[BTS] SGOTM 13 - Gypsy Kings

[mabraham]

Originally Posted by Ronnie1

I am unclear here a bit. I have tested a bit both ways, and I think it comes down to how fast we need to settle off the island. If there are better sites off island, then Sailing 1st makes sense for the early galleys as much as for the Lighthouse. But if we can live with just 2 cities for a bit, the early Granary makes our whipping dramatically more efficient. I think if there is land north, which could create a stronger #2 city, then I would lean to Pottery 1st. Sailing 1st if we have to settle #2 east where we have been testing.

Another factor suggesting early Sailing is putting a mine on the off-island GH. IMO it is the second-highest worker priority after the corn farm. The corn farm adds 2, that mine adds 2/turn to our output. Mining the PFH adds only 1. Road on the corn adds 1 which is sometimes 1. Road also on the PFH gets the 3E city up a turn faster. The early galley will also help scouting.

This is going to sound like heresy, but the granary seems too good to build. If we run a 6/3 cycle with granary, we need 20+21+23 food to regrow to size 6. At size 4 we have 14 excess food. So regrowth will take about 5 turns. At size 6 working four food resources and two mines, we produce a settler in 8 turns. We want to put on a settler for 3 turns before whipping it with max overflow. So sometimes we might have to delay a turn whipping. Alternatively we could grow and run an 8/4 cycle, preferably using < 14 overflow hammers onto a settler for the 4-pop whip. Or we could donate two of those food sources to the second city, which I think is best.

I think planning to build an early granary in the capital makes sense only if there's a second city sharing the tiles, and we know that the 90 delay is not critical (e.g. no AI pressure on our available land)

That 90 is our chunkiest early build apart from possible wonders.

Originally Posted by Ronnie1

The standard bulb path to Astro requires 4 GS's....bulbing Machinery/Optics/Astro(2). We need to research Alphabet/Calendar/Compass and avoid Meditation/Civil Service.

So how do you generate those GS's?

The Great Library gives 2 free scientists + 2 assigned + NE nets 26 GPP/t. 24 of 26 are GS points....we need 150/300/450/600 for a total of 1500 points which is 58 turns AFTER all of the above is in place. We know that we will actually be generating some GPP BEFORE all is in place, we just don't know exactly how much. So how fast could we actually generate 4 GS's?

OK, that can certainly do the job. However it'll need our first off-island city to be capable of a large pile of pretty fast - that's 900 of wonders, plus granary and library (at least). We may be lucky and have marble. Those need to be delivered somewhat later in the game than (say) the 750 on the Pyramids, but they sound like broadly comparable investments to me. In the GL+NE scenario, we also built the Colossus somewhere, so that's 1275 all up. Recall also that I made the point earlier that the Colossus strategy was predicated on there being not a great deal of land suitable for Caste workshops. If that's generally true apart from this magic site, if we were running Caste, this GL+NE site is still going to produce a pile of for useful stuff - like the start of a navy.

Originally Posted by Ronnie1

What if we don't NEED Astronomy? How much faster could we move if it was researched even if all the AI's were galley accessible? Galleons move twice as far and carry 50% more, seems like a decided advantage on the surface at least.

True, those are serious (or even necessary) advantages. Circumnavigating reduces the marginal gain in movement points, but we'd take both if we could get them cheaply enough...

Originally Posted by Ronnie1

The GE adds 150 points to our GPP total IF we want to bulb to Astro also. Not a huge deal, but a definite delay if Astro is required. Astronomy is also very effective for trading overseas obviously, which is something to consider. On the downside, it negates the effect of The Collossus.

... and negates our +1 monuments.

[mabraham]

Originally Posted by bcool

You can upload the save after the warrior move. They have changed how this works (they rename the file to xxx2). We did this numerous times in OSS last SGOTM.

OK. I gather the spirit of the discussion is that people were happy to get the new data that will be available, before Ronnie pointed out a problem, which bcool said is not a problem. I realize Ronnie didn't actually assign me the job, but in a few hours' time, I will plan to download, move the warrior, take screenshots, save & quit, and upload assuming bcool's right. The worst-case scenario is that I have to take a full first turn set. Please yell if that's not OK.

I also recommend people make a subdirectory in their Civ saved game folders, e.g. "SGOTM13 actual game" so that they have to actively choose to open the real save, and won't open it thinking it's a test game or something.

[adrianj]

I'm happy with Mab (no objection to shortening your name?) taking the save and moving the warrior.

[Grifftavian]

Originally Posted by mabraham

...

I also recommend people make a subdirectory in their Civ saved game folders, e.g. "SGOTM3 actual game" so that they have to actively choose to open the real save, and won't open it thinking it's a test game or something.

That's a good idea, and I do that on a regular basis. I created a TEST SG13 subdirectory for the test games, and also have TEST BOTM 39 subdirectory (thanks for the game R1) and a BOTM 39 subdirectory for the actual game.

[mabraham]

I ran another test on the usual SIP, 4*WB, worker, lighthouse, galley, monument. I Duckweeded one settler onto the Pyramids for the 3E site. Then I didn't whip the capital until the Pyramids would finish, and they were built T138 (550BC). I gave two food tiles to the 3E, and worked the capital's mines ASAP to maximize the natural hammer-output, rather than growth. I worked the two grassland forests as I grew, and at size 7 (with the whip- gone away) I even ran a citizen, for a peak output of 11/turn on the Pyramids. I chopped all three forests into the Pyramids, but mis-timed things so that I did only a 3-pop whip from size 8 at the death finish. Thus I could have been about 2 turns faster. Meanwhile the side city had gotten granary, lighthouse, monument, warrior, workboat, settler finished, and the third city was settled.

Clearly with this strategy the empire-wide output was somewhat lower, but the rate of finishing the Pyramids was significantly higher. I could probably have Duckweeded a second settler onto the Pyramids, really. Takes about four turns away from the Pyramids, but returns 40-odd , so it breaks about even. Edit Nah, who am I kidding? Takes about 10 turns to re-grow from 3 to 6 even with all four food resources, and that's the source of delays on the Pyramids.

My best full-Duckweed Pyramids run was T155 (with five settlers total built). If you give this game another 17 turns, I'll get out at least two more settlers (side city has one half-built). So I estimate that the cost of the increased certainty of the Pyramids is delaying the settling of cities by about 15 turns each, and maybe one fewer city. So, if the Pyramids looks like the way to go, I think we need to give serious consideration to such a hybrid.

[mabraham]

Originally Posted by adrianj

I'm happy with Mab (no objection to shortening your name?) taking the save and moving the warrior.

Sure, "mab" is fine. Ronnie asked that earlier - sorry, I forgot to reply to it.

[Ronnie1]

and we know that the 90 delay is not critical

I would not delay for 90H.....I would overflow a settler or worker into a granary at size 2 and whip it the next turn to size 1 and regrow so fast it will make your head spin!

So have you moved the warrior yet or not?

Fog gazed the real save...we are definitely on a 6 tile island. So our testing to date should be very accurate!

Had a meeting this evening, so haven't done the roster yet.

[mabraham]

Got the save. Sorry, RL intervened for a bit. Will move.

[mabraham]

OK we learn a big pile of nothing.

Spoiler:

There were no surprises on the settings screen. The original map script has not been disclosed. No info screens said anything interesting.

I did some fog-gazing and the results are now noted on the screenshot. I found no evidence of land other than where indicated.

There is an unanticipated issue. When uploading the save after moving, I get

ERROR
There was a problem with your submission:

Your entry has not been recorded. Please correct the items marked * and resubmit it.

# You can't upload a new start file.

I'll PM to inquire.

[mabraham]

Originally Posted by Ronnie1

Fog gazed the real save...we are definitely on a 6 tile island. So our testing to date should be very accurate!

Agreed.

Settling thoughts
The 3E city we have been considering looks to me like it will have access to a grassland hill mine on the SE side island. (I'm basing the my deduction that the tile is a hill on a spur/ridge that stands out when you use the Ctrl-left and Ctrl-right alternative views.) That's great, because at size two it could work both grassland hill mines and pump out warriors and workboats (and later siege) even when it doesn't get access to the food resources.

If we don't settle in place, there may be similar merit in settling on the W-island forest, because there's another (unknown) land tile in its BFC. That plus settling the PFH is not as good a scenario as the SIP-and-3E combo, however.

One could settle PFH and on the W-island hill, but I cant see that addressing our primary need of more natural .

[mabraham]

Originally Posted by Ronnie1

I would not delay for 90H.....I would overflow a settler or worker into a granary at size 2 and whip it the next turn to size 1 and regrow so fast it will make your head spin!

We-e-ll. Suppose the food box is half full at size 1 when you're done the double-whip. You'll spend about 9 turns regrowing to the happy-cap of size 4 (pre-monument). Now if you can't share those food tiles off to another city, you're going to have to put that food excess of 14/turn onto settlers for about 19 of the next 21 turns. Call that two whole settlers. You get to grow to five in the middle somewhere. Then you can organize another round of whipping.

That's all well and good, but there's going to be a massive ongoing food glut that we can only apply to building more settlers. That means we need lots of settle-able land, and the ability to pay for those cities.

The post-granary food glut is much more manageable if it is spread across two cities in the early phase. Now more of it can be used growing population points more efficiently, and these can then be whipped into things other than settlers - as well as into settlers.

This is why I think building an early granary in the capital makes good sense only if we're planning a second city to share the food tiles.

[bcool]

sorry apparently I was wrong, you can upload multiple saves on the same date except apparently the start save.

[Grifftavian]

FWIW, I've updated my TEST SAVE based on mab's guesstimate of the land in the nearby fog, and R1's suggestion to give the GLH to Ragnar!

We should probably move the Warrior SE on the next turn to verify if that is a Grass Hill off to the SSE of the 3E second city site. Should we go ahead and SIP on Turn 0, or wait a turn for the next Warrior move?

[Ronnie1]

Originally Posted by mabraham

We-e-ll. Suppose the food box is half full at size 1 when you're done the double-whip. You'll spend about 9 turns regrowing to the happy-cap of size 4 (pre-monument). Now if you can't share those food tiles off to another city, you're going to have to put that food excess of 14/turn onto settlers for about 19 of the next 21 turns. Call that two whole settlers. You get to grow to five in the middle somewhere. Then you can organize another round of whipping.

That's all well and good, but there's going to be a massive ongoing food glut that we can only apply to building more settlers. That means we need lots of settle-able land, and the ability to pay for those cities.

The post-granary food glut is much more manageable if it is spread across two cities in the early phase. Now more of it can be used growing population points more efficiently, and these can then be whipped into things other than settlers - as well as into settlers.

This is why I think building an early granary in the capital makes good sense only if we're planning a second city to share the food tiles.

I agree completely...in what I think was my best test, I whipped a worker on T74, and the Granary was whipped on T76. City #2 was already up and made use of the Clams and the FPH while cap was regrowing!

Cap builds went like this...Warrior(partial, MM so cap grows to size 2 coincides with Fishing)>WB#1>Warrior finish(3t)>WB#2>WB#3>Settler starts T48(whipped T58)>Lighthouse(T59, 1 turn build)>WB#4 starts T60>T70 Cap grows size 4 and switches to worker, whip worker T74>Granary starts T75(whipped T76)>WB#4 finishes>Galley starts T78...

I did not take great notes and plan to run more tests today. I was able to whip the Oracle about T105 I believe. The biggest slow down on Oracle date was actually researching Priesthood. Research went Fishing>Mining>BW>Sailing>Pottery>Myst>Poly>Priest learned T83

[Grifftavian]

Originally Posted by Ronnie1

...

I did not take great notes and plan to run more tests today. I was able to whip the Oracle about T105 I believe. The biggest slow down on Oracle date was actually researching Priesthood. Research went Fishing>Mining>BW>Sailing>Pottery>Myst>Poly>Priest learned T83

Couldn't we save 2 or 3 turns getting to Priesthood by researching Meditation instead of Polytheism? Any reason why we don't want to get Meditation that early?

[Ronnie1]

Originally Posted by Grifftavian

FWIW, I've updated my TEST SAVE based on mab's guesstimate of the land in the nearby fog, and R1's suggestion to give the GLH to Ragnar!

We should probably move the Warrior SE on the next turn to verify if that is a Grass Hill off to the SSE of the 3E second city site. Should we go ahead and SIP on Turn 0, or wait a turn for the next Warrior move?

I am not positive we have a consensus on everything we need to start Turnset #1, but I personally am pretty clear that SIP is our best option, and I THINK we have consensus on that fact.

It would be great if we great if we could get everyone to weigh in on that issue.

I will do the roster after I have some more coffee! Obviously mab is going to settle our cap as he has to continue at this point.

I think we need to get consensus on a few issues and then set #1 can continue.

Please remember, we are NOT in a hurry, we want to make great decisions, and we want to avoid errors in both our preparation/thinking and in our actual game play.

[Ronnie1]

Couldn't we save 2 or 3 turns getting to Priesthood by researching Meditation instead of Polytheism? Any reason why we don't want to get Meditation that early?

We could indeed, but it takes away the OPTION of bulbing to Astronomy.

[Grifftavian]

^OK, Thanks!

[bcool]

Let's try to come up with a plan to the end of fishing so we can test with full knowledge of the starting island and resources that might be within reach.

I think we SIP and count ourselves lucky if the 2nd city 3E has a seafood resource of its own.

Tech Fishing seems obvious.

I guess what to build isn't.

I would think it either a barracks for a slower decay of invested hammers (although even these hammers decay in my test games)
Or a warrior, which I personally never finish, but instead start a workboat asap. (which means those hammers invested in the warrior decay away)
I think the worker first plan has been shown to be weaker.

MM might be an issue, but I think running the corn until fishing seems the best to me.

So i propose tech fishing, build barracks, work the corn, explore the entire island, then stop to develop more accurate test games.

[Ronnie1]

I actually like Warrior 1st for these reasons....1) we will need another MP sooner or later anyway, 2) once cap is size 2, it is only 3 turns to complete, 3) I hate wasting those 10 hammers we invested if it can be avoided, which it can, for very little delay in WB #2.

MM might be an issue, but I think running the corn until fishing seems the best to me.

Corn for sure early on, we grow to size 2 with 10 hammers into whatever we build and we are 1 turn from size 2 when fishing comes in. Grow to size 2 and add in the FPH, the WB completes T22 and nets same turn, now working nets and FPH for 4H/t and warrior will complete in 3 turns (T25).

The first set is always the most boring, but those 25 turns should be the start we are looking for I think!