SGOTM 21 - Plastic Ducks

[Doshin]

(1) Warrior/City 3

Capital builds: Settler ---> Worker ---> Warrior (T30)

The Warrior can move:

Spoiler :

^^^^^ the Warrior is built in time to fogbust the Gems in advance of the Workers.

T35: the Warrior reveals new tiles.
T36: the Workers are in place to leave our borders. (Warrior reveals some more new tiles).
T37: both Workers outside our borders.​

I slow build the second Settler in this test. We could whip and still be safe, as long as the Warrior lives.

The Warrior gives a few turns of new map information before we settle, but not enough time to effectively change the Worker micro (or change tech path).


(2) Scout/City 2

Planned stopping point 2SE, on T11:

Spoiler :

Settler built on T22. Settler and Worker move out of borders. Scout moves:

Spoiler :

T23, Scout moves:

Spoiler :

(3) Scout Before City 2

We'd have 5 turns to go in one direction, and 5 turns to head back to fogbust.

Does this appeal to anyone? At this stage, I would prefer to simply fogbust.

NE would be the obvious direction, if we do want more land info, but we just moved away from the Lion in this area.

(4) Scout After City 2

Gems region:

Spoiler :

This will arrive before the Warrior, and with 5+ turns to explore.

However, barbarians will have spawned for 10+ turns by the time it arrives.

West/South West:

Spoiler :

------

^^^^^^^ This is mostly for reference, but let me know if you have any thoughts (especially about point #3: Scout movement before City #2 is founded).

 

[Duckweed]

As I suggested before, you could let scout go through the north of the wheat if we want 100% safe (move 1 tile a tile after 2N of the wheat) and let it back to guard the Cow site before T23.
Just a suggestion, I have no strong objection to stay.

After T23, you could also let scout lead the way before warrior to be safe. This is strongly recommended.

 

[Doshin]

Ok. I'll try to explore by the Wheat, if the Scout is healthy and there is time. Regardless, the main priority is to fogbust for the Settler party.

Playing now...

 

[Doshin]

Wait, not playing. The file crashed.

Message:

"Load Failed"

"The save file you have selected is protected to ensure that the assets in your mod folder have not been changed."

-----

I'll see if I can fix it.

 

[kossin]

Tried Assets Checker?

 

[Yamps]

It's hard, but looks like doable. We probably don't need to save the forest on Gem. Pre whip something should give us 30H.

Capital producing 2 GPs with pollution is too much. I could accept 1 early non GS GP, which can be used for a golden age, but the 2nd unlucky GP could cost us more than 5 turns, which is more than the loss of Oracle.

Busy with work, but I've been reading along. I'm interested about this less than 5 turns estimate. Are you're saying that it would cost us less than 5 turns if we don't build the Oracle at all?

It also sounds like there would be a lot of effort required to build it in our third city, Kossin mentioned 25 worker turns. GPP pollution is a small percentage risk, but how much do we lose for sure in development if we don't build it in the capital?

 

[Doshin]

It took 8 hours, but I've now opened the real save.

Steam BTS is truly awful. There are sooooooooo many problems and errors with mods, and BUFFY in particular.

I'll play tomorrow, in the early afternoon.

 

[Yamps]

@Doshin good to hear you've managed to sort it out!

Regarding moving the scout after city 2: barb warriors could pillage border improvements when the average number of cities gets larger or equal to 2. To protect our first gem, I'd send him like you suggest in the first screenshot and fortify him for fog busting until our warrior arrives.

 

[BSPollux]

Eight hours, what a pain

 

[Duckweed]

Busy with work, but I've been reading along. I'm interested about this less than 5 turns estimate. Are you're saying that it would cost us less than 5 turns if we don't build the Oracle at all?

It also sounds like there would be a lot of effort required to build it in our third city, Kossin mentioned 25 worker turns. GPP pollution is a small percentage risk, but how much do we lose for sure in development if we don't build it in the capital?

It's just a roughly estimation. Because the tasks, currency becomes a necessary tech to learn early in this game. Therefore the gain from Oracle is the beaker cost of Currency ~600B plus other economical gain for about 20 turns. Let's assume 1000B totally. The cost of building Oracle is 150H(probably = 300B) plus the techs and worker turns, let's assume another 300B. Therefore the net gain from Oracle is around 400B. If we don't have mids, then the research rate in the final period to Astronomy would be around 150bpt, which needs 3 turns to output 400B.

Of course, the real situation is more complicate, the most important thing is whether the GS or the research rate is the bottleneck. With Phi trait, mostly the bottleneck is the research rate, that's why I insist on bulbing Math and Oracling Currency instead of building GLH.

It's also hard to estimate the chance of getting Oracle. Even in Monarch, AI could build Oracle before 1800BC, or in deity, AI might build it later than T65. Generally, T60~65 have ~50% chance of getting it, mainly depends on whether AIs who is likely to research PH have the marble. However, one thing for sure is that "If you don't try it, you'll never get it". The thing is that we should control the process to minimize the cost in case that we lose. For instance, chop the forests in the last minute and complete it in the shortest time to avoid wasting hammers.

 

[Yamps]

^ Hmm, these decisions are hard to make. If it really only saves us 3 turns in the end and if we have 50% chance to lose it, then couldn't we make up for this 3 turns early on by playing differently? For example, we could just expand and then build research along with 100% slider if get enough cash from GM trade routes.

 

[Duckweed]

3 turns are a lot, imagine how much you could save with those micros. There were only few things that could give you 400B, which is a strategical decision, not technical. To emphasize, 3 turns are net gain, which already count the cost, such as the important 150H early hammers.

Edit: Let's take some time to review the cost and value of a GM.

1st, we can only produce a GM after we have CoL, with Oracle, T60, without Oracle T70+. After CoL, if we want a GM early, we need to switch to CS, which greatly hampers the setup of GP farms, especially the island cities have to rely on whipping to produce granary and LH. If we produce the GM until all GP farms are mature, it's worse than a GS, since a GS bulbs Optics instantly, while 100% research slider still needs 4~5 turns to research Optics manually. You also need to count the time needed to travel for trade mission, and usually the next 3 GPs takes less than 10 turns, which means not many turns can be used to take the benefit of 100% slider. Moreover, you have to count the cost of not bulbing Math and not Oracling Currency. Therefore producing the 1st GP as a GM, no matter with any way, is way worse than a GS to bulb Math.

If you still have questions, I'd suggest you to take a look at our playing of SGOTM14 (the commercial situation is more similar as this game with Phi trait, so we produce 7~8 GPs) and SGOTM17(commercial situation is much worse and without Phi trait, so we only produce 4 GSs). Both games had mids, but this game probably not.

 

[Doshin]

Started playing.

 

[Doshin]

@ Doshin re:scouting.
I'd be happy if we fortified the scout 2SE.
IDK what there's to find north of the gems that would affect our immediate settling plan.

I like Doshin's plan to get to Turn 30, I just want us to be a bit more aggressive with Scout.

Maybe Turn 24 after our 2nd city is settled.
Or maybe the Wolf attacks right away, the Scout takes no damage, and it can explore up the east coast 1 tile at a time briefly.

As I suggested before, you could let scout go through the north of the wheat if we want 100% safe (move 1 tile a tile after 2N of the wheat) and let it back to guard the Cow site before T23.
Just a suggestion, I have no strong objection to stay.

No animals yet on T12. Current (optimal/no animal encounters) scouting plan:

Spoiler :

The Scout could be back on the Gems on T21.

I'll turn back if the Scout is attacked, or progress is blocked by the Lion.

 

[Doshin]

Played to T24 and the second city:

Spoiler :

The Scout is on the Gems, as you can see, and is yet to move.

Am I ok to move the Scout 1NE towards the third city, Gems area? I would like to fogbust before real barbs appear, if possible.

No animal encounters during these turns.

The scouted land isn't terribly exciting. What we thought was a second island is probably part of our continent:

Spoiler :

I can post more thorough information later, but I'm meeting someone for dinner in 20 minutes.

If anyone comes online, please let me know if I'm (not) ok to move the Scout NE!

Also, I will post demographics info later, but for now Willem's Land Area % was the most interesting thing I noticed:

Spoiler :

^^^^ probably screws up my earlier analysis.

 

[shakabrade]

I agree with Doshin's scout movement and reasoning behind it.

 

[shakabrade]

It's just a roughly estimation. Because the tasks, currency becomes a necessary tech to learn early in this game. Therefore the gain from Oracle is the beaker cost of Currency ~600B plus other economical gain for about 20 turns. Let's assume 1000B totally. The cost of building Oracle is 150H(probably = 300B) plus the techs and worker turns, let's assume another 300B. Therefore the net gain from Oracle is around 400B. If we don't have mids, then the research rate in the final period to Astronomy would be around 150bpt, which needs 3 turns to output 400B.

As I understand, Oracle route involves skipping TW and Pottery. That is a huge waste. In 50T, you can get 480, which is roughly 580 from just 3 riverside cottages. No risk. And postponing Granaries hurts our production. Delaying TW wastes worker turns and makes us less flexible against barb threats.

Of course, the real situation is more complicate, the most important thing is whether the GS or the research rate is the bottleneck. With Phi trait, mostly the bottleneck is the research rate, that's why I insist on bulbing Math and Oracling Currency instead of building GLH.

I agree that research is a bottleneck here. GLH seems safer to me than Oracle though and we don't need to skip important techs.

It's also hard to estimate the chance of getting Oracle. Even in Monarch, AI could build Oracle before 1800BC, or in deity, AI might build it later than T65. Generally, T60~65 have ~50% chance of getting it, mainly depends on whether AIs who is likely to research PH have the marble. However, one thing for sure is that "If you don't try it, you'll never get it". The thing is that we should control the process to minimize the cost in case that we lose. For instance, chop the forests in the last minute and complete it in the shortest time to avoid wasting hammers.

I fear that the fact of having 10AIs leads to increased chance of losing Oracle due to following reasons of which some are tangled:

- Wonders are RNG thing, more AIs, more chance, some AIs have their flavours, but 10 AIs will, I am pretty sure, have better average Oracle dates than 6 AI games
- We already have Izzy on the map
- 10 AI's will have no land for cities, especially with 26% landmass which includes mountains, and we can see mountains are present, so, more chance to go for wonders
- more chance AIs will connect Marble

 

[Kaitzilla]

Good work getting our 2nd city up safely and quickly Doshin.

8 hours trying to fix Steam Civ, sorry to hear that.


I'm a bit surprise we haven't met any AI yet.
No idea why Willem has 0% land.
He must have a city because he leads us in .

Fogbusting with Scout to keep real barbs away seems ok to me.
Team seems very cautious using the Scout and really wants 3 cities down without any problems.

Since units fogbust in 2 tiles every direction, I'd say 2N2E of gems would be best if it is safe to stand there.
Otherwise 1NE of gems is ok.

 

[Duckweed]

@shakabrade

I'm not sure how you got the 480C from earlier pottery. We don't skip pottery, only delay it. The difference is probably 15 turns. We can still get Pottery before T60. Capital won't have much time to work on cottages. Moreover, cottages are against the bulbing strategy and also you did not count the gain from a farm.

When you evaluate something, you have to consider the loss from the other ways.

The problems of the GLH are

Its cost is more since LH is not an urgent building in capital.
Its gain is less since when you count the gain from GLH, you also need to count the loss from delaying Math and Currency.

 

[shakabrade]

@shakabrade

I'm not sure how you got the 480C from earlier pottery. We don't skip pottery, only delay it. The difference is probably 15 turns. We can still get Pottery before T60. Capital won't have much time to work on cottages. Moreover, cottages are against the bulbing strategy and also you did not count the gain from a farm.

When you evaluate something, you have to consider the loss from the other ways.

I am simply comparing risky approach with safe and boring approach = cottages.
You also have a granary in such a city.
15T of Pottery is 15T of Granaries, which will bring more cities more quickly up to generate more research or GPP more quickly.

I thought this was understood.

I may miss some of the factors but my reasoning is always trying to take opportunity costs into an account, so thanks for the advice but I'd rather like to be persuaded into Oracle thing by reducing opportunity costs of delaying Pottery, researching unnecessary techs, losing chops or postponing them (also a cost).

Oracle's worth depends on production vs. research potential ratio.
I don't think we are that good with production, we are ok at best. We don't have enough of happiness to utilize our food, we don't have that many forests, we don't have a granary. Research is good with 3 Gems and an island city. If we Oracle something, Oracle is worth it, no problems with that, but there's a high chance that we'll lose it and wait with our Granary and Library chops for some turns. GP pollution is a problem too but one GP is okay for GAge. And we can forget about Music.


The problems of the GLH are

Its cost is more since LH is not an urgent building in capital.
Its gain is less since when you count the gain from GLH, you also need to count the loss from delaying Math and Currency.

GLH is expensive, yes, but less risky and we need only one unnecessary tech compared to Oracle. So it is less expensive than you originally calculated. It saves us beakers from start and it can be built with cheaper hammers since we'll have Granary in that case to OF at least 3 whips: Lighthouse, settlers or workers, maybe Library.