SGOTM 14 - Plastic Ducks

[Bebekija]

My understanding was that getting Silk city works best for the gpp plan. If it is not required (i.e. can you finish GPP prior to or on T104?) then let's stop breaking our heads open and plan something a lot easier.

It depends : 7 or 8 GPs ? Cause 7 I can do in ~12 turns, but 8 I doubt it.
The thing is the point of forgetting Silk is to focus the worker turns on Wash instead, making it a much stronger GP farm, able to get out 2-(3?) GPs.
But the micro around Wash is precisely where I fail the most (farms + chopping the Mids), so I'm definitely not getting the most out of this option.

 

[kossin]

Just 7 GPs.

1-Math (Wash)
2-Philo (Wash, provided we're growing NY)
3-GA (Boston)
4&5-Astro (NY, ?Wash?)
6-Machinery/Optics (?Canal?)
7-Engineering (?Silver?)

The timeline for 8 is too slow, we can generate the extra research the GS will bulb just by playing those extra turns. All he'd do is allow us to research side techs.

The one thing I am unsure of though, is even if we beat my T104 attempt (by 1t, 2t if we get some lucky forest spreads, confu spreads and trades), we'd have fewer developed cities to start the total war. Which is kinda pointless, 3 turns might be worthy but 1t no (you're either sending a SoD that's weaker or delaying the start...).

 

[Bebekija]

Just 7 GPs.

1-Math (Wash)
2-Philo (Wash, provided we're growing NY)
3-GA (Boston)
4&5-Astro (NY, ?Wash?)
6-Machinery/Optics (?Canal?)
7-Engineering (?Silver?)

So yes, 7 can be done without Silk.
In my last test, what I did was Wash/NY/Wash/Boston/Silver/Canal/Wash, with Wash starting at Pop 10 and Canal/Silk at Pop 8 (Silver didn't have the religion). It took 12 turns, which could of course be improved if Wash started stronger and Silver had the religion.

But...

The one thing I am unsure of though, is even if we beat my T104 attempt (by 1t, 2t if we get some lucky forest spreads, confu spreads and trades), we'd have fewer developed cities to start the total war. Which is kinda pointless, 3 turns might be worthy but 1t no (you're either sending a SoD that's weaker or delaying the start...).

I completely agree with that.
That's another reason why you didn't hear me push the "No Silk" option forward, because I know that having this extra city ready when we start to build an army will prove infinitely more useful than 1-2 turns earlier Astro (it won't be 3 turns earlier, unless I did something very wrong in my test, and I don't think I did)

So I believe that settling the Silk will prove to be the better choice, it's just that I can't manage to make it work on my own, so I kept trying something else.

 

[Deity player]

Fastest settler in Wash: mine the non-riverside PH with 2 workers

Fastest settler in Boston: start T67 at size 4 with 1 chop

Fastest settler in NY: start after worker with 1 chop

These advices really helped me. I am not very experienced and it is difficult to balance things for me. For example, I apparantly planned to chop too much in the PPP. I can only catch up when I know what to do and try to find the best way to acomplish that. So thanks for the nice summary.

This is a real headache, as we also have to balance out getting workers where they need to be for optimal city growth when they are settled (btw, I haven't checked this in a long time).

Yes, headache is the right word .

Can we get a few T75 saves trying out the different approaches? They need not be perfect of course, just the general positioning of workers and workboats as well as improvements.

I have played until T75 right now following your reccomendations and I finally feel that I did well enough . I will try again and post the save in several hours.

 

[Deity player]

So here is my T75 test. I followed the fastest ways to construct the settlers as kossin pointed out. Note that I played very differently from my PPP (and almost surely better).

 

[kossin]

Good! All we need are a few different approaches to make up our mind.

 

[kossin]

I like the micro around Canal and Silk. However, Silver needs to be 1N for optimal growth... I know it sucks we're losing a lot of worker turns and 1C/t until Sailing but it's worth about 30H more which converts in quite a bit of food.

I'm still getting T104 as best result. Maybe we can find a way to accumulate some earlier GPP instead of growing the cities large and that might push it to T103. It'd also mean we have better chances of getting scientists than Artists.

 

[kossin]

Something pretty similar to Deity's T75. Main differences being the Boston wb is out 1t earlier (=2F2C in Silver) and Silver settled 1N which costs a few worker turns, but the city develops faster.

 

[Bebekija]

I'e been doing some more tests on optimal growth of the GP farms.
Or rather, on sub-optimal growth...

Given that, with optimal growth, it takes 16 turns for Silver city to reach Pop 8, and 16 turns as well for Canal city (15 turns if we can manage to have IW and Calendar by T85), I looked for the worst scenarios that still allowed to reach this population in 16 turns.

I came up with this :

- Silver city (WB late): @T0, 2 workers ready on the forested PH 1W of the city, nothing prechopped, the fish is improved @T0+9. The city will reach Pop 8 (19/36) on T0+16.

- Silver city (workers late): @T0, 2 workers on the DH 1S of the city with no movement left, nothing prechopped, the fish is improved @T0+5. The city will reach Pop 8 (18/36) on T0+16.

- Canal city : @T0, 1 worker on the sheep and one worker on one of the hill forests (doesn't matter which one), first seafood improved @T0+10, the other one improved @T0+11. The city will reach Pop 8 (18/36) on T0+16
The worker turns are more critical in Canal city since the workboats are bound to be late anyway, so I didn't explore the "worker late" option.

Still not tested Silk....

Hopefully this will provide some guidelines when trying to optimize the worker turns/WB building


Also, I started a full test focusing on Silver and Canal, settling Silk only as an afterthought, to try and provide a different save from yours. But I have no time to work on it tonight, so don't expect the final save before this we.


Finally, note for later : I believe the turnset should end just before the cities are settled. The reason being that once we reach a decision on our plan up until Astro, we are probably going to fine tune seriously the GPP plan, and the order in which the cities are settled may matter.

 

[kossin]

Yes for set length.

That's why we need a plan until the completion of the 3 settlers. We need 6 workers in good locations and there are some things that must absolutely be done by then.

For example:
-mine vs chop in Washington
-2 chops in Boston is a must if we want the settler out early at size 5
-1 chop in Boston if we start settler at size 4
-road to Canal

There aren't too many ways to do this quick... so I suggest you try playing some tests until T91-94 and compare the end result. You will need to get the Philosophy GS and start working on the GP from Boston.

Thanks for the city micro. This means we should have a few extra worker moves (but not much). This is something we need to determine before finishing this set so that we can figure out where the workers need to end up at.

 

[Duckweed]

^ What's your test results and conclusions?

From Deity player's and your test saves, it seems that you are growing NY. What's your thought, 2nd GS? or GP farms later?

The luck factor is going to play an important role in the plan.

1. How's the free religion spreading? My guess it's a low odd that all designated GP farms get free religion. Therefore basically our plan should consider the case of producing at least 1 missionary or producing the last GS without religion.

2. The odd of GA. Over 10% chance of at least 1 GA is not negligible. How do we handle this situation? That's the reason why I mentioned 7~8 GPs, we need a backup plan for suboptimal result.

 

[kossin]

^^

T104 is my result for 7-GP. I haven't played further to 8-GP because it would be a very different path (i.e. Washington needs to pop 2 GPs).

NY: we can use it to pop the 2nd GS on T91, this means more food+hammers in Washington, and also more turns of scientist specialist which means better odds of a GS than a GE.

Luck is indeed very intense in this plan:

1-some games I get all spreads without even using the free missionary
Other games I get 1 or 2 spreads only.
2-This is something we can try to deal with by starting to run specialists earlier... less GPP but more scientists = better odds.

Best case scenario
We get 4 GS for bulbs (Machinery/Optics, Astro*2, Engineering).
We are able to trade for Alphabet, Sailing, Pottery, IW, Calendar, Aesthetics, Construction.
Free religion spreads.
All done on T103 (I don't know if it's possible, but I guess yes.)

Typical scenario
Trade for Alphabet, Iron Working
Religion spread in Washington, NY and 1 of the 3 GP farms => 1 missionary required
We get 1 GA instead of a GS... skip Engineering bulb (it sucks, but that's the better option imo).
All done on T104

Worst case scenario
No religion spreads. Whatsoever. (I think it's about 10% per city when they are close (function of distance) each turn, so this would be sooooo bad luck).
Can't trade for Alphabet... so no techs at all.
We get 3 GAs and 1 GE.

If this happens, we're more or less screwed

Backup plans
1-We can use 1GA+GE to trigger a new Golden Age and try to farm 1~2 extra GS to bulb Astro.
2-Depending on religion spread by T91, we grow cities further and go for the 8-gp approach instead.
3-We abandon the total war option (ugh..)

 

[Deity player]

Deas anyone plan to post more alternative saves? (I tried to do some but the results were very bad.) If not, I will write a new PPP tonight or tomorrow (trying to follow kossin's test).

 

[Bebekija]

I want to try something tonight, but don't mind me and write your PPP anyway.

While everyone is reviewing the plan, I will try to finish this test, and report if I've come up with something good before you get the final green light from everyone.

 

[Bebekija]

Backup plans
[...]
3-We abandon the total war option (ugh..)

If we're screwed with the total war option, what would be the best way to recover?
Would it still be possible to switch to the diplomatic option or are we already too far on the domination path? And with team diplo in action, is it even possible to achieve a diplo win? (I'm not talking about being competitive here, because if we're in that much trouble, we can assume that any laurels are out of the question)

 

[kossin]

Bulbing Math and CoL from Oracle pretty much forces us to go total war. We should focus on developing cities and try to adapt depending on religion spread. The 8-gp approach is safe enough to not have to drop total war.

EDIT: we can try a few alternatives for getting the settlers out, but what I proposed (as seen on Deity's and my T75 test) is the best I could come up with.
Alternatives:

Washington: worker->settler via chops
Problem: silk is settled later than other 2 GP farms

Boston: settler @5 via 2 chops
Problem: again settled late, 3 more worker turns required in Boston instead of somewhere else

NY: I don't think anything else even remotely makes sense

 

[Duckweed]

^^

T104 is my result for 7-GP. I haven't played further to 8-GP because it would be a very different path (i.e. Washington needs to pop 2 GPs).

NY: we can use it to pop the 2nd GS on T91, this means more food+hammers in Washington, and also more turns of scientist specialist which means better odds of a GS than a GE.

Luck is indeed very intense in this plan:

1-some games I get all spreads without even using the free missionary
Other games I get 1 or 2 spreads only.
2-This is something we can try to deal with by starting to run specialists earlier... less GPP but more scientists = better odds.

Best case scenario
We get 4 GS for bulbs (Machinery/Optics, Astro*2, Engineering).
We are able to trade for Alphabet, Sailing, Pottery, IW, Calendar, Aesthetics, Construction.
Free religion spreads.
All done on T103 (I don't know if it's possible, but I guess yes.)

Typical scenario
Trade for Alphabet, Iron Working
Religion spread in Washington, NY and 1 of the 3 GP farms => 1 missionary required
We get 1 GA instead of a GS... skip Engineering bulb (it sucks, but that's the better option imo).
All done on T104

Worst case scenario
No religion spreads. Whatsoever. (I think it's about 10% per city when they are close (function of distance) each turn, so this would be sooooo bad luck).
Can't trade for Alphabet... so no techs at all.
We get 3 GAs and 1 GE.

If this happens, we're more or less screwed

Backup plans
1-We can use 1GA+GE to trigger a new Golden Age and try to farm 1~2 extra GS to bulb Astro.
2-Depending on religion spread by T91, we grow cities further and go for the 8-gp approach instead.
3-We abandon the total war option (ugh..)

Not peculiarly the answers I expected. Since you have run most tests in different situations and known the results and answers to your own questions in post#710, I'd like to see your reasons behind your choices.

Regarding the religion

From my previous tests, usually Washington could get free spread. The odd of 1 more free spread in 1 of the 3 new cities is quite high. However 2 out of 3 is not optimistic. I'd like to see the difference of producing missionaries (100 H) and producing the last GP without religious bonus.

Regarding unexpected GArtist

The chance of 2 ~3 GAs is very low (<1%) and hence can be neglected. However the odd of 1 GA is about 1-0.94*0.94*0.94, not very high, but surely it can happens. The question is when do we start the golden age (for safe consideration, it better to be late, especially if Boston pop a GS). The case of producing 8 GPs vs manually researching Optics, which one is better?

Regarding worker micro

Canal city -- In my previous tests, I spent 4 worker turns on chopping the forest of city center tile (8H to NY) and chop the south 2 forests 1st. The 2 road on the hill tiles in your test does not look helpful to me.

Silk city -- How about improving Pig tile before chopping forests?

 

[kossin]

Not peculiarly the answers I expected. Since you have run most tests in different situations and known the results and answers to your own questions in post#710, I'd like to see your reasons behind your choices.

Regarding the religion

From my previous tests, usually Washington could get free spread. The odd of 1 more free spread in 1 of the 3 new cities is quite high. However 2 out of 3 is not optimistic. I'd like to see the difference of producing missionaries (100 H) and producing the last GP without religious bonus.

Regarding unexpected GArtist

The chance of 2 ~3 GAs is very low (<1%) and hence can be neglected. However the odd of 1 GA is about 1-0.94*0.94*0.94, not very high, but surely it can happens. The question is when do we start the golden age (for safe consideration, it better to be late, especially if Boston pop a GS). The case of producing 8 GPs vs manually researching Optics, which one is better?

Regarding worker micro

Canal city -- In my previous tests, I spent 4 worker turns on chopping the forest of city center tile (8H to NY) and chop the south 2 forests 1st. The 2 road on the hill tiles in your test does not look helpful to me.

Silk city -- How about improving Pig tile before chopping forests?

I didn't bother finding out optimal city growth, so I used Bebekija's guidelines. I'll run a quick test trying to do something else later for silk city and compare. Obviously, running the artist at size 2 might be preferable but might yield fewer food overall as well.

I chopped the Canal forest a few times also... the problem is 8H is 2H/t for the job, it's not very efficient. Compare to the roads on hills:
-1 more turn of roads than through wheat+sugar
+2 turns saved when sending workers there to chop
+1~2 turns saved when going to improve the sheep (basically, 1 worker would be 2 tiles away with no road movement in between)
---
-2 turns later on sugar [this is depending on when we get IW]
-2 turns later to make a road on wheat [we will never need the extra health so this is a moot point]
+saved turns if we ever wish to mine the hills [doubtful, but a grass hill is good]
+++a lot more food from faster city development
The earlier worker turns are worth more than later worker turns, otherwise it's about worker turns neutral, just faster in the beginning.

Religion
Not having the religion in 1 city, the worst case scenario would be the 7th gp, it'd cost 18 gpp/t, or about 180gpp total. Generating 180gpp with 6 specialists will take 180/(6*6) = 5 turns delay
It adds difficulty with the triggering of the GA so that we don't waste 1~2 turns of Anarchy after the GP-spree. If this were the case, best to revolt into Taoism and use that religion like Bebekija suggested.
2 options for missionaries: Washington after Pyramids with forests AND/OR Boston instead of granary+LH

Unexpected GA
Another alternative: skip the Golden Age (needs 4 extra turns of specialists = better odds, 1 turn of revolting... 5 turns delay). '7th GP' can be backup or used to bulb Paper/something useful.

8-gp plan (I can't recall exactly, but it shouldn't be too far off) might add 5~8 turns.

Hand-researching Optics (stable slider) should be 4~5 turns.

EDIT: after a quick look at previous tests, these numbers seem accurate. It's much safer to assume we'd have all research done with a 8-gp approach, whereas on 7-gp we're relying on the AIs for Alphabet, Calendar and 400~500 beakers more.

 

[Duckweed]

Fair enough about worker micro on canal city.

Since I saw a road on the pig tile, so I thought there might be some room to get the pig improved as soon as possible. Granary can be delayed a little bit.

I saw Bebekija posted the result of producing the last 1 or 2 GPs without religion somewhere, IIRC it was 2 or 3 turns slower.

The balance between fastest GP production and safer way is tricky. IMO skipping GA is not an option since all we are worrying about is the unexpected GP, which can only be used for GA. Anyway, let's focus on setting up farms now.

BTW, x-team seems to succeed in getting Oracle.

 

[kossin]

1road in silk is done while the city is not settled and completed once the city is settled to allow the 2nd worker to get there without wasting a turn. Alternative is to pre-chop another forest but then we lose 2 worker turns = road (and further movement when chopping the other silk). Perhaps improving the pig first yields a better result, I did not check yet.

Delayed GP 2-3 turns is probably by using Taoism, which is a good alternative for a few turns.

I don't like skipping the GA either but I raised the issue because it's an option... bad as it is

X-Team indeed, it looks like the map isn't really punishing which is bad, we need the AIs to trade us techs