succession game " - Hueij
"The Great Library has been built!"
"A short cut has to be challenging,
were it not so it would be 'the way'." - Paul Craven

**La Fayette** · May 18, 2001, 08:25

WoW again samson.
This makes me very sad, thinking about all those 10000000 beakers all of us use in vain researching junk techs, instead of reading you seriously and optimizing tech tree and gift of techs.
I bow.

------------------
aux bords mystérieux du monde occidental

**samson** · May 18, 2001, 09:02

Thanks for the kudos, all.

Solo,

For me, your 776 landing is still the one I want to beat.
Your approach, in that game and your 1076 game, is quite different
than what I used here. You build a perfectionist-style 747 jumbojet
of caravans and research bonuses that cruises in for a smooth landing.
I built a OCC-style flying-gascan which sputtered across the finish line
with the wings falling off. Your research pace was slow at the beginning
but hit a 1 tech/turn stride and maintained it. Mine was faster upfront
but only got to 2 turns per. I think a strategy which combines elements
of both approaches might be a real winner for early AC landings.
Perhaps a strategy discussion on this topic is in order.

I've been bitten by the early landing bug.

samson

**hbourj** · May 18, 2001, 11:46

I did never start with 5 start tech, in Deity level. It seems impossible to me. Then if i had MONARCHY as you, i should have done it in -3950 ! So i think possible to do my landing before ! Really I am very surprised by the kindness of IA with you.

**solo** · May 18, 2001, 22:43

samson,

I'm back and look forward to a discussion about early landings and will kick it off here.

I took a little time to compare your game with my 1056 arrival game, and have a few things worth mentioning:

1) What first surprised me was that my 2 turn advances started about the same time as yours, although yours were by science alone, and mine were assisted by commodity deliveries.

2) It took me until 2950 BC to acquire my first 3 techs. Since I started with 2 freebies vs. your 5, it took me 20 turns just to catch up, and research costs throughout the game were always 3 techs heavier for me.

3) We were both even in the first flurry of tech trading with the AI, with about 7-8 techs each. Later on, you managed 1 or 2 more useful trades, whereas my embassies helped me to 7 more useful techs through trades.

4) You did better with hut technology with 8 free techs to my 1. So we came out about even in useful techs NOT acquired through research, with about 16 each overall.

5) Where you really outdid me in learning techs was the period before when we both started getting 2 per turn. By that time, you had all science improvements and wonders in place, and all I had was a library in my size-8 SSC. You were quite far ahead on the tech tree. My helper cities were full size and the profitable trade stream had begun, but my SSC was still in the process of growing.

6) I was too far behind and ended up launching 7 turns later.

I'll list some problems with my approach and what I think are some problems with yours:

Me) My SSC develops too late, and as said in the earlier post, there is a lot of extra overhead associated with it, especially in keeping my guys happy enough. It also takes me a steady stream of cargo deliveries plus extra science improvements in helper cities to maintain 1 turn advances. Once your SSC is set, your 2 turn advances are pretty much automatic. I also have a problem with boats that are so vital to overseas trade, and building Leonardo's is a very expensive way to maintain an active fleet in place that will be happy under Democracy. The transition to that government comes much later for me and I miss many opportunities where I would like to set science over 80%. I am also slower to Republic, and must take a tech tree detour to Monotheism. Sometimes I have to race the AI to an important wonder I'd rather build later.

You) Most all of your beakers come from one place, the SSC. With helper cities and commodities to trade, I have much more flexibility, and get the most out of available beakers. You often have too many, or not quite enough to complete the next advance. Your income is less dependable than mine, and a bit smaller. Your cities are more prone to "minor disasters" and your SSC is more vulnerable. Not much trade with the AI makes them slower learners, without much tech to trade after the first stage of the game. Colossus is necessary for you, and it expires with the discovery of Flight. In general, if things don't go according to plan, you have fewer options than I do. Establishing embassies is not as easy for you as for me.

I think there are things we can borrow from each others' approach that will eliminate many of the weakenesses listed above. I also think much more can be done to expedite the trip through the tech tree taking full advantage of your key civ discovery, embassies, and by designing an approach somewhere between yours and mine.

**solo** · May 19, 2001, 00:19

samson,

Looking over your log tells me that your shoestring strategy is the quickest way to AC when cararvan rehoming is not an option. I have several ideas for refining it, that I would like to test out in an actual game using your start or one similar to it.

I am leaning to your method as the superior one because who cares if you can only manage 2 advances per turn if they start coming early enough. You get a tremendous headstart by just concentrating on the bare essentials. My successful streams of one turn advances incurs a lot of overhead in the form of additional wonders (Michaelangelo's, Leonardo's, Hanging Gardens), and many extra units you don't need to bother with, such as additional settlers, boats, diplomats, and especially all those extra caravans and freights needed to generate enough trade to justify the system. About 6-8 size-8 helper cities were big enough for profitable trade, and seemed easy enough to establish and maintain. The game is skewed so favorably towards trading commodities, that I had never seriously considered an alternative such as yours that zeroes in on science in such an efficient way. No one seems to be able to sift through all the options (I am tempted to say rubbish) the game distracts us with better than you do.

(Sorry, I have more to add but have to go)

**samson** · May 19, 2001, 00:41

quote:

I did never start with 5 start tech, in Deity level. It seems impossible to me. Then if i had MONARCHY as you, i should have done it in -3950 ! So i think possible to do my landing before ! Really I am very surprised by the kindness of IA with you.

hbourj,

There's a great piece of work on starting techs by the Scouse Gits in the GL.
Look under Technology/Starting. According to their data, Monarchy occurs
as a starting tech about 1% of the time on random maps. If your starting
civ is "civilized" that goes up to about 3%. However, if you play as one
of the "militaristic" civs, you will never get Monarchy.

5 starting techs is not all that rare. I can usually get at least one start
with that many or more techs in 15 minutes of random map restarts.
It's not the kindness of the AI that creates them, but the combination
of probability and a tolerance for tedious tasks.

I, too, believe it is quite possible that you or others can best my landing
date with a comparable start. Have a go.

samson

**samson** · May 19, 2001, 15:34

Solo,

I agree with your analysis.

My game was faster in research at the start, yours at the end.
Yours was a strategy you had used successfully before. Mine
was made up as I went along, except for the OCC-based part. I'm thinking
a good compromise might be a OCC-style capital with ICS-style outlying
towns, low on infrastructure, building lots of caravans for the money
and beakers. We've got to deal with the happiness problem, though.
Temples and size3? Or a Wonder?

I also think an aggressive hut-hunt should be part of the strategy
as well as using diplos and tech-gifting to soak the AI for as much
as we can get out of them.

I've got to think on this a bit more. Later,

samson

**Mark_Everson** · May 19, 2001, 20:39

My congratulations to both of you on your excellent games! Not long ago I thought it was basically impossible to get a 1500 landing at deity!

My question is for Samson on the 'aggressive hut-hunt' (say that 10 times fast ). IMO huts are very much a two-edged sword, and I'm not talkin' about barbs. Huts can Really screw up your tech progress if you get lots of 'junk' techs that don't fit your strategy. So doesn't aggressive hut use mean you may well criple yourself in otherwise perfectly good games?

-Mark

BTW I play MGE so maybe the huts are nicer in 2.42 if that's what you're using.

**solo** · May 19, 2001, 20:41

Mark_Everson,

Just noticed your post and will give my opinion whether you want it or not! Sure there is a chance of getting unwanted techs from huts, but odds are that the tech will be useful, and it could be one you desperately crave! So many good things can come from huts, that the potential gain far outweighs the risks. An unwanted tech from a hut is an inconvenience, but hardly a major disaster.

samson,

When you say ICS-style outlying cities, do you mean as many as possible? I tried this first on the 776 landing game map, getting 20 tiny cities, and it took until 1020 AD to get Space Flight. I only used the Elvis approach to control happiness in the tiny cities. As long as each one produced a few arrows and at least 1 shield, that was all I cared about. So far, all ICS-style attempts I know about by the most able practitioners are landing at 1075 AD or later.

For me, fewer outlying cities on the same map worked much better, with a 776 landing. Granted that knowledge of the map helps, but not that much. You also did very well on a first attempt with a non-ICS approach.

Yes to aggressive hut hunting!

A few ideas, based on a combination of the best features of both approaches:

1) Maybe priority 1 should be a fully operational size-12 SSC to get science going quickly. Size 12 may end up being optimal, which could eliminate Sanitation as a required tech. While science techs are pursued, an AI is apt learn Medicine (needed for Chemistry) and Sanitation, which can be traded for later, if it is needed.

2) In the meantime, aggressive exploration and hut-hunting and embassies as early as possible with each AI.

3) Add helper cities around the SSC early, slightly overlapping on its specials. Founding the SSC and 1st helper immediately overlapping on gold, will give a nice boost to early research. SSC development takes precedence, but other helpers can be added, as convenient. They can attend to the Queen bee, supplying caravans for wonders, military units, and settlers for development. Depending on surrounding terrain and specials, probably about 4-6 planned helpers is good. The rest can be minimalist advanced tribes as in your game, accumulating caravans for the spaceship.

4) Hanging Gardens is probably best for early happiness, facilitating celebration of the SSC, and acting as a temple in the helpers, and even getting them to celebrate spontaneously when their workers have access to trade specials. If the AI research to Monotheism, there will be an option later to replace HG (which expires with Railroad) with Michaelangelo’s. If the AI don’t get to Monotheism, Oracle is worth considering, instead, since Theology is apt to remain undiscovered. Oracle may have to be captured, which would add some extra adventure!

5) Let the core helpers grow naturally. When the SSC has all science in place, techs should be coming every two turns, and the government should be Democracy, and Magnetism should be known, so that some galleons can be built. The focus can switch to trade, which will be more profitable with bigger cities, and the income can be used for happiness improvements, as needed, and for science improvements in the best helpers. With caravan and helper city beakers, quite a few 1 turn advances are likely.

6) When Railroad is acquired give it to all the AI. They love Railroad and one or more will start to build Darwin’s. If a needed tech is being studied (checked with embassies) all is well. One free tech is assured, maybe two, the AI pays for the wonder, and no research time is wasted. The cost of embassies is justified for this alone! Caravans can be saved for the rocket.

7) Give all AI Monarchy, Republic and Democracy quickly so they will study harder with their “great minds”.

8) Leapfrog techs, when possible. The early parts of the Warrior Code and Horseback Riding tech paths, of which the AI seem so fond, can be left for them to enjoy, as can selected techs elsewhere. Missing techs are even better than freely acquired ones.
[This message has been edited by solo (edited May 19, 2001).]
[This message has been edited by solo (edited May 19, 2001).]

**samson** · May 19, 2001, 23:18

Mark,

Yes, huts can be a problem. Unwanted techs can slow research of crucial
early discoveries, mess up the tech tree, or trigger side effects.
But like Solo says the benefits outweigh the risk. A NON settler, an
Advanced Tribe, a timely bit of cash, a vital tech. These can impact
a game enormously.

I think hut-hunting would be most effective, tech-wise, in the phase
after the first 20 techs are acquired. That's when the research lags
and a good tech can cut 7 or 8 turns off your score. Also, at that point
the penalty for carrying a "junk" tech is less than earlier in the game.
So perhaps some strategy can be applied as to when to hunt.

Solo,

I was just thinking out-loud when I said ICS.
I really think that each city founded (Advanced Tribe or chosen spot)
should be developed as the terrain and strategic location dictate.
A too rigid guideline (only size3 cities or all perfectionist-style cities)
is going to miss opportunities. If outlying cities have great trade specials
they probably should be developed as mini-science cities or trade centers.
A gold special in a hilly region might be good for a size 1 city.
But in any case, the focus must be relatively short term. Too much infrastructure
would be a waste. There isn't enough time to fully develop much except
the SSC.

I think our next goal should be to land a ship by 500AD.

#1 - Yes, a size12 SSC with great trade specials can bang out almost
as much tech as a size 30 SSC. And it can hit the ground running.
With Gold you can be doing techs in 3 or 4 turns right from the start.
In my game I eventually went to size 16 to bring those ocean squares
on board, but adding a sewer late in the game isn't too hard.
I never built ST. Didn't need it and that saved 6 caravans.

#2 Embassies, definitely. They allow fine-tuned control of tech-
gifting and, in theory, multi-plexed research. I think a city founded
near another civ should push out a diplo in quick order to get an embassy.

#3 Helper Cities. A couple could be good. 5 or 6 sounds like too many.
What is their exact role? Caravan production is one. I can't quite
see the overlapping specials though. The SCC is going to quickly leap
to size 6 or 7 once Republic arrives. I'm a big fan of Early Republic
in OCC. But what about with multiple cities? Is there an optimal time
to switch from Monarchy? Is it worth it to delay Republic and simply
celebrate in the SSC under Monarchy to get the extra arrows? Or should
we use Republic to celebrate, zoom up to size 7, then drop back to
Monarchy for expansion while celebrating in the SSC?

#4 I don't know about happiness Wonders. I'd love to get along without
them. HG might be the best if built in the SSC, but I can't see delaying
COPERNICUS to get HG first. And if HG is delayed too long, it becomes
pointless. Railroad comes up fast. I had it by 200BC and we have to do
better than that. That's the problem with all the infrastructure stuff:
the faster we go, the more it becomes the thing that slows us down.
MIKE's is the best of bunch, but it's two techs off the main path.
I almost think temples and colosseums, as needed, is the answer.

We won't be capturing any Wonders, I think. We're going to have to move
too fast for that. Maybe it all comes down to cash. If we can get enough
money, we can buy anything we want.

#5 In my games, I can usually get a great start. Hitting 2 turns/tech
early. I want to graft on your end game with its 1 turn/tech phase
without slowing down my quick start game. I know this is going to involve
trade and multiple cities. Without using caravan rehoming.

#6 Letting the AI build DARWIN's is a risky business. They often waste it
on Communism or some such drek. I think it best to build it ourselves.
Preferably soon, to help get to Automobile as quickly as possible.
With Superhighways the SSC can hit 1 tech per turn for a short time.
If we have a brigade of other cities cranking out 100 or so beakers
and caravans waiting to be delivered, we can sustain that 1 turn/tech.

If we're already getting one turn/tech Darwin's may not be necessary.
But at two turns/tech it save 4 turns (80 years). That's worth it.

#7 Agreed. Give away new government types to all.
Also, Writing and University.

#8 Yes.

samson

**Jaguar** · May 20, 2001, 18:48

IMO it is also extremely advantageous to give the AI refrigeration and sanitation, especially your trading partner, because those fat 20+ AI cities really give you huge trade bonuses.

**solo** · May 20, 2001, 21:56

(Edited to correct the spelling of Xin Yu's name. Sorry Xin Yu, it was a typo.)

samson,

I agree with your general comment that any non-SSC city’s terrain and circumstances should be the main thing in determining its role and development. Adhering to any rigid rule of thumb would be silly.

With time so much a factor, especially in pressing for arrivals like 500 AD, the fewest number of cities needed to make things work out needs to be determined. The reasons for having extra cities:

a) Fast spaceship construction

Sometimes working backwards can help provide an answer. The fastest ship needs 24 parts. At least 3 turns will be needed to construct it, since the last 3 required techs: Plastics, Super Conductor and Fusion Power, follow Space Flight, so certainly no more than 8 cities are a necessity. This is good, because more than 8 start to cause extra unhappiness in a Republic. If 4 or more turns are used to make a spaceship, which is more likely, no more than 6 cities are needed. More than 6 cause extra unhappiness in a Monarchy.

b) Enough beakers for 1-turn advances

From my experience, one size-8 city with 3 trade routes and without any science improvements will produce about 25 beakers. To keep a steady flow of caravans going, so that one can be delivered each turn, will require about 5 cities. However, additional caravans are needed for wonders and for the spaceship, and cities don’t always produce good commodities, so a few more than 5 cities may be a good idea. An average of 250 gold/beakers is possible from demanded cargos delivered overseas from size-8 cities, so trading generates many beakers. Unfortunately, a delivery each turn is usually required to sustain 1-turn advances. So the optimum number of cities is probably between 5 and 8. This would settle the question, if it were not for the happiness problem.

I’ll reply now, to the comments you made:

1) On to size 16 is good, but we should probably wait until an AI learns Sanitation and trade for it.

2) No difference of opinion about embassies. I should add the sooner the better. The AI are only good for so many advances in a game, and hopefully each one will count, without repeats of something already discovered by another AI or us.

3) The role of helpers is partly explained above. Their presence speeds construction of wonders. They build units needed for trade and defense and terrain development. Their raw science contributes towards 1-turn advances. But mostly they are there to produce caravans for trade. The reason for having them overlap the SSC‘s specials, besides speeding early discoveries, are: mainly because the specials can be used to maximize a helper’s trade arrows at the moment commodities are delivered (imagine the difference in gold/beakers received, when two workers jump onto gold specials from their normal locations); that last caravan needed for Cope’s is only 1 turn away from the SSC instead of two; better defense. Obviously, if there are enough quality trade specials around for every helper to have its own, overlap is not needed, but gold is fairly rare in random maps. A thing to keep in mind is that size 5 cities can hire Einsteins and tax collectors. A variation of Xin Yu’s strategy bearing the same name may be useful.

4) Very good points and I’ve given this a lot of thought. I have to agree that happiness outside the SSC is the killer. I would say start the SSC and one helper right away, with the helper overlapping the gold. Build a settler first with the helper and have it replace the one normally used to develop the SSC site. Now develop the SSC just the way you normally do, as if the helper were not there. Build Cope’s there ASAP, etc. It will be completely up to the first helper to happy up itself and to introduce companions as best as it is able. Advanced tribes may produce others, nearby. Later in the game, after an AI has discovered Monotheism, trade for it and build Michaelangelo’s if it seems prudent. There will be more cities to justify the cost, and expensive colosseums can be avoided. Later Mike’s would also cover an SSC expansion up to 16 citizens. How does this plan sound? It starts things your way, but provides a chance to switch to my way later.

In your game, your first minor city came much later in the game, too late to get the helper ball really rolling.

In my game, (largely patterned after the success of the 776 landing game, which was designed for 2 advances/turn and rehoming) I would have done better by founding my SSC earlier and should have built HG before Michaelangelo’s. If I had, my SSC would have been growing from size 8 to 12, while my helpers were going from 3 to 8.

I was not really serious about capturing Oracle!

5) Implementing trade will come a little later, and may not be enough for sustained 1-turn advances, but maybe 2 advances every 3 turns after a fast SSC start will be enough. Tech leapfrogging may be successful enough that not as much trade will be necessary. Games will have to be played to find out by trial and error.

6) There is no risk if an AI is building Darwin’s. Imminent completion of AI wonders is always announced a turn before the event, allowing a rush job to beat them to it, if needed. If our techs are coming every 2 turns, and we let them build it after checking to see that their current project is useful to us, the most we can lose is 2 turns if we let them complete it. The cost is even a bit less, because we can trade for what they get later, to keep our beakers lower now, plus we save 6 camels. The earlier the arrival goal is, the harder it is to accumulate enough caravans to build Darwin’s and do everything else. Wasn’t it tight enough in your game?

7) yes, writing and university, too.

That’s enough for now.

Good point, FireDragon, but we want to skip as many techs as we can.

Announcement

Pre-1000AD landing

Pre-1000AD landing

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment

Comment