SGOTM 13 - One Short Straw

[shyuhe]

Running an engineer while we farm scientists is fine with me, although it puts a kink in our tech route doesn't it? Namely, we'll want metal casting for a forge but we would want to put off machinery in case we get an engineer?? Or would we just burn the engineer on the pyramids or a golden age in that case? Or maybe even MoM?

 

[Dhoomstriker]

^^ I think I got you now, it is more about being able to use everything we got.

Yeah, pretty much. Sorry that I used so many words, but I'm still thinking through the idea and am coming up with new info to add the more that I think about it.

I have no idea what "synthetic" means, but I am guessing that you are asking me to try to be more concise, right? i.e. Say the same thing with less words?

I will try both approaches (MC and machinery from oracle). the philo variant is basically the same than machinery (COL costs less than MC) so I won't go on much with it except to try a pyramids build at the same time (option 4 from LC's post).

Sounds good! Yeah, Lightbulbing Philosophy would simply take less time than Lightbulbing Machinery in terms of teching required, but we'd probably face the same production-based bottleneck.

Representation is nice, but if we're going to get it, we might as well run Caste System and hire Specialists everywhere, right?

Personally, I have been convinced by others on the forums that the best value of early Specialists are the Great People that get generated, not the actual output of the Specialists themselves. This claim has been made regardless of whether or not Representation is in use.

But, if you can find an efficient way to also get The Pyramids, I would be very excited to see your results!

note about pacifism though (you are talking of generating more than 5 GPs in the pacifism scenario): we can't really plan on how many turns we will be able to run the civic. As soon as u need units, the civic starts to be expensive (and u don't whip units while being in pacifism).

That's a very fair criticism. We'll just have to weigh the cost versus the benefits when the time comes.

For example, the question might be: "does paying an extra 10 to 20 Gold per turn by being in Philosophy for an extra 30 turns get countered by having a second Golden Age?" The size of our army (bigger costs more) and the size of our empire (bigger benefits more from a Golden Age) will determine the answer at the time.

Running an engineer while we farm scientists is fine with me, although it puts a kink in our tech route doesn't it? Namely, we'll want metal casting for a forge but we would want to put off machinery in case we get an engineer?? Or would we just burn the engineer on the pyramids or a golden age in that case? Or maybe even MoM?

Well, I was thinking that we'd be putting off Machinery until after getting our 4th Great Person. There are still other techs to research regardless of how we play... Alphabet, Currency, perhaps Construction if we can get Math in trade early on, Compass, etc.

Whatever. It's just an idea. I can see the desire to try for a high-Flask-count tech like Machinery with Oracle because it feels like we'd be saving Flasks. But I think that I showed that the Flask count would actually be equal between the two scenarios, so we wouldn't be saving Flasks at all. This last fact is the most compelling part for me--getting our Great People more efficiently gives us enough additional Flasks to be ahead by 2 techs (Code of Laws and Pacifism) while still having an equal number of Flasks for the other techs that we wnat to research as taking the riskier approach of Oracling Machinery.

 

[babybluepants]

I was playing around with Oracle MC > Alphabet > trade for IW > hook up copper > Colossus. Getting it about 500 BC. Probably could be improved, but let's say that's the approximate date. To make this play, we ideally wanna settle stone as our fourth and copper as fifth (before 1000 BC), to be followed by pig/fish, delaying furs. This is to simplify worker logistics getting to copper. If we're aiming to bulb Astronomy by 1AD (?) that gives us only about 30-ish turns of Colossus. If I assume something in the order of 6-7 working an average of 2 coastal tiles (?) that amounts to somewhere around 350-450c, for the cost of forge + cheap Colossus (Paris hammers that could be used in REX). It also contributes GM points, which is potentially good for bulbing CS, but most likely mixes them with GP points from the Oracle. I'm wondering if it's even worthwhile?

 

[shyuhe]

We'll almost certainly build a forge in Paris regardless of the Colossus. So it's basically a question of the cost of the Colossus (125 hammers) vs. the coins. I think giving up a settler is worth 400ish coins?

 

[LowtherCastle]

Dhoom, how are you hiring merchants without CS? WHere do you build the forges and when? Are you considering the time it takes for those cities to regrow?

 

[LowtherCastle]

Meanwhile here you will find the Excel sheet I was talking about yesterday (credits to Niklas and Renata).

Raz, what sort of a file is this? What is ".7z" for a file type? I can't this to open up.

EDIT: nvm...downloaded 7-zip...

 

[RRRaskolnikov]

Yeah sorry, I should have converted it in .rar or something

To actually use the sheet, u have to copy/paste the tile you use from the first page and fill the tile column with it (in a city page). You enter whips in the relevant column (0 or 1) and substract pop and add hammers in the respective columns. Iirc, there is also a column to enter anarchy turns. Tell me if u have troubles.

@shyuhe: I am not sure about Colossus, we have to take into account what that city would bring during those turns as well.

 

[mdy]

There may not be a very big difference in the time it takes us to tech astro with 2 or 5 GP's. The big difference is that with the additional bulbs we can divert our research into techs which aren't required for astro. In an ideal world we would manually research Civil Service the turn after we bulbed astro and either have finished, or may a good start into researching engineering. If we can go to war immediately after astro with maces/trebs we will be in a much stronger position than if we still have to rely on cats/axes. We would also not have to worry about obsolensence/upgrades. We could put our slider to 0% after astro and put ourselves on an all out war footing which we could not afford to do otherwise as we would need better units at some point. It also means we could afford to whip more heavily, would take fewer losses, and could therefore keep the war going much faster.

Here's an outline of how I think we could get astro, CS, and engineering before 0 A.D.

We build the Oracle in marble city, taking either MC around 1300BC, or machinery just after 1100BC.

The capital whips a library for 3 pops as soon as it is at size 6 and we have writing. It then runs 2 scientists almost continuously and generates our first GS before 1000BC.
If we Oracle MC we will have manually reserached COL by this point so we can immediately bulb philosophy. If we go for machinery we will manually research COL next and then bulb philosophy. Then revolt to pacifism+whichever religion will be most convenient.

Around 1000 BC gold city starts running 2GS which will give us our second GS for bulbing optics 25T later if we Oracle MC. If we Oracle Machinery we would probably need to revolt to CS a few turns earlier than we otherwise would. Note this would give us caravels to look for AI's over the ocean if need be much earlier than we would have otherwise.

Running extra specialists in these cities over this time period is entirely optional allowing the capital to be whipped to support our REX (or maybe to rush some units for an early war if need be).

We then revolt to CS if we haven't already done so. The next 3 GP's will require 450, 600, and 750 GPP respectively. With 5 specialist being run in each of the capital, fish/pig and gold we can generate all these in 25T. In pigs/gold we will run scientists to bulb astro. In the capital we run merchants to bulb CS. Note this would allow us to build the Collossus/pyramids in the capital if we wanted. We would also be able to hire an engineer as an engineering bulb would be almost as good.

In order to get pigs/fish set up as a GP farm in time I think we should settle this city fourth.

Other Techs we need to research (This assumes we take MC with the Oracle, but if we get machinery things would obviously be even better)

Techs there is a 100% chance we would have to research ourselves:

Machinery= 1365
Compass= 780
engineering= 1950
CS=643 (Assuming 1100 beaker GM bulb)

Techs we will need but might or might not be able to trade from the AI

Construction= 683
Calendar=683
ironworking=390

I think there is a 100% chance of being able to trade for any other tech.

So after COL we would need another 4738-6494 beakers after COL. In the approximately 60 turns from discovering COL till 5 BC this implies we would need an average beaker rate of 79-109 beakers a turn.

In my 1050BC save we were running at around 50 beakers/turn at the break even point. Running the extra scientists for 25T would add an average of around 23 beakers a turn by itself, so while an average of 109 beakers/turn may be optimistic we should easily be able to do better than 79 beakers/turn. With a little help from the AI it should be doable, and Oracling machinery would obviously make this easier.

The pyramids would come too late to help much/at all with the research, but it may be worth building for the +3 happies to increase whip potential, or maybe for the hammer bonus from police state.

 

[LowtherCastle]

My main concern with Dhoom's and mdy's GP plans is having the population to run these specialists. I would like to see someone play out one of mdy's machinery test saves to show that this works by 1AD. Normally, GP spamming works with mature empires, in contrast to growing/REXing empires.

My gut feeling is that using the same machinery 1050BC save, the Astro/2GS beeline would be far ahead in REX and Astro date.

 

[LowtherCastle]

To actually use the sheet, u have to copy/paste the tile you use from the first page and fill the tile column with it (in a city page). You enter whips in the relevant column (0 or 1) and substract pop and add hammers in the respective columns. Iirc, there is also a column to enter anarchy turns. Tell me if u have troubles.

I don't have excel. I use OpenOffice Calc these days and the macros don't work. Alas...

 

[mdy]

I've been testing an alternative to T82 that I'm interested in feedback on, in particular from mdy, if it improves your results. After the galley but while Paris is still at pop4, switch 1 clam to the mine for 1t. This delays pop6 by one turn, but gets the lh done in advance.

Since Marble gets the fish nets slowly anyway, I don't see a problem on this.

Another possible fine-tuning on the mdy-machinery variant, would be to chop the northern forest into the lh before quarrying the marble, since the worker sleeps on the marble anyway, waiting for Masonry.

I get a 1075BC machinery sling with this approach (though i think this was mainly due to me actually remembering to do binary research and to put the slider to 0 whilst the library is built in the capital this time). I don't think it would improve our Oracle date, but it would leave the gold city in a stronger position, so I think it would be a good move if we decide to oracle machinery.

My main concern with Dhoom's and mdy's GP plans is having the population to run these specialists. I would like to see someone play out one of mdy's machinery test saves to show that this works by 1AD. Normally, GP spamming works with mature empires, in contrast to growing/REXing empires.

Both gold and the capital have three food. Getting these cities big enough will present no problems whatsoever. The only one we need to be careful with is pigs/fish, which is why I favour founding it fourth to get it up to speed on time.

My gut feeling is that using the same machinery 1050BC save, the Astro/2GS beeline would be far ahead in REX and Astro date.

The 2GS plan would be substantially ahead in terms of rex, but behind in tech. We can get 5 GP's in the time in takes to get 2 without CS/pacifism-depending on when we chose to revolt to CS I expect the 5GP plan would actually get astro earlier.


In strikes me that one really important test has not been done yet: What dates do the AI's tend to build the Oracle when they are as unusually ******** as they appear to be in this game.

 

[LowtherCastle]

The 2GS plan would be substantially ahead in terms of rex, but behind in tech. We can get 5 GP's in the time in takes to get 2 without CS/pacifism-depending on when we chose to revolt to CS I expect the 5GP plan would actually get astro earlier.

I ran your 1050 save forward, focusing on REX. I stopped around 400BC when an unusually ******** barb galley killed our galley at 1.6/2 HP and pillaged the Pigs fish in the same turn. Our REX was also unusually ********, imo, for ramping up any sort of war machine by 1 AD. This makes me think the only viable alternative is MC slingshot with max REX from the get-go. Otherwise, we'll be sorry. If our empire is not well established by 1 AD, this game will take forever.

 

[LowtherCastle]

Both gold and the capital have three food. Getting these cities big enough will present no problems whatsoever.

Well, I'd like to see your save, because I"m not finding that to be the case. Both cities need a forge and they're the primary ones for building settlers. Combining that with running 5 sci so soon doesn't work in my tests.

 

[Dhoomstriker]

Dhoom, how are you hiring merchants without CS? WHere do you build the forges and when? Are you considering the time it takes for those cities to regrow?

A Market from Currency... we did say that we wanted to research Currency. I'm considering the fact that we'd need to regrow, which is why I say that it's slower than the non-Pacifism approach. Probably only one City would need a Market, which could be a City that generates, say, Great Person #2, so that it has time to grow and prepare for whipping and running so many Specialists when generating the optional Great Person #5.

Ideally, the idea will be put to a test run. However I am still torn on other ideas that come before then... such as...
If we're going with an approach that Oracles Metal Casting, we might as well try and get the Oracle as soon as possible, right? If so, then does it make sense to do soemthing like:
a) Settle on the Marble and build a Monument
AND
b) Build The Oracle in the capital
AND
c) Build 2-Clam in the east or Pig + Magical Fish and research Animal Husbandry before Fur, so as to focus on starting 3 Great Person Farms (counting Marble) that don't use the capital's Oracle
AND
d) Keep spamming Settlers while using whip overflow into the Oracle, then spam some more so that we can pick up Fur and possibly Stone or Copper soon thereafter

versus

What you guys seem to be doing by not settling on the Marble and manually improving it by researching Masonry early
But perhaps combined with building the Oracle in the capital and then very quickly getting up 2-Clams and Pig + Magical Fish, since the Marble location will no longer be a suitable Great Person Farm

versus

What you guys seem to be doing by trying to build the Oracle in Marble City.
The concern is that if we're building The Oracle at a late date due to a production bottleneck, then will people complain that we took "such a cheap tech as Metal Casting" with the Oracle?


Then I start wondering if I really need 3 early Great People Farms at all, and that if I don't, is it better to try and dual-use Marble City for both Marble and a Great Person Farm, so that we can settle Fur sooner, again taking into the fact that if we only need 2 Great People Farms for a while, then the capital won't have to be one and can keep pumping out Settlers while first overflowing Hammers into the Oracle and regrowing while building the Oracle.


So, basically, I feel a bit stuck on which path to try out because all of these possibiilities happen before we'd even start using Pacifism, and I am hesitant to run a Pacifism test run if it turns out that I picked a really inferior City-location plus Oracle-location choice (given that Metal Casting is immediately Oraclable and thus building the Oracle ASAP is probably, but not necessarily, a good thing to do).

Still, I will try to make some time soon to play out some of these ideas.

 

[Dhoomstriker]

Raz, what sort of a file is this? What is ".7z" for a file type? I can't this to open up.

EDIT: nvm...downloaded 7-zip...

Would anyone care to re-upload a .zipped version of the file to this thread?

EDIT: Nevermind, I'll do it myself and will upload a copy attached to a new message (because I haven't been spamming this thread enough already! ).

 

[Dhoomstriker]

My main concern with Dhoom's and mdy's GP plans is having the population to run these specialists. I would like to see someone play out one of mdy's machinery test saves to show that this works by 1AD. Normally, GP spamming works with mature empires, in contrast to growing/REXing empires.

Well, his idea of using Caste System for most of the Great People does carry the advantages that:
a) He will not need to whip any buildings besies a Granary plus some optional buildings like a Lighthouse or a Monument, allowing him to skip building Libraries
AND
b) With most of our Cities hiring 3 to 5 Specialists, Slavery won't be of much use as a Civic during that time period anyway

A third, undiscussed, but strong idea that builds on his approach is:
What if we use Great Person #2 or Great Person #1 on a Golden Age?

Sure, LC has mentioned using a Golden Age to help us cover the costs of a quickly-expanding empire, but if we aim to raze rather than capture many Cities, we might not have a major issue with negative cashflow, particularly if we're turning off research after Civil Service and/or Engineering.

Do we really need Engineering for extra land movement if we have Galleons for movement, anyway? Okay, okay, Trebs are awesome, but we can make that decision later.

However, one key benefit of a Golden Age is that it equal's Pacificism's bonus on Great Person generation.

In fact, if we were smart about it, we could save the 2 turns of Anarchy from switching in and out of Caste System that his idea would require, particularly assuming that he'd plan to revolt into Pacifism earlier than he would Caste System.


So, we could still optionally spend 2 turns of Anarchy on getting into Pacifism and our State Religion, in order to generate Great Person #2 for our Golden Age. We would then save 2 turns of Anarchy from switching into Caste System and out of Caste System. During the Golden Age, we would aim to work on generating at least 3, if not 4 Great People, using Caste System. When the Golden Age is about to end, we'd switch out of Caste System and would just use Pacifism and 2 Scientists per City (plus an optional Engineer in one City... we probably wouldn't have the population to have whipped a Market anyway) to finish off the Great People.


If we wanted to get REALLY fancy, then we'd Oracle Philosophy. Suddenly, we would be able to shift the 2 turns of Anarchy from switching into our State Religion and switching into Pacifism into our first Golden Age, since we'd use our first Great Person on a Golden Age instead of on Lightbulbing Philosophy.

Prior to the Golden Age, we would spam our Great Person Farms, build minimal infrastructure to help them grow as soon as possible (Granary, Lighthouse, possibly a Monument, probably a Warrior for Happiness purposes, and that's probably it). When we've grown our 3 to 4 Great Person Farms, we'll launch our Golden Age, switch into our State Religion, switch into Caste System and Pacifism, and save 4 turns' worth of Anarchy.

We'd still plan to only switch out of Caste System at the end of the Golden Age, while continuing to run Pacifism and 2 Scientists (plus an optional Engineer if we can whip a Forge in time... which we probably won't be able to do in this scenario since we didn't get Metal Casting early on, so probably just 2 Scientists) per City that still needs to complete its Great Person.


I haven't worked out the math yet for how many GPP we'll make during the Golden Age period versus how many turns would be remaining to generate our Great People after the Golden Age is over, but at least we would be leveraging some of the strong uses of a Golden Age:
a) Extra GPP
AND
b) Anarchy-free Civic switches


So, yeah, now we have even more ideas to test!

 

[Dhoomstriker]

Attached are copies of the spreadsheet that RRRaskolnikov wanted us to look at.

I used a file conversion website, http://www.convertfiles.com/, to switch from .7z to .zip format, then used that website again to switch from .xls to .ods format (for Open Office).

I have provided a .zipped file for both the .xls version of the doecument and the .ods version of the document.

I don't have Open Office, so I don't know how well the .ods version will work out--let us know!

 

[Dhoomstriker]

allowing him to skip building Libraries

Well, we'd probably still need a couple of Libraries, depending upon which Cities would generate a Great Person outside of Caste System.

Alternatively, we'd run Caste System beyond the end of the Golden Age and would just have to lose 1 turn of Anarchy switching into Slavery later.

With Pacifism and a Golden Age, each Specialist generates 9 GPP per turn. A Golden Age on Epic Game Speed lasts 10 turns. That means that each Specialist will be worth 90 GPP during that time period.

90 GPP * 5 Scientist Specialists = 450 GPP, which would allow us to spawn Great Person #3 in a City that focused on growing quickly without building a Library... but we'd need to stay in Caste System on the last turn of the Golden Age, meaning that we wouldn't get an Anarchy-free switch into Slavery. That fact may be a good thing, anyway, since if you were to switch out of Caste System on the last turn of the Golden Age, then you would lose the Golden Age's last turn worth of extra GPP on Specialists.

A different City, such as Gold City, could feasibly just "continue" completing Great Person #2 (since it had already hired Scientists in its Library prior to the Golden Age), then gain a good head start on Great Person #4 or Great Person #5, for example.


So, yeah, with the Golden Age, we should probably plan to still spend 1 turn of Anarchy at a later date for switching from Caste System back to Slavery. But, we'd at least save 3 turns of Anarchy and would get extra GPP during the Golden Age... depending upon how many Specialists we'll be able to hire during this time period, it may be worth doing.

 

[LowtherCastle]

My summary conclusions

1. Writing comes next: 1) OBs for our exploring wbs; 2) trade route cash; 3) research into overdrive; 4) start building GSes; 5) GS for Philo possibility. Bonus: 6) the Ducks didn't do it... (They're going all-out REX. Maybe going for Stonehenge, though I doubt it.)
2. Our main priority is city development/REX. When to settle which city depends on: 1) wkr/wbs available; 2) prereq techs; 3) our long-term needs; 4) whether the city needs 15 asap; and 5) building the Oracle.

Based on those factors, here's my current best logic, in sequence:
1. Settle Marble. Gives the wkr time to chop the forest into the Gold granary. Frees Paris to build other stuff. Allows us to beeline Priesthood.
2. Settle Stone. No prereq techs. Needs a monument asap. Can work the Paris mine. No big hurry for the worker. Pop6+forge = 20hpt (@-1fpt). ***My testing shows that around 1000BC we'll start needing some city to crank out wbs and other hammer-heavy stuff.***

-------------------------------------
Techwise, all slingshots are still open.
-------------------------------------

If we don't go for Machinery:
3. Furs. Requires a dedicated worker. Needs an immediate monument. Hunting is next in the research sequence. Gives our research a great boost. Gives us +1 and +1, empire-wide.
4. Pigs. AH is next. Needs no monument. Expands rapidly. +1
5. ...

-----------------

Note: Don't forget that we still have three unfogged tiles that may change all our plans noticeably, especially the Marble tile.

 

[babybluepants]

Oracle: I'm for MC asap, not least because I don't trust dates closer to 1000 BC or believe that we can reliably test for the safe date (Oracle date is notoriously unreliable - I'm pretty sure I've seen it built even before SH/GW).

GPP: the tests I've run to 500 BC felt very slow, even with a bit of a focus on expansion and city development. While that plan sounds excellent, I find it very hard to believe that we can run off 5 GP under Caste and have anything resembling a military production machine by 1 AD. If you guys can prove it in a test game - great. I've been wrong many times before. In the meantime, I'll focus my testing on MC followed by Currency and further expansion. Might be useful to run some variants closer to 1 AD, so we have a real sense of the state our empire could be in by then.

Furs: problem is the dedicated worker, I think. That's probably fine, but if we want Colossus, we need a couple of workers down in the SE (which makes both stone and pig/fish more desirable). I think we need to decide on Colossus almost right away - it's a commitment and only makes sense to really beeline, since we're obsoleting it quickly.