SGOTM 13 - Gypsy Kings

bcool · Mar 1, 2011

I played out a test game going for the pyramids with whip OF from settlers rather than the early GE.

I can whip the pyramids at the start of Turn 151 (going down to 2 pop) or finish it in 4 turns without the whip. Might be able to do a little bit better.

quick stats
7 cities 15 pop
No forge so Colossus is ways off (could have prioritized the forge instead of the 6th settler however)
1 problem perhaps is we likely get a Great Prophet first (I ended up saving him could have settled it in the capital for slightly better result perhaps, under representation he isn't much worse than an academy at least early on. I did mix the gpp pool so it was about 30% GS)

see screen shot for other buildings

Spoiler :

It is a bit hard to compare to mabraham's triple wonder games, but initial feeling is it might be a bit better. I should have gone for the Colossus too for better comparison. And I bet mabraham can improve on the results

mabraham · Mar 2, 2011

I tried a game on bcool's map where I did not attempt either Oracle or Pyramids. Instead, I played the start Duckweeding settlers and workers onto infrastructure. That worked fine, but I expanded way too fast - I had too much food for what I could use. I did a fair amount of exploring, but deliberately avoided meeting the AIs, to get a feel for the worst-case scenario.

Stats at T151: 8 cities, 22 pop, total :commerce:

= 55,

at 100% = 57, :gold:

at 100% = 17 (so sustainable 30% slider). I'd built a Maoi and one-third of the Colossus. Teched MC and that's about it. I could have had Currency by now instead, but that's a whole 16 :commerce:

, and the Colossus will be worth more than that. There were 5 cities at size 1 that so far are working one netted clams, one marble and crap tiles between them (while waiting for border pops). They'll be OK-to-good once they come on-line, but it feels like a long time before the machine will really roll. :yuck:

and

caps were really putting a crimp on things in the two cities that were really working well.

I was forced into an early Maoi Statues just to have something useful to build in the capital. Since then, the capital is really only working a few extra coast tiles per turn, which is not a great rate of return, since I still have to whip to hold down the food. I haven't had a chance to get to Monarchy for some MP-

, which is where Bureaucracy/Moai/Colossus starts to shine. That machine requires Civil Service, MC and Monarchy and a pile of :hammers:

before it all gets together, and the empire needs Currency and CoL to sustain itself. If we're isolated, that's a scary amount of :science:

to chew through. Maybe I need to aim at an early library in the capital, and not whip it much. Still not great returns, though. I could have aimed at GLib, which I would probably be starting about now, but what would I have been spending my :hammers:

on in the meantime?

Rep (for

and a good reason to run specialists), Caste (for artists for quick border pops in new cities, and Courthouses, and Clam Chowder scientist-spam) from some kind of Oracle->MC, Pyramids-while-teching-CoL play would have made this position heaps better. You grow slower in those scenarios, so you get more time to put necessary infrastructure tech on the board. Also, you don't need to tech Monarchy and Civil Service before the machine can move.

Take-home lesson

If we're isolated, we absolutely need an economic plan that keeps up with our growth. If we're not isolated, trading with the AIs might buy us enough time for the above position to come good, but we still have about 50 turns of teching before that.

mabraham · Mar 2, 2011

mabraham said:
MM thoughts

Food is normally king, but hammers that increase food later can be better...

Gory details

T0-10 we work 3 corn to grow to size 2. This has put 1 on the WB. Then four plans come to mind:

A) T11 we work PFH+GF to get WB in 9 turns, switch GF to nets to get WB in 11 turns (no overflow). If we then go for a fast settler, we'll start a third WB, and switch to settler at size 4. This can be whipped T47.

C) As for A, but if we then go for a third WB, we'll work two nets at size 2, then two nets+PFH, then one turn of nets+PFH+2GF, then start settler with one turn of 2 nets+PFH+GF, then three nets+something (gives up a turn on a net to get WB a turn earlier and thus another net turn back, plus earlier start on settler). Settler whippable T51 (on the nose).

OK if the plan becomes "3WB, then settler", then I can improve on C - call this C'. It gains 2 :commerce:

and moves 2 surplus :food:

from the

box into the :hammers:

box, by planting the second WB late!

Gory details
Basically, you move the second WB to the SW clam, giving up :food:

-profit from working a net early, still growing to 3 at the same time, not needing to do any fancy footwork during the transition growing to 4 to get the third WB out in time. This earns back the turn working a net. In C' you end up working an extra turn of nets and PFH, growing 1 turn later (so eating 2 :food:

less when we have no good fourth tile to work), and not working grassland forest 3 times. This breaks even on net food and hammers and picks up two :commerce:

. The (small) gain is improved because the timing is such that we (effectively) overflow 2 :hammers:

onto the settler, rather than 2 :food:

into the food box. That doesn't change the time we can whip the settler for minimum or maximum overflow, but we'll still have our 2 :food:

=>2

conversion in place at the end of the settler. The subsequent regrowth from 2 to 3 will happen on the same turn. So if we build 3WB then settler, I think we'd be silly not to gain 2 :commerce:

and convert 2 pre-granary- :food:

into 2

.

I know, small things, small minds. It may matter next turn set when we will be deciding our build plan and thus when the second WB is planted. It could matter in other build-orders too, I guess.

Here's an update of the table I showed earlier. Numbers are :food:

surplus (i.e. what went into the :food:

box), total :food:

(i.e. including that which we ate, but not :food:

building a settler) / total :hammers:

(including :food:

building a settler, but ignoring whether we've whipped or not) / total raw :commerce:

(including palace). These metrics are invariant over whether we work PFH or unimproved corn when building a settler (for example).

Code:

   T22                T33            T48         Settler whippable?
A  37/103/ 64/202 67/177/102/327 108/324/180/509 T47
B  43/109/ 56/198 65/175/100/319 113/327/170/501 T48
C  37/103/ 67/202 67/177/102/327 111/317/174/511 T51
C' 37/103/ 64/202 63/177/102/327 109/313/176/513 T51
D  37/103/ 64/202 67/177/102/327 ???/148/509 T54 (third WB in place T48 but not planted)

(Small correction to T48 stats for total food, also - buggy measuring.) Here we see that C' produces and eats 2 fewer total food than C, moves 2 food to the hammer box, and gains 2 commerce. Note that growing to size 4 requires 108 :food:

.

mabraham · Mar 2, 2011

bcool said:
I played out a test game going for the pyramids with whip OF from settlers rather than the early GE.

I can whip the pyramids at the start of Turn 151 (going down to 2 pop) or finish it in 4 turns without the whip. Might be able to do a little bit better.

quick stats
7 cities 15 pop
No forge so Colossus is ways off (could have prioritized the forge instead of the 6th settler however)
1 problem perhaps is we likely get a Great Prophet first (I ended up saving him could have settled it in the capital for slightly better result perhaps, under representation he isn't much worse than an academy at least early on. I did mix the gpp pool so it was about 30% GS)

see screen shot for other buildings

Spoiler :

View attachment 284348

It is a bit hard to compare to mabraham's triple wonder games, but initial feeling is it might be a bit better. I should have gone for the Colossus too for better comparison. And I bet mabraham can improve on the results

Yeah it does look better. Key difference was probably building an early granary.

mabraham · Mar 2, 2011

Pre-Play-Plan T10-end of second WB

BUILD QUEUE
Clam Chowder: WB, WB

EXPLORING
I can't. I won't even get a border pop.

UNIT MOVES
None. Clam Chowder isn't large enough to need an MP, we have no defensive requirements, we may as well bust fog on the hill 3E on the off-chance we make an AI contact.

WORKER ACTIONS
N/A

CITY MM
Clam Chowder: Option A, per previous discussion: grow one turn on corn to size 2, then work max hammers (grassland forest + PFH) for 9 turns to finish WB with 1 hammer overflow. Net the NE clam, work nets+PFH to build second WB in 11 turns. Do not plant the second net, in case delaying it is useful for the plan for the second turn set.

DIPLOMATIC ACTIONS
Hopefully none. I'll play nicely with the other children if I have the option.

STOP CRITERIA

Finish building second WB.
Anything weird, especially if we might react to it.

OTHERS
Research: Mining then BW
Civic change: Slavery immediately, but it won't happen this turnset.
Religion change: Definitely not.
Cities settled: None

THINGS TO DO EVERY TURN

Make a saved game
Demographics screenshot

I plan to play in about 40 hours from this post, unless most people say they're happy for me to go earlier.

How about I stop one turn before the second WB is finished and upload that? This might mess with the head of someone on another team :evil:

mabraham · Mar 2, 2011

I was bored at work and went looking for threads on Musketeers... The search feature excerpts about 20 relevant words... I complained about that in SGOTM9, but nobody cared. You'd think they could exclude the current SGOTM threads from the searches.... Tonight, I hit our thread once and Sporks once with "Musketeers". Irritating. I edited away my references, referring instead to "UU musketmen".

mabraham · Mar 2, 2011

While thread-trolling I found the triple wonder idea is not new. One possible side benefit of the Pyramids is the option of running Universal Suffrage to rush courthouses. Unfortunately, even once you put some hammers on the courthouse, you still pay 3 :gold:

/1

. Neglecting the :gp:

and representation- :science:

, that means a merchant in an established city can effectively contribute 1 :hammers:

to an outlying city. So you run merchants for a while under Rep, and then cash in courthouses under US. Or a Great Merchant's trade mission can build a bunch of courthouses. The anarchy would still suck, so it's hard to see this idea working. Or we could use it as the first part of our Golden Age as we switch into war mode, to buy some theatres for access to the Globe Theatre, or barracks for

in draft cities, or whatever seems vital at the time.

Note that if we continue running a specialist economy through the science and war phase, we should be planning to tech Drama to use the slider for

. This is a silly proposition in a cottage economy because each slider position costs you 10% of your :science:

. However running Rep+Caste we're only going to be getting a fraction of our :science:

from the slider. Theatres double the effect, of course, and we'll probably have at least six of those. Colosseums work on each 20%, so not so useful for us. The limiting factor now becomes food. Suppose we've got 12 cities, and four of them are

-limited, with excess :food:

. (Remember, we haven't been whipping out markets and forges for happiness, Rep only works in 5 cities and is only worth 3

...) The average city is only pulling in 8-12 :commerce:

from trade routes and a few tiles (most of which is going into :gold:

to pay the bills), so losing 10% of that :commerce:

to run four extra merchants (12 :gold:

12

) is a clear gain. Obviously you might prefer other specialists given the right conditions. Even in some crappy city where all you can do is run another person working a coast tile for 2 :food:

2

(with lighthouse and not Colossus) you've shown a small profit (and have another draftable dude!). Concentrating the theatres in the high-food cities is best, because now each 10% slider allows two more specialists. Colosseum maybe, and only in the best such cities if the slider will get to 20% or 40%.

If we don't go for the Pyramids (i.e. don't tech Masonry), then the Great Artist from Music can be used to bulb part of Nationalism. That's 1400 :science:

to get back about 2400 :science:

. Free 2400 if we actually want Military Tradition later. That might free up Liberalism for popping Steel or something.

A Great Prophet is useless for bulbing because they go straight down the Mysticism+Masonry line for ages.

Fast Astronomy to steal the Great Lighthouse city in a surgical strike could pay big dividends.

If we plan to run a heavy espionage economy in the lead-up to a war phase, then we may not need to worry (as much) about siege units. Use a spy to cause a city revolt to get rid of culture bonuses, muskets ignore walls and castles => BAM. Perhaps this is only realistic for war-opening "critical strikes" on strategically important cities.

Alternatively, muskets + cuirassiers might be more effective (i.e. faster-moving) than muskets+cannons. That said, the fast-moving UU is great for covering a straggling stack. You can leave some siege to heal up in the last city, and until the AIs get knights or muskets, a fortified musket is fine siege cover as the late-arriving siege move up. The extra movement of the UU allows one or two of them to join in the attack a turn sooner than you would otherwise suppose. The UU is also great sitting in a stack outside a city while the AI straggles in more defenders past the stack (e.g. a coastal city with limited land routes to it). You can pick off all medieval units at acceptable odds, and can move back into the stack after cleaning up the last target.

mabraham · Mar 2, 2011

... under Slavery...

it takes 37 turns for a city to whip out a granary and a courthouse if it has immediate access to two netted clams and several coast tiles. That's thanks to being organized, so that two 2-whips are enough.
it takes 23 turns for city A to whip out just the courthouse.
it takes 55 turns for the same city with a grassland forest, but the nets in the outer ring (so need a border pop), to whip out a monument at size 2, wait for the border pop, 1-whip the granary at size 2 (the accumulated help here), and 2-whip the courthouse at size 4 (with 80 juicy overflow onto lighthouse or something)

Note that all the above ended at size 2.

... under Caste System...

it takes 86 turns for a city with just one grassland tile (onto which we put a workshop in 8 turns) and two nettable clams in the outer ring (so need a border pop) and plenty of coast to build a monument, a granary and a courthouse, ending at size 6, resources permitting.
it takes 57 turns for a city with two grassland tiles (so our single worker builds two workshops in due course) and two nettable clams in the outer ring (so need a border pop) and plenty of coast to build a monument, a granary and a courthouse, ending at size 5.
it takes 55 turns for city 2 to run an artist for four turns for the border pop, and then build granary and courthouse (whether courthouse first, or granary first), ending at size 5. (Monument as well in 8 more turns).
it takes 29 turns for city 2 to run an artist for four turns, grow to 2, then work the workshops as they come online to build just a courthouse, ending at size 2.

How does this help? If we're in a scenario where we might plan to switch to Caste after a whipping period, we need to plan when each city site might get founded, how much the maintenance will cost us until we get a Courthouse up, and how good a post-Caste settlement might have to be in order to be worth the cost. In particular, settling sites with food sources and some land for workshops together with some pre-planned worker support will permit ongoing empire growth under Caste System.

Obviously, all the above examples relied on support from the empire to have workboats available as needed. We should arrange for this to be true. Nothing will crash the economy so badly as a city that has to build its own monument, and then its own workboats before it can start being useful. Minimising the time-to-courthouse for each new city has to be a top priority for the empire, and that means that cities 2-4 will be doing a lot of heavy lifting with workboats, warriors, galleys, triremes and workers, assuming the capital is off building wonders or spamming scientists.

Take-home lesson

With judicious choice of land and suitable planning with worker support, city development times under Caste are comparable with slavery. However, city sites with (say) several seafood and no land need to be settled early to get some infrastructure whipped before the switch. A good supply of workers needs to be in place before switching away if growth is to continue.

Grifftavian · Mar 2, 2011

mabraham said:
Pre-Play-Plan T10-end of second WB

Spoiler :

BUILD QUEUE
Clam Chowder: WB, WB

EXPLORING
I can't. I won't even get a border pop.

UNIT MOVES
None. Clam Chowder isn't large enough to need an MP, we have no defensive requirements, we may as well bust fog on the hill 3E on the off-chance we make an AI contact.

WORKER ACTIONS
N/A

CITY MM
Clam Chowder: Option A, per previous discussion: grow one turn on corn to size 2, then work max hammers (grassland forest + PFH) for 9 turns to finish WB with 1 hammer overflow. Net the NE clam, work nets+PFH to build second WB in 11 turns. Do not plant the second net, in case delaying it is useful for the plan for the second turn set.

DIPLOMATIC ACTIONS
Hopefully none. I'll play nicely with the other children if I have the option.

STOP CRITERIA

Finish building second WB.

Anything weird, especially if we might react to it.

OTHERS
Research: Mining then BW
Civic change: Slavery immediately, but it won't happen this turnset.
Religion change: Definitely not.
Cities settled: None

THINGS TO DO EVERY TURN

Make a saved game

Demographics screenshot

I plan to play in about 40 hours from this post, unless most people say they're happy for me to go earlier.

How about I stop one turn before the second WB is finished and upload that? This might mess with the head of someone on another team

The plan looks good to me. I especially like the messing with the heads of the other teams part! :yup:

bcool · Mar 2, 2011

Take-home lesson

With judicious choice of land and suitable planning with worker support, city development times under Caste are comparable with slavery. However, city sites with (say) several seafood and no land need to be settled early to get some infrastructure whipped before the switch. A good supply of workers needs to be in place before switching away if growth is to continue.

Interesting analysis. 1 Forest chop under mathematics allows a 1 pop of a courthouse. That can significantly accelerate the profitability of a city under slavery or caste system. Or with 1 forest chop under mathematics you can 1 pop a granary (probably better than a courthouse under slavery)

If we do get Music relatively early, we have the option of building culture instead of a monument. Even with the +1 happy from a monument this could be an important option for some of our late settled cities in the last stages of slavery.

Or we could even consider running 10% culture just for the border pops if we get drama early.

edit: PPP looks good

Ronnie1 · Mar 2, 2011

I like the PPP....and the head messing also.

da_Vinci · Mar 2, 2011

mabraham said:
I was bored at work and went looking for threads on Musketeers... The search feature excerpts about 20 relevant words... I complained about that in SGOTM9, but nobody cared. You'd think they could exclude the current SGOTM threads from the searches.... Tonight, I hit our thread once and Sporks once with "Musketeers". Irritating. I edited away my references, referring instead to "UU musketmen".

I think you can restrict searches to specific forums (or fora?) by being in the forum of interest and searching that forum only. It's one way not to get active SGOTM stuff in the results, but it means not seraching the historical SGOTM info as well.

dV

da_Vinci · Mar 2, 2011

Great thinking in the posts above mab, you must get REALLY bored at work! :eek:

PPP looks good to me. I am so used to worker first, that things like finish the growth and have two pop work for the hammers for the WB get right by me.

On broader strategic issues, what do we know about the mechanics of capitulation?

I assume we will want to fight an AI to captulation, then move on to the next?

So here is a "tragedy" that occured in a recent game. I am attacking an AI, with the intent of making her capitulate, so I have an ally when I go to war against the others. I raze one city, capture one, which seems too little to force surrender. And indeed, capitulation is not an option.

Now I go take a second city (a large one of size 10 or so, but AI still has her capital, and a total of 5 cities left (I killed one and took two of 8). I still figure I need the capital to compel surrender, and indeed, capitulation is not an option. The AI is not at war with anyone else.

In the interturn, she impales a stack of about six units on my troops in the just captured city. She's lost a lot of troops in the three cities I have taken, and in field combat during the sieges. Then, and the end of the interturn, the war horns go off, she now refuses to talk, and some other AI also refuses to talk. Look at the status list in the lower right corner: She is vassal to another AI and I am now at war with both! :eek:

A Voluntary Vassalization ... because it made war, not peace (and because I could not bust her free by taking more cities).

This would not be a happy outcome for us in our game I would think, an AI making such a self-protective deal to put a kink in our war. So why did the AI not offer me capitulation, but was willing to vassal to some other AI? Did the troop losses in the interturn put the AI over some threshold of surrender? Is the voluntary vassal threshold of "fear" lower than the capitulation threshold?

Anyone understand these mechanics in detail?

I saved the autosave before that event, so I could re-test it. So far, it played out the same in one re-try (except she vassaled to a differnent master).

I could test out whether the interturn troop losses made the difference by not taking the city in the inital assault, and let AI put those troops in the city, and kill them and take the city in the same turn one turn later ... see if capitulation is an option. Of course, I will have a few extra losses in that case ... maybe that will change the calculus.

I suppose the two AI that became master had to be willing to go to war with me, or they would not have taken on the vassal? Maybe we can manipulate that to prevent such vassalizations.

Be good for us to understand this so we don't have unpleasant surprises.

dV

Ronnie1 · Mar 2, 2011

I assume we will want to fight an AI to caiptulation, then move on to the next?

Options: Vassals are disabled. No random events. No goody huts.

or not

adrianj · Mar 2, 2011

PPP looks good to me :goodjob:

Universal Suffrage is quite a nice idea. I often neglect it, but I can definitely see how useful it would be once we reach the war stage. Combine with nationalism + military tradition and we could have a very fast moving stack of drafted musketeers and bought cuirassiers. What US is not good for is buying courthouses - since AFAIK cash buying ignores production bonuses, eg, from our Organised trait.

I also really like the sound of espionage to knock down walls for our fast moving stack. This ties in neatly with our cheap courthouses which can run spy specialists when the time is right, and Nationalism giving +25% :espionage:

adrianj · Mar 2, 2011

I know a *little bit* about the vassal mechanics.

To capitulate someone you need to a) have more land than them b) a lot of power (something related to the average of all players, yours, them, etc...) and c) inflict some decent losses on them. I think these are all dependent on the AI involved.

To have someone voluntarily vassalise to you, you first need them to like you, ie, there is an XML field for minimum attitude level to consider offering it. On top of this, I think for an AI to accept a vassal they will consider if they would go to war with the vassals enemies. Ie, are they pleased/friendly with them, and how do their power ratings match up. If they're already WHEOOHRN then I *think* they won't go for it either.

Pretty sure there is a forum thread out there with much more detail.

Bottom line: if we're very powerful we need not fear. If our target has no friends we need not fear. If the other AI are busy warring with eachother, then we're probably ok... until we change targets...

Ronnie1 · Mar 2, 2011

da_Vinci said:
snip...

I assume we will want to fight an AI to captulation, then move on to the next?

dV

adrianj said:
snip...

Bottom line: if we're very powerful we need not fear. If our target has no friends we need not fear. If the other AI are busy warring with eachother, then we're probably ok... until we change targets...

Again, sorry boys, but....from the game signup and maintenance threads...

Options: Vassals are disabled. No random events. No goody huts.

We are going this alone!

da_Vinci · Mar 2, 2011

I assume we will want to fight an AI to caiptulation, then move on to the next?

Options: Vassals are disabled. No random events. No goody huts.

Ronnie1 said:
or not

OK, one less thing to have to worry about. :blush:

dV

da_Vinci · Mar 2, 2011

adrianj said:
I know a *little bit* about the vassal mechanics.

To capitulate someone you need to a) have more land than them b) a lot of power (something related to the average of all players, yours, them, etc...) and c) inflict some decent losses on them. I think these are all dependent on the AI involved.

To have someone voluntarily vassalise to you, you first need them to like you, ie, there is an XML field for minimum attitude level to consider offering it. On top of this, I think for an AI to accept a vassal they will consider if they would go to war with the vassals enemies. Ie, are they pleased/friendly with them, and how do their power ratings match up. If they're already WHEOOHRN then I *think* they won't go for it either.

Pretty sure there is a forum thread out there with much more detail.

Bottom line: if we're very powerful we need not fear. If our target has no friends we need not fear. If the other AI are busy warring with eachother, then we're probably ok... until we change targets...

Useful information. What I think is missing:

1. What are the thresholds for when the AI will run in fear to mommy? I can imagine a threshold for feeling threatened, fearful, or weak that is independent of whether anyone qualifies to be a voluntary master or a capitulation master.

2. Is there a different threshold of "fear" for voluntary vassaling vs. capitulation? Does the AI have one set of criteria for offering to voluntarily vassal to a friend (in terms of its fear factor) compared to capitulating to a war enemy? Hmm, might the fear factor be dependent on the leader type (warmonger or peacemonger)?

It would be nice to be able to predict which wars might end with a larger war on one's hands ...

dV

Thorn · Mar 2, 2011

PPP looks okay to me. Are we all supposed to take a demographics screenshot every turn? I'll be asking more questions as this gets rolling I'm sure.

SGOTM 13 - Gypsy Kings

Deity

Attachments

Deity

Deity

Deity

Deity

Deity

Deity

Deity

Emperor

Deity

Deity

Gypsy Prince

Gypsy Prince

Deity

Deity

Deity

Deity

Gypsy Prince

Gypsy Prince

King

Similar threads