SGOTM 13 - Gypsy Kings

[mabraham]

I ran another Pyramids test on Grifftavian's test save. This time I got a brain, settled in place built 2WB, WB 1-whipped onto WB, worker 2-whipped onto lighthouse, galley 2-whipped onto monument, Pyramids. I settled the second city 3E and did a Duckweed-style settler 3-pop whip max-overflowing onto Pyramids, while tile-sharing a couple of turns to the second city on the corn per whip cycle. I got Pyramids T155 (275BC) with four side cities settled and one more ready to settle. I had Maths/Alpha/Monarchy/MC/IW, all the backfill techs except useless Horse Riding, and two thirds of Currency. An AI beat me to CoL while I was teching MC, which was my primary trade tech, as well as barb-galley defence.

One critical improvement was that I managed to have enough food at the start of the cycle so that the whip at size 6 went down to 3 and immediately re-grew to 4. The side city got one turn working the corn then, and then usually a couple more towards the end of the whip cycle while I switched off the corn to work a grassland forest so that I got some more raw , didn't wastefully grow , and didn't fail to work good tiles when they could be worked. I also delayed the two later forest chops as much as I could, and was lucky to trade for Maths to pick up 15(=2 turns) on the last chop.

Stats

Spoiler :

Hammer values are only for the Pyramids, and exclude those from chops - effectively this Pyramid was size 645, not 750. Whips occurred on T70 and every 15 turns until the last one.

Turn Population Hammers Delta-hammers
 70  4            0       -
 72  4            9       9
 87  4          113     104
102  4          225     112 (health trades and both mines were online by now)
117  4          331     106 (somewhere before this I gave the side city an extra corn turn or two to get to size 2 faster)
132  4          442     111
147  4          553     111
155  3          649      96

Ignoring chops, overflow in and final whip, this delivered 6.8/turn to the Pyramids. It's 7.6/turn counting overflow in and final whip. It's 16.4/turn average counting the five settler builds. This is lower than my previous average rate (17.2/turn) because I had less overflow in, finished with one population point more, and spent less time at the peak output after T95 or so.

The 6.8 number is a the one that I think people can usefully use as a lower bound in (say) a Colossus-building scenario, if they want to. The mines and trades will be up by the time we got MC+forge up somewhere.

Post-pyramids saved game

I did try a run sneaking out a second worker rather than donating food to the second city, but it was poor. There's not enough work around for two workers early enough, and if you don't get the second out early, the rot.

 

[da_Vinci]

Hehe, I'm sure dV has plenty to say on the topic of assuming safety from barbarian invasion

Goes without saying ... if you were around for my blunder.

If you were not, short anwswer is "Never assume, it makes an ### of you and me."

If we are on small islands, then our early expansion does not need defenders, which is very useful because we are -limited early on.

Hmm, if we whip as much as discussed, then we are not hammer limited, we are food limited (since we turn all our food into hammers, and run low pops).

I think it is possible for both cities to run around size 4 whipping to size 2 on what we can see already. Capital growing to 6 to whip settlers to 3 needs all the food for long enough that the other city basically doesn't whip.

I'm not sure that 8/4 is a good cycle for us to run, given our strategic considerations.

If we whip 4/2 cycles, we get one unhappy face for each cycle, right? And the cycles will come faster at 4/2 than the whip unhappy will wear off, right? At emperor we have what, 5 happy in cap and 4 in city 2? Seems to me that after 2 or three cycles, we start having the problem of unhappy citizens NOT working the food tiles if we are whipping aggressively. I agree that a mad citizen is not a bit deal if it would otherwise work a weak tile, but one idle citizen out of two would be a killer in this start. Are the test games finding that unhappiness is not an issue with our intensive whipping campaign?

So far, the leading suggestions are some degree of Moai/Colossus/HR/Bureaucracy capital, or some kind of NE/Pyramids-Representation/Caste/Pacificism GP farm running 11 specialists.

I'm skeptical that a plan to move the capital will bear fruit fast enough - we have to find a good enough site within about the first 100 turns, settle it, cottage it, build a palace, switch to Bureaucracy, grow the cottages, and then later maybe switch to Nationalism.

Well, if the capital stays on Elba, then the second city is really useless long term if we want the cap to be working lots of tiles. Even if the cap moves and we GP farm Elba, city 2 is still pretty useless, but it might make a catapult farm.

Seems to me that no matter how late in the game, there will be a better place for the capital than Elba.

Two 4/2 cities has to be fastest for crazy-REX (more tile usage more often, more efficient whipping at smaller sizes), however that's no good unless we have a plan to be able to parlay that REX into enough lead to conquer the world, and enough to make it work. There's no Great Lighthouse, and just Currency is not enough. That means we have to

• Oracle some critical tech that will do the job,
• build the Pyramids,
• build the Colossus, or
• run HR with a zillion warriors while working a pile of coastal tiles when we have nothing better to do
• find some real Bureaucracy capital site and move there

One tech from oracle can be nice, but maybe a nice tech trade is almost as good? Early representation, especially with our food rich Elba, seems like a very useful path.

Final thought ... no one even groaned, (much less laughed) at my Elba room joke?

dV

 

[adrianj]

If we whip 4/2 cycles, we get one unhappy face for each cycle, right? And the cycles will come faster at 4/2 than the whip unhappy will wear off, right?

I think the definition of a cycle is basically 15 turns, ie, the time for the previous whip to wear off. So a cycle defined in this fashion is sustainable long term.

Well, if the capital stays on Elba, then the second city is really useless long term if we want the cap to be working lots of tiles. Even if the cap moves and we GP farm Elba, city 2 is still pretty useless, but it might make a catapult farm.
Seems to me that no matter how late in the game, there will be a better place for the capital than Elba.

I'm all for moving the capital. I'm sure we can find a better capital location elsewhere, meanwhile for the whole early part of the game our first two cities do not a lot more than produce settlers/workers/whatever - and later can generate navy/army.

Final thought ... no one even groaned, (much less laughed) at my Elba room joke?

I totally missed it. It's funny now - more so since you had to go to the extra effort to point it out

 

[da_Vinci]

I think the definition of a cycle is basically 15 turns, ie, the time for the previous whip to wear off. So a cycle defined in this fashion is sustainable long term.

So if every 15 turns for the unhappy to wear off is the rate limiting step, then maybe the fastest thing to do is to whip more hammers in each 15 turn cycle? Which means whip more pop in each 15 turn cycle (not just two, but perhaps three ... although maybe just works in cap where happy cap is 5? Getting a hapcap 4 city to pop 6 could be a slog.).

Alternatively, if we think it has enough accelleration potential, two whips in 15 turn may be the way to go (at least once), then cycling from -1 to -2 whip unhappy instead of 0 to -1 whip unhappy.

dV

 

[bcool]

So if every 15 turns for the unhappy to wear off is the rate limiting step, then maybe the fastest thing to do is to whip more hammers in each 15 turn cycle? Which means whip more pop in each 15 turn cycle (not just two, but perhaps three ... although maybe just works in cap where happy cap is 5? Getting a hapcap 4 city to pop 6 could be a slog.).

You don't really get more hammers for the wonder if you pop more population if you are using the overflow for the wonder. You get a maximum of 44 hammers in OF from the whip regardless of how much pop you whip. Of course you could whip something larger with more pop, but there really isn't anything bigger than a settler that we might want to whip early on. And whipping the wonder directly imposes a penalty, so that is usually not the best way to go.

If I remember correctly the happiness with a monument is 7 in the capital. Especially if we whip 3 pop settlers than the unhappiness we accumulate probably won't be too bad. But even if we do accumulate some unhappiness because we are whipping faster than 15 turns, the extra production is worth it if we are producing something big.

If we are isolated with just small islands I think we should try to get all 3 big wonders...

1) Oracle some critical tech that will do the job,
2) build the Pyramids,
3) build the Colossus, or
4) run HR with a zillion warriors while working a pile of coastal tiles when we have nothing better to do
5) find some real Bureaucracy capital site and move there

Oracle Metal casting so we can get a GE for the pyramids, and and perhaps whip build the Colossus in the same city we are getting a GE or whip build the Colossus in another city (where we would have to build a forge as well).

 

[Ronnie1]

@mabraham...are you ok with me using some sort of abbreviation when referring to you?

Can you re-post your test games please, I have downloaded them twice and get an error message saying the file is invalid.

 

[mabraham]

If we whip 4/2 cycles, we get one unhappy face for each cycle, right? And the cycles will come faster at 4/2 than the whip unhappy will wear off, right? At emperor we have what, 5 happy in cap and 4 in city 2? Seems to me that after 2 or three cycles, we start having the problem of unhappy citizens NOT working the food tiles if we are whipping aggressively. I agree that a mad citizen is not a bit deal if it would otherwise work a weak tile, but one idle citizen out of two would be a killer in this start. Are the test games finding that unhappiness is not an issue with our intensive whipping campaign?

Yes, per adrianj, cycles are intended to be normally 15 turns long so as not to accumulate . Thus, because we have enough food to grow faster than that, we will spend time converting the excess food into settlers and workers. We should plan life so that we put the whip on the settler/worker with max overflow so as to convert that excess food to hammers for the other things we need to build - wonders, galleys, buildings, etc. See bcool's post in this thread referring to a strategy Duckweed employed in the Christmas ice-fest BOTM.

Well, if the capital stays on Elba, then the second city is really useless long term if we want the cap to be working lots of tiles. Even if the cap moves and we GP farm Elba, city 2 is still pretty useless, but it might make a catapult farm.

Sure. Early on it makes workboats and warriors working its GHmine and whatever else is going. Later, galleys or cats or whatever. Its role is to make sure we get best value for our early tile improvements. Its worst-case scenario is to sit at size 3, working the GHmine, a coast and the SE grassland island. If we add a lighthouse, it can alternate the coast and the PHmine.

One tech from oracle can be nice, but maybe a nice tech trade is almost as good? Early representation, especially with our food rich Elba, seems like a very useful path.

OK, but the facts of life seem to be that we're going to struggle to get a wonder out "early". So far, T155 is the best we've done a Pyramids, and T97+T170 the best for Oracle+Pyramids.

Final thought ... no one even groaned, (much less laughed) at my Elba room joke?

I saw it

 

[mabraham]

So if every 15 turns for the unhappy to wear off is the rate limiting step, then maybe the fastest thing to do is to whip more hammers in each 15 turn cycle? Which means whip more pop in each 15 turn cycle (not just two, but perhaps three ... although maybe just works in cap where happy cap is 5? Getting a hapcap 4 city to pop 6 could be a slog.).

Fortunately, we're Charismatic and can build a monument in the capital to be happy at 6 with a whip- already.

Alternatively, if we think it has enough accelleration potential, two whips in 15 turn may be the way to go (at least once), then cycling from -1 to -2 whip unhappy instead of 0 to -1 whip unhappy.

That definitely has merit for the final cycle or two of a wonder.

 

[bcool]

Yes for some reason I can't open mabraham's zip files either. I get the same invalid error Ron seems to be getting.

edit: I can't seem to open the save file mabraham posted after he finished the pyramids either. Civ IV crashes while loading it.

 

[mabraham]

You don't really get more hammers for the wonder if you pop more population if you are using the overflow for the wonder. You get a maximum of 44 hammers in OF from the whip regardless of how much pop you whip. Of course you could whip something larger with more pop, but there really isn't anything bigger than a settler that we might want to whip early on. And whipping the wonder directly imposes a penalty, so that is usually not the best way to go.

Agreed.

If I remember correctly the happiness with a monument is 7 in the capital. Especially if we whip 3 pop settlers than the unhappiness we accumulate probably won't be too bad. But even if we do accumulate some unhappiness because we are whipping faster than 15 turns, the extra production is worth it if we are producing something big.

Actually, my experience with running the 6/3 settler-Pyramids cycle described above was that starting with a full food box, we would grow . Putting the spare food on a worker would be OK if we're not building a wonder, but the timing is bad and those rot before we get the worker out. The second city comes into its own here - now we can usefully share a food tile so that our empire works all its good tiles with every person happy.

If we are isolated with just small islands I think we should try to get all 3 big wonders...

Wow, ambitious. It is feasible for SIP to Oracle MC, and the 3E city to then work corn+clams+2 mines to build a forge in the 25-turn GP window (e.g. pre-build granary there, grow to size 4, give clams back to the capital, work at most 12 turns returning 8 and whip 2 population for 180). Ideally, we can start working the engineer about T115, to pop in T165. Meanwhile the capital has regrown a bit and starts a settler/forge cycle as best it can, moving to full speed from about T115. Pyramids get popped around T165. At around 7/turn sustainable running a 6/3 settler-forge cycle in the capital, the forge will finish around T135. Then a 6/3 settler-Colossus cycle for a 375 will finish about T185. We'll have put out 5 more settlers from the capital by then, so if all goes to plan, we'll have around 7 cities, Pyramids and Colossus at about 300AD.

My concern with this is that Representation and Colossus do not stack all that well. We'll get 20 or so extra /turn from working the seafood across our empire. Our capital will run 7 scientists for much of the next 50 turns. Our side cities should mostly run merchants to pay for that. While they're growing to their :happy:-cap, new population will work coastal tiles consuming 2 to produce 23 each. Once they're at their :happy:-cap then working merchants consuming 2 for 333 is best. So the combination of the two is good, but is it good enough to justify the cost of the Colossus?

 

[Grifftavian]

...

If I remember correctly the happiness with a monument is 7 in the capital. ...

I checked one of my test game saves, and this is correct; Capitol City with Monument has a happy cap of 7. It's due to Napoleon's Charismatic Trait, we get an extra +1 happiness/city, and +1 happiness from Monuments & Broadcast Towers. That may make Mysticism a higher priority tech than it normally would be, especially if we'll be doing a lot of whipping. Charismatic also give us a reduction of 25% Experience Points needed for Unit Promotions, so fighting Barbs early on is a good way to build up some veteran military units.

Sorry about not re-posting my test map from the SG11 team thread, but the forum will not let me post the same attachment twice. Attached is Version 3 of the same map, but I added a school of Red Snapper in the fog (which the Warrior will find when he moves to the FPH). And before anyone asks, in Version 1, I forgot to give away the Great Lighthouse to an AI Civ, so it's not really usable for test scenarios.

FYI, for these Test Maps, I used the Big and Small map script, with the Tiny Islands & Island Regions Separate settings. Our starting location is in the separate tiny islands region, and is therefore a fairly isolated start.

EDIT: cross-post with R1, mabraham x3, & bc!

 

[mabraham]

Yes for some reason I can't open mabraham's zip files either. I get the same invalid error Ron seems to be getting.

edit: I can't seem to open the save file mabraham posted after he finished the pyramids either. Civ IV crashes while loading it.

Hmm weird. I've re-uploaded everything making sure I use FTP binary mode. Maybe that's it. I've broken the saved games out of the .zip archive too. You can browse the directory here., or try the old links.

 

[Grifftavian]

...
Can you re-post your test games please, I have downloaded them twice and get an error message saying the file is invalid.

I had trouble getting them as well. The Windows 7 file extractor didn't like it, and neither did the 7-Zip program. I downloaded a trial version of a program called winrar, and it was able to extract two maps, but said the third had a bad archive, or something to that effect.

EDIT: another cross-post!

 

[bcool]

Thanks, I now can open the 3 test games, but the pyramids finished test game still crashes Civ4.

 

[bcool]

Wow, ambitious. It is feasible for SIP to Oracle MC, and the 3E city to then work corn+clams+2 mines to build a forge in the 25-turn GP window (e.g. pre-build granary there, grow to size 4, give clams back to the capital, work at most 12 turns returning 8 and whip 2 population for 180). Ideally, we can start working the engineer about T115, to pop in T165. Meanwhile the capital has regrown a bit and starts a settler/forge cycle as best it can, moving to full speed from about T115. Pyramids get popped around T165. At around 7/turn sustainable running a 6/3 settler-forge cycle in the capital, the forge will finish around T135. Then a 6/3 settler-Colossus cycle for a 375 will finish about T185. We'll have put out 5 more settlers from the capital by then, so if all goes to plan, we'll have around 7 cities, Pyramids and Colossus at about 300AD.

My concern with this is that Representation and Colossus do not stack all that well. We'll get 20 or so extra /turn from working the seafood across our empire. Our capital will run 7 scientists for much of the next 50 turns. Our side cities should mostly run merchants to pay for that. While they're growing to their :happy:-cap, new population will work coastal tiles consuming 2 to produce 23 each. Once they're at their :happy:-cap then working merchants consuming 2for 333 is best. So the combination of the two is good, but is it good enough to justify the cost of the Colossus?

These are good questions and a good point that they don't stack super well. And if we need astronomy to reach the AI then the Colossus is even less attractive since it would obsolete as soon as we can beeline Astronomy. As we learn more I think we can decide if the Colossus is worth it. Also we can probably do a better job at estimating its value in game.

The Temple of Artemis is another wonder to at least consider. Without the Great Lighthouse it does lose some of its limited appeal. But it is an early wonder that really does increase the great people production. And it does become more attractive with pyramids and representation.

Maybe we will have Copper or Marble somewhere nearby to help us decide if Temple of Artemis or the Colossus is something to shoot for or not.

 

[Grifftavian]

Hmm weird. I've re-uploaded everything making sure I use FTP binary mode. Maybe that's it. I've broken the saved games out of the .zip archive too. You can browse the directory here., or try the old links.

OK, I got your thrid test game from the directory link. Tried re-downloading the .zip file; the 7-Zip program gave me some sort of error about an archive. The WinRAR program gives me this...

! C:\Users\Tom\Downloads\sgotm13 test games.zip: The archive is corrupt
! C:\Users\Tom\Downloads\sgotm13 test games.zip: The archive is corrupt
! C:\Users\Tom\Downloads\sgotm13 test games.zip: CRC failed in third sgotm13 test.CivBeyondSwordSave. The file is corrupt
! C:\Users\Tom\Downloads\sgotm13 test games.zip: The archive is corrupt

 

[bcool]

I tested a few variations on the oracle build but I get similar results as mabraham did here:
http://forums.civfanatics.com/showpost.php?p=10238229&postcount=38

So yeah the Oracle---> metal casting, build forge pop GE for pyramids plan is a bit slow if the 2nd city has no food resource for its own. You really can't get that 2nd city up and running anywhere close to 8 pop before the oracle comes along. You can get the forge into the window and beat the prophet from Paris, but still it does seem a bit weak.

If the 2nd city has no food resource than I think a pyramids build with OF from 4-5 settlers is probably a better start. We need to know more to make more definite plans.

I would like to get a look at what mabraham did with the pyramid plan if anyone can open that save let me know.

attached a save nothing special T97 Oracle with 2nd city that could grow and get a forge before window on getting a GE 1st closes.

 

[mabraham]

I would like to get a look at what mabraham did with the pyramid plan if anyone can open that save let me know.

I re-downloaded it and it was broken. The version I was uploading is fine, however. I renamed it with a simpler name, uploaded, downloaded and it seems fine now. Will update old link.

 

[mabraham]

I tried for three wonders after SIP. I tried the effect of getting a granary before the wonders, on the theory that this would mean I could share two of the food resources to the 3E city. In practice that means the capital needs three of the food resources for just a few turns per cycle, and for the rest, just two. The +25% from the forge meant that a 3-pop whip on the settler comes after about two turns of go on it.

Oracle->MC 103
Forge Orleans 123
CoL 129 (2 turns late)
Start Colossus T130 (4 overflow in), it got built T153 when I was 240/375 (but the first GFchop just happened). So maybe I could get T165 or so.

Unfortunately the GLH-bearing AI teched Alpha and MC by the time I met him, and I couldn't get Alpha out of him for CoL. Yet another AI beat me to CoL (after all, we are going to pop a prophet some time, and who wants to bulb Theology anyway?). Then the original AI built Colossus in front of me. This was an unlucky run, I feel.

 

[Ronnie1]

So I was reading Duckweeds article again and thinking about our overall strategy and it got me to thinking.

??? What wonders do we REALLY need to get us up and running and off of Elba ASAP?
??? What will allow us to GROW ASAP, and stay financially solvent while we do it?
??? If we need Astronomy, how best to get there ASAP?
??? Domination or Conquest?
??? How far into the Military tree do we need to go?

GLH is not an option of course, so what do we have to work with? I think we need to assume that we won't have stone or marble at least in the very beginning. We know we have seafood and lots of it. This means working coastal tiles almost all the time. I think The Colossus could prove to be the most valuable early wonder especially if we get copper on 1 of those grass hills.

Do we really NEED the Oracle? Could those hammers be better spent on galleys and units as we work our way out into the islands?

Pyramids? We are going to be whipping like mad. We may not have the population to actually run very many specialists, especially early. The Pyramids are only really valuable if you are planning on running a specialist economy, is that our plan? I am not convinced.

I have been convinced after more play testing that settling in place and getting city #2 up ASAP is a better option than settling on the plains hill UNLESS there is room to settle north on the island somewhere.

What about bypassing the religious techs all together, Mysticism possibly excluded because of our trait.

Tech path something like Fishing>Mining>BW>Pottery>Sailing>MC>Writing>CoL>Math....CS

The problem I see is if we do indeed need Astronomy. Once we get CS, we lose the opportunity to bulb our way to Astronomy.