SGOTM 13 - Gypsy Kings

[adrianj]

Emailed another test game out for review at T151.

Sorry Ron, that isn't particularly helpful for me. Most of my contribution is made while at my desk at work. I can't actually play the game, so most of what I contribute is based on assumptions and memory

What are the main pros/cons?

 

[Ronnie1]

The progress page culture graph shows we definitely built a wonder at T94!

Maybe we should not get in too much of a hurry at this point....

 

[Ronnie1]

Sorry Ron, that isn't particularly helpful for me. Most of my contribution is made while at my desk at work. I can't actually play the game, so most of what I contribute is based on assumptions and memory

What are the main pros/cons?

Well the 1st test I played it as planned mostly....Pyramids in CC on T141, 2 more side cites up, 2 clams and marble, GLib started in CC T151, GS settled in CC

Test 2 I went Pyramids in FH, whipped T150....3 side cities settled, same 2 + copper, extra worker built, GLib started in CC T151, no GPers born yet.

The text in the emails give some details also.

 

[adrianj]

Well the 1st test I played it as planned mostly....Pyramids in CC on T141, 2 more side cites up, 2 clams and marble, GLib started in CC T151, GS settled in CC

Test 2 I went Pyramids in FH, whipped T150....3 side cities settled, same 2 + copper, extra worker built, GLib started in CC T151, no GPers born yet.

The text in the emails give some details also.

Any extra detail on actual produced? Eg, how far through whichever current tech are you?

Am I right in thinking that Test 1 is similar to Case 3 from my earlier post, and Test 2 is similar to Case 5? Is there any merit to trying something along the lines of Case 4, that is building Pyramids in CC and letting GProphet come out slowly?

 

[mabraham]

Any extra detail on actual produced? Eg, how far through whichever current tech are you?

Am I right in thinking that Test 1 is similar to Case 3 from my earlier post, and Test 2 is similar to Case 5? Is there any merit to trying something along the lines of Case 4, that is building Pyramids in CC and letting GProphet come out slowly?

The test game I reported a few posts back followed aj's Case 4. I can report more stats for that later today when I have the file. I think I over-expanded a trifle. The need for workboats, workers, triremes and galley turns was acute, and a plan to expand fast is only worthwhile with a thorough infrastructure support plan. I had seven of the eight planned side cities up by T144. That's respectively two and one more cities than Ron's recent test games.

We probably want to compare dates for finishing the GLib in the leading strategies also. I got mine about T168. We also want to report tech status, sustainable tech rate, GP status at T150 and (say) T170 for real comparisons.

 

[mabraham]

I think we can abandon the library in FH as long as we can efficiently use the food in FH. I assume we can manage the food with efficient whipping and settler/worker builds. There was some concern that it would run into happiness issues without the scientist outlet.

Certainly FH happiness must be managed either way, but a combination of building a trireme or workboat while re-growing, switching to settler or worker according to timing once it is at full size, and Duckweeding a settler or worker onto the boat as the decays worked well in my test game and gets full value for all the available tiles - which would often include the corn and crabs from CC.

I would like to try test games both ways but I don't think I will have the time tonight. Or this week really. However the Great Prophet is looking okay to me.

Me too. I thought bcool's analysis a few days ago concluded that an early GScientist does not deliver enough value settled or as an academy compared with waiting and bulbing. Once we've bulbed Astro, we'll still have bulb opportunities on Paper and Education. The early GProphet gives us a significant return then and for the rest of the game, unlike a settled GScientist whose return drops off once we finish our tech path.

The prophet would probably accelerate the pyramids and GLib as well. Those 2 bonus hammers are pretty useful I imagine.

Yep. Our base hammer rate comes from the 8 from the mines and central tile. The GProphet is somewhat like a limited-functionality forge in that regard.

About scientists in FH,....a few thoughts.
Pros for running scientists early
1) We could build early academy with a GS if we get one.
2) If we get a GP, we get him way sooner, almost 40 turns sooner I believe.
3) The factor of those scientists
4) 25% science boost in city when running 100%

Cons for running scientists early
1) Give up & , the food of course gets converted to with the whip.
2) #1 leads to slower expansion

I'm sure I'm missing something, what is it?

Settling the GProphet gains earlier, when it matters most. The best uses for Gscientists accrue later.

Are we ever going to build OxU? I don't think it pays off unless we really miscalculated the level of military we need for the VC's. This quote is the only thing that has me slightly worried..."Level: Notionally Emperor, but this is a difficult map for warring." Are we underestimating the difficulty of this map in some way?

I think we're in trouble if we need OxU. We have no city site worth building it in. At best, our capital produces beakers from 7+2 scientists under Rep (54), 5 from tiles and (say) 3 trade routes at 5 (by the time OxU comes up). At 100% slider, that's 74 to double, which is pathetic. I think OxU is playable only if we capture an AI capital or discover the New World.

I am tempted to settle Paired Clams next. It needs a Granary + LH before Maoi, 2 WB's for the nets and worker turns on the mine and workshops.

Sure, it's a lively candidate for our fourth city. I settled it fifth and got Maoi T189. Earlier would be better.

Marble city needs Granary + LH + 1WB + marble quarry before we start GLib, and it is closer.

Actually, Marbled Clam only needs quarry and road before GLib. Its city infrastructure is decoupled from the wonders.

Piggery...not a great site other than access to the pigs, it can wait IMO, maybe never depending on other good sites availability in the fog.

Sure, it's not a great site. Its value is good in the last 15 turns before the switch to Caste, when FH can switch off working its fish and give the food to a late-settled Piggery for a last-gasp poly-whip of its infrastructure. Long term, Piggery is just a draft farm.

Accelerating the prophet/scientist in FH also increases the likelihood of us generating ANOTHER prophet for the second great person. By slowing it down we can then let CC take the lead on GP production.

Nah. FH (or any other side city) will only be competitive with CC for if they run several specialists for a long time, to compete with the 5-9 specialists, 3 wonders under the NE in CC. In my played-out test games I did get some GMerchants in those cities that ran merchants, and sometimes that included FH, but that was by design, not by force of the Oracle.

My opinion for the ideal GP scenario is that FH builds the Pyramids (and associated settlers) while CC builds a Library and starts growing/running scientists/building workboats/warriors.

Going into more detail:
FH pops Prophet on T169 (using BCool's guess, I also don't have game access) For simplicity's sake I'll also assume this is when the Mids get built here.
I'll assume the library is next build in CC, and from T130 it is full-time running 2 scientists. So at T169 it has 39*6 = 234! What the heck! It just *surpassed* FH in producing a great person did it not?

Bah - forget this attempted analysis! Why don't we just make CC produce a GS first at 100% odds?


Ok, some detail on this *other* option.
IF FH builds pyramids say ~T150, then it will produce a Prophet/Engineer on T160.
To guarantee a GS from CC before T160 we need the library built and functioning before T135. Is that possible? Is it reasonable for FH to be busy with the settlers/mids buildings while CC slows itself down a bit producing a GS?

If we take the idea of CC focussing on scientists a little further, and assume it builds GLib approx T170, then the second great person will come out from there on T190 - this is also before FH produces an unwanted Prophet. So, we could in fact have TWO Great Scientists before we worry about Prophets spoiling the mix.

As above, I don't see value in early GScientists if we will eventually get a GProphet anyway. An early GProphet and later two GScientists that don't have to wait around so long before they are useful for bulbing is a much better return, unless someone can demonstrate that settling a GScientist or building an academy is better than bulbing.

Also, the engineer in CC contributes to the GP production rate, and even two GEngineers are OK - one bulbs Machinery and the other builds the GT (or whatever seems best later). In Fish Hills, its value is lower, if any.

The game says 1st GP 61 turns from NOW at T108 = T169 for the GProphet.

I like the idea of getting GS production going in CC.

Can we alter our plan to get Pyramids up in a reasonable amount of time in FH?

We probably could (and you've now demonstrated that we can) - but what is the value of GS production before we get the NE up, compared with building more stuff and running the GS later?

 

[mabraham]

Eastern exploring WB can keep heading east. My preference is to not cut corners, as it would be irritating to miss possible coastal links.

I would argue that finding coastal links when we're already 20+ tiles from our capital is not a priority. Even if we found one, we'd be guessing which route would meet our goals best (AI contacts, trade routes, homelands, and circumnavigaton). In such a case, our investment in thoroughness need not have any payoff, will have a payoff only when we guess to follow the route to the new island and that works well (i.e. comparing to making faster progress along the coastline we're already on), and still could lose the vital circumnavigation bonus.

Perhaps a better way of looking at this is that (IMO) we want Cook to make progress east. So cutting corners that might expose routes to the west, or due north or due south are not a big loss, as above. Corners that expose possible routes to the NE, E or SE should not be cut, because they offer us the best chances to gain.

 

[adrianj]

Given how much research we still have to do the old fashioned way (ie, not bulbing Machinery or Astro) I think a settled GS or Academy will still be valuable.

If we do want 3 Great Scientists to bulb Machinery+Astro, then these could be the 3rd, 4th and 5th. At a cost of 1800 our GLib/NE capital can get out all three of them in about 30 turns under Caste System.

I agree with what you say about sending Captain Cook east.

 

[mabraham]

Given how much research we still have to do the old fashioned way (ie, not bulbing Machinery or Astro) I think a settled GS or Academy will still be valuable.

True, there is a bunch of manual research to do as well, and that goes well past the time when we might be able to bulb Astro. We will need to bulb Astro before doing the Civil Service and Guilds trees, of which we will probably need to do most ourselves. Astro is bulbable in the T220-240 time frame, but there's no possible way to get through Nat, Chemistry and Lib->Steel in that time. A settled GPerson and academy keep working all through that time. A settled GProphet keeps working through the war phase, as well.

If we do want 3 Great Scientists to bulb Machinery+Astro, then these could be the 3rd, 4th and 5th. At a cost of 1800 our GLib/NE capital can get out all three of them in about 30 turns under Caste System.

In practice, we'll generate further GScientists, because our primary source is still likely to be the 7+2 specialists in CC. These can bulb Paper and Education, of course.

 

[mabraham]

Decisions
1) Quite important - Where to build Pyramids. CC is the original plan, but FH may also be worth considering.
2) Once we decide 1), what build order will get us there fastest? Eg, when do we fit library into the queue? Does FH build a lighthouse?
3) Do we prioritise in FH to get a GProphet sooner, and chance at a GScientist?
4) Shoot for Currency or Literature first? Where does Maths fit in?

My thoughts on these things:
1) A plain forgeless Duckweed in FH on my spreadsheet came out T159. I could arrange to whip it T157. Doubtless I could do a few turns earlier. The lack of the forge, a chop and the palace- contributes to the slower rate. So far, the jury is out on the advantage of the earlier settled Gscientist from CC (which slows growth), but it has to be large enough to offset the 10+ turn delay on access to Rep and the increased risk of losing the Pyramids.
2) Library never comes before Pyramids in either city. FH needs a lighthouse before the pyramids, because it works two seafood every turn and even with the lighthouse we have to work the corn rather than the Pmine on a few turns to grow fast enough (the Palace- would be beneficial here...). It needs a monument also so that it has enough to work two mines, two seafood and corn to grow back to size 6 in the 11 turns we have after whipping before we have to work another settler to max the next overflow.
3) I like expanding, but I will run a comparison.
4) I prefer Aesthetics->Literature before deciding about Maths, and then probably Currency+CoL before Maths - AIs will trade Maths much more often.

 

[mabraham]

The test game I reported a few posts back followed aj's Case 4. I can report more stats for that later today when I have the file. I think I over-expanded a trifle. The need for workboats, workers, triremes and galley turns was acute, and a plan to expand fast is only worthwhile with a thorough infrastructure support plan. I had seven of the eight planned side cities up by T144. That's respectively two and one more cities than Ron's recent test games.

We probably want to compare dates for finishing the GLib in the leading strategies also. I got mine about T168. We also want to report tech status, sustainable tech rate, GP status at T150 and (say) T170 for real comparisons.

Stats from my test game (aj's Case 4)

T143 6 cities, one ready to settle, done Pyramids

T150 7 cities, 19 pop, 33 sustainable , GPro due in 20, done Lit, just started Currency (traded for Alpha, IW, Mona)

T170 8 cities, 32 pop, 49 sustainable , GPro just popped (about to settle), Currency 99% done along with Lit (traded for Alpha, IW, Mona).

T179 Finished GLib (not T168 like guessed earlier!) and CoL

T193 11 cities, 42 pop, 125 sustainable , second GP just popped, just done Compass, Lit, Drama, Currency, CoL (traded for Alpha, IW, Mona, Maths, Calendar). Next GP is due from CC in 10, FH is not in the game. I definitely did not have enough workers, workboats, galleys and triremes to sustain all the new cities. NE had been up for a handful of turns.

 

[mabraham]

Thanks for the valuable analysis aj, but I'm going to pick some holes in it.

Assume: CC is building Forge -> Mids (T150) -> Library,+2 scientists (T160) -> GLib(T170).

Actually Pyramids should go up in CC around T141-143 depending whether we whip it to finish. We do not have time to run scientists when we finish the library, we have to keep Duckweeding the GLib. 10 turns for GLib is very optimistic. Our two mines take 30 turns to build it, reduced somewhat by Duckweeding, and this can't afford scientists on the side.

This means the second GPerson is rather later that T183 (I got T193, above) but probably has comparable chances for GE:GS.

The previous analysis underestimated the returned by settled GPersons and the Academy (Rep is earlier), overestimated the Academy's bonus from the natural scientists and GLib (there are no scientists, GLib is later). So Case 2 looks a fair bit worse with respect to cases 1 and 3 than it used to. Cases 1 and 3 are fairly close - the extra hammers get the forge bonus and the extra beakers get the library bonus.

Cases 4 and 5 are unaffected.

Cases 1-3 and 5 are all about a faster GPerson, case 4 is all about faster expansion. So the questions boil down to: do we want an earlier settled GPerson, or faster expansion?

(I think expansion, but we need more testing)

 

[mabraham]

My play-through of aj's case 5

T150 6 cities, 15 pop, 46 sustainable (incl 2sci and settled GSci from T146), done Lit, started GLib (marble due next turn) about one-quarter done on Currency (traded for Alpha, Mona, IW)

AI popped a GEng T152, luckily they built ToA...

T170 8 cities, 24 pop, 72 sustainable , whipped GLib this turn, finished Pyramids T166 (way way way too late), done Lit, Currency and 3/4 of CoL (traded for Alpha, Mona, IW)

AI built HG T175

T193 10 cities, 38 pop, 154 sustainable , (GSci T186->academy), done NE T184, done Lit, Currency, CoL, Drama, Compass, Feudalism and 1/5 of Music (traded for Alpha, IW, Mona, Maths, Calendar). FH is still a long way from being in the game. CC is ready to switch to CoL to run 7 specialists in 3 turns' time. Next GP in about 8 turns.

I can only presume my woeful Pyramids time was because I Duckweeded it after lighthouse and monument. Ronnie must have built it straight, and probably not put out lighthouse or monument first.

 

[bcool]

Thanks mabraham for the testing and stats!

If we can improve the pyramids time in case 5, then I think that is looking better. The tech rate is significantly faster (Drama, Compass, Feudalism and 1/5 of music extra is significant). If that is the cost of 1 fewer cities then that is pretty good.

Maybe tonight ~12 hours from now I can give it a shot.

 

[Ronnie1]

Thanks mabraham for the testing and stats!

If we can improve the pyramids time in case 5, then I think that is looking better. The tech rate is significantly faster (Drama, Compass, Feudalism and 1/5 of music extra is significant). If that is the cost of 1 fewer cities then that is pretty good.

Maybe tonight ~12 hours from now I can give it a shot.

We can improve the Pyramids time here for sure....I built it here on T150 (whipped), WITHOUT a forge/lighthouse/monument. Ducked in 1settler + 1 worker.

Based on the date of Stonehenge(1725) and the Great Wall(not built), I think we have a bit of time to build Pyramids.

I'm going to play with it some more and see if I can improve on it and take some better notes while doing it.

 

[bcool]

Thanks ron for testing too, we need to see what an early pyramids sacrifices. It looks like we sacrifice a lighthouse/monument (and FH forge?) which isn't terrible in my opinion. If we get similar expansion then a case 5 T150 pyramids looks competitive.

 

[bcool]

AI popped a GEng T152, luckily they built ToA...

this is pretty unusual isn't it? They went for early metal casing and ran an engineer I guess. I suppose the GLH AI is going to pop Great people faster because of tech acceleration, early lighthouse for more pop and the free early from the GLH itself. Even after the 1st merchant perhaps.

 

[mabraham]

Stats from my test game (aj's Case 4)

...

T193 ... just done Compass, Lit, Drama, Currency, CoL (traded for Alpha, IW, Mona, Maths, Calendar).

My play-through of aj's case 5

...

T193 ... done NE T184, done Lit, Currency, CoL, Drama, Compass, Feudalism and 1/5 of Music (traded for Alpha, IW, Mona, Maths, Calendar).

Thanks mabraham for the testing and stats!

If we can improve the pyramids time in case 5, then I think that is looking better. The tech rate is significantly faster (Drama, Compass, Feudalism and 1/5 of music extra is significant). If that is the cost of 1 fewer cities then that is pretty good.

Note that Drama and Compass were common to both runs.

 

[Thorn]

Okay, here's my rough draft PPP. It's not complete and need mega feedback! I will update this post as changes are made.

PPP Turn 108 to Turn 122

Diplomacy
Sign Open Borders treaty with Vicky

Research
Start Aesthetics
Literature

Civics Changes
None

City Builds

Clam Chowder

1. Forge
2. Library

Fish Hills

1. Workboat
2. Lighthouse

Stone Mountain

1. Monument
2. Workboat


Unit Moves

Worker 1

1. build Mine
2. move to CC if needed or build workshop on SM grassland

Cook

1. fog bust going due east
2. continue to fog bust

Galley 1

1. ferry Worker 1 to CC if needed

Warrior 1 > sentry in Fish Hills
Warrior 2 > sentry in Clam Chowder


Unplanned Pauses (if any)

* when anything not talked about/new happens!!!

Spreadsheet details <<< not complete and need lots of help!

Also, do we want to have any diplomatic interactions with Vicky? Open Boarders treaty comes to mind.
I will wait for the change requests to pour in.

 

[adrianj]

Good start Thorn.

I definitely think we should OB with Vicky. I don't think we have anything to lose by it. There is the added bonus that this would enable trade routes, which gives us a little knowledge as to where she is. Ie, if we *don't* get trade routes then she must not be reachable by galley.

I don't think we yet have a concensus on who is building the Pyramids. I actually think Ron's version where FH builds Mids after the current WB is quite nice - so in this case FH does not build Lighthouse and is always working the mines.

Referring to my cases a few posts ago - I guess there is also a 6th case: Build Mids in FH and still produce GProphet/GEngineer. It's similar to Case 5 but it puts less pressure on CC to run scientists

Tech path: Aesthetics->Literature seems to have a concensus. Are you familiar with "binary research" Thorn?