SGOTM 11 - Who Dat?

[Fluroscent]

What matters is how many GPs can be created by building Parth. You are absolutely right that we do not want to get SciMeth faster than neccesary, and of course, the bulbs should be fired after it (and all but 1T from lib) is teched. This is true whether we have GLib/Parth or not, as we are likely to have some monastaries anyway (especially in the capital). But equating beaker counts with game length is flawed, as this would only matter if the bpt would be contstant.

Parthenon is quite expensive at an early point in the game, and while it is surely better than nothing, I find it difficult to see where we can spare the hammers.
We would need to build it in:
-A city that will never itself produce a GP
-A city that has enough hammers to build it quickly
-A city that has no other priorities

I don't see this kind of city often in speed games.

If a candidate city would present itself, it is still likely that:
-It has better use building military to prepare for conquering the needed resources
-It may be the place to build UN, in which case its forests should be spared
-It is better simply building wealth.

By the way, listing the GPP cost of each succesive GP may be a great help in determining how we run our GP farm and how much valure Parth/NEpic/Pacifism etc. have.
Ideally, we would be able to generate (before finishing SciMeth):
-8 Scientists. One for academy + seven to bulb Physics+Electricity along with the Physics scientist.
-1 GE, for UN (2 also works, but we can surely find a city with enough forests to cover the rest in 1T)
-However many needed to fill the quota, although we may be able to generate what we need AFTER Electricity (where no more bulbs are useful).

Getting 8 Scientists is probably going to be unattainable, and every one that needs to be saved for Eldine costs a bunch. So, once we know how good a GP-farm we can build, we should calculate an approximate total of GPP, and then adjust to the closest GPP value that will pop a GP. (Go Go Gadget Spreadsheet)
IF, and only IF, Parthenon can push an additional GS, should we consider sacrificing anything to build it.

Example (WARNING, random numbers with no relation to the actual amounts): Say we need 5000GPP for Scientist nr.6, and 10000GPP for nr.7, and Parthenon can be estimated at a GPP value of 2000, then we should be able to generate less than 10000, but more than 8000 without Parthenon to make it worthwhile. If we can only produce 6000, then the Parthenon will only get us the last GS faster, but not an additional one. Ideally we pop the last GS the turn we complete SciMeth.

 

[Fluroscent]

@building UN
We should definitely build it ourselves, but we may want to build and then gift the city.

 

[beestar]

Hi Fluro,

What matters is how many GPs can be created by building Parth. You are absolutely right that we do not want to get SciMeth faster than neccesary, and of course, the bulbs should be fired after it (and all but 1T from lib) is teched. This is true whether we have GLib/Parth or not, as we are likely to have some monastaries anyway (especially in the capital). But equating beaker counts with game length is flawed, as this would only matter if the bpt would be contstant.

Beaker counts are a good measurement in the context of GP bulbs, which are a constant number of beakers no matter whether early or late game. I've been valuing GP points in terms of their bulb benefit.

I do of course agree that later beakers are easier to come by than earlier beakers.

- A city that will never itself produce a GP

I'm not worried about polluting the GP pool by building the Parthenon. If we produce a surplus Great Artist or two, they can be one of the Eldine greeters.

Actually based on my last post, it should be GLib that we would consider building rather than Parthenon. It's cheaper, produces beakers, and generates 6 GPP/turn * multipliers, which should easily exceed Parthenon's 10 GPP/turn in the midgame.

By the way, listing the GPP cost of each succesive GP may be a great help in determining how we run our GP farm and how much valure Parth/NEpic/Pacifism etc. have.

Yep, that exact calculation is in the Google Docs spreadsheet. The Parthenon calculation in my previous post is based on those figures. Anybody who doesn't have access (or would prefer it to be reposted to the forum), let me know.

 

[Fluroscent]

Maybe a bit of misunderstanding from the (over)use of the terms GPs (Great People) and GPP(Great People Points)
Maybe I am just blind, but I do not see the number of GPPs required in the spreadsheet. And I still maintain that only a whole number of GP will work .

I did mistype once, and I meant that the relevant figure concerning Parth (but partly GLib too, although it has pure beaker value to be added) was GPPs, not GPs. And GPPs cannot (at the moment) be directly converted to a beaker value, since:
1. GPPs are only of value if they produce a GP and
2. GPs cost succesively more GPPs, which means that the beaker/GPP value depends on the total number of GPs born, with more meaning a lower beaker/GPP value.

Once we can estimate how many GPPs we can produce by the time we hit SciMeth, we can make a proper calculation.

About Parth vs GLib. Glib is harder to get if we wait and trade for Literature than getting Parth if we trade for Aesthetics. Other than that, Glib is the clear better build of course. But my gut still tells me that shooting straight for CS->Edu after we have our economy techs is superior in this game.

Also, GPP pollution should not be taken lightly. A GPfarm with both NE and Parth will IMO have an unacceptable risk of generating multiple GAs. One fluke from NEpic is fine, given that we proably need one "regular" GP for Eldine anyway, but more and it begins to hurt.
Of course, if Lib->Radio is not possible, then GA's can be used to help bulb there, but at a lower beaker return of course.

Which makes me want to make another point entirely - We really want to delay the AI in getting liberalism!
-This is best done by getting Philo first, and not trading it, and of course holding on to Paper-Edu as long as possible.
-This means that we should probably bulb philo with the 2nd GS popped (the cost of which should be alleviated by the extra monastary in the capital).
-If there are AI with Free Religion favourite civic, here is another reason to beat them down.
-Other than letting us get Radio from Liberalism, this will also make it less likely that AI's will begin using Free Religion.

By the way, I know that Asoka has Free Religion as a favourite civic, but he is obviously not in the game (us being India and all). How many other AI have that fav civic?

A final observation, it seems that Fifth Element took an early religion from the culture graphs. Everything else looks normal. I wonder if they choose to delay the worker too?

 

[beestar]

Sheet "GPP Calculator", in the top section "Required GPP"

First GP: 150
Second GP: 450 (i.e. 300 more points)
Third GP: 900 (i.e. 450 more points)
4th 1500
5th 2250
6th 3150
7th 4200
8th 5400
9th 6750
10th 8250
11th 9900

The bottom half of that sheet shows that 57 Great Person Points per turn can be generated with National Epic and PHI multipliers only, 5 specialists, and 4 Great Person Points per turn from miscellaneous wonders.

 

[Fluroscent]

Ah, ok I didn't see the additional sheets . Great work with these spreadsheets beestar .

Still some conceptual pitfalls to be wary of, like the GPP "integer" factor and the implied value decrease when generating GPs from more than one city .
We might want to input a more general formula that expresses the GPP yields as a function of base GPP, so that we can modify it as our GP options change. We certainly cannot be sure to even find a site that can support 5 scientists (certainly not with the land revealed so far ), so we may want to squeeze out GPP from various sites instead if we time it just right.
We can certainly use this to figure out how quickly we need to generate a GE before the 3 or 5 base GPP/turn becomes too little. I'll try to fiddle around with it tomorrow.

I can't help feeling like good GP management could be one of the most important parts of a strong finish here.
At some point, we need to make an educated guess at when we are going to arrive at key moments (Scimeth in particular), preferably accurate to within a handful of turns. The value of this information is evident when studying some of the top teams performances in recent games.
So much depends on how good a GPfarm site we can find though...

 

[orb]

OK. - To draw us back to the next turn.

We all seem to like the idea of the stone (especially if there is extra food) and the assistance it gives us with Mids and soon after HG. - So I think we should plan towards settling the stone resource or similar depending on what the Delhi Warrior reveals. - The Warrior will be complete in 3T and can be on the stone in 8T.

The NE warrior should continue to explore (safely)

The worker will be busy farming and mining up to BW becoming available then switch to settler chop. -OPTION - Chop 2nd worker first for faster chop and/or road to the northern cow block site.

Also BW will currently be complete in 9, but as Delhi grows this will change based on the FFM plan (F1 complete).- Not sure on the numbers here.

Delhi grows to size 2 next turn and works the corn for extra growth, then the silver mine when it becomes available.

Position of copper and Zara's REXing are the 2 things that could alter our plans so I propose I play up until BW complete or Zara settles another site (which ever is first) then post and review.

Any comments? - I can play Saturday night if this gives folk enough time to comment?

 

[pnp_dredd]

The main optimisation that can be done for this turnset is worker moves. Mine first or irrigate corn first? The worker can be on the silver next turn, and mining silver first will save us perhaps 1 worker move in the long term.

But what's really going to matter is when Delhi grows.

Is it worth running both scenarios in a test save and seeing how they work out with chopping/whipping?

 

[cas]

Not that it matters for a little while...but you don't think we should use a GSci to bulb part of Edu to get Oxford faster ? I usually bulb one or two GS pre SciMeth when I have a heavy Buero/cottage capital.

And try to leave a few forests around the 6 cities where I plan to build Universities so I can get those up asap too.

I think faster/well-planned Oxford build would make up for using the GSci pre SciMeth ?

cas

 

[pnp_dredd]

Note: this is really detailed. I've put my favorites into post #117, so look there first, and only read this post if you care about other options.

I've tested some options. I've run the tests until the three squares (2 corn, 1 silver) are improved [and continued some until I get first settle]. Delhi is building warriors, working unimproved corn at size 2. I've teched wheel-pottery next but this isn't necessarily what we need to do.

[Option 1: Mine first]
3175 complete mine. Delhi can work mine this turn and still grow to size 3 next turn.

[option 1a: mine first, chop next, then irrigate]
3150 Delhi grows to size 3, begins worker2
worker1 moves north to riverside plains forest to chop
3125 BW in, begin wheel-pottery [not necessarily the optimal path]. No switch to slavery just yet.
3025 chop complete
3000 worker2 complete (with 0 overflow). Both workers to irrigate corn.

[Option 1aa: mine first, chop next, then irrigate, grow to size 4]
2925 Corn irrigated. Wheel complete, 9 beakers into pottery. Have 2 workers, 3 warriors and 4 hammers overflow. Delhi size 3 with 18/39 food. Not in slavery.
[gets a settler in 2750 with 2 forest chops and a whip, 55 overflow!, in slavery, Delhi size 2 and 3/36] [can create another worker in 3 turns]

[Option 1ab: mine first, chop next, then irrigate, settler at size 3]
3000 switch to settler
2925 Corn irrigated. Wheel complete, 9 beakers into pottery. Have 2 workers, 2 warriors, 1 warrior 11/22 and settler 30/149. Delhi size 3 with 3/39 food. Not in slavery.
[gets a settler in 2775 with 2 forest chops, 19 overflow, not in slavery, Delhi size 3 and 3/39]​

[option 1b: mine first, irrigate next]
3150 Delhi grows to size 3, continues to build warrior
worker1 moves towards corn
3075 Delhi completes warrior, begins warrior (planning to grow to size 4 for worker - may not be optimal)
3000 mm so that warrior/growth happens at the same turn
2975 complete warrior, begin worker2 (size 4)

[Option 1ba: mine first, irrigate next, slavery]
switch to slavery
2950 hurry worker with 17 hammers overflow.
2925 Corn irrigated. Wheel 137/139. Have 2 workers, 4 warriors and 23 hammers overflow. Delhi size 2 with 1/36 food. In slavery.

[Option 1bb: mine first, irrigate next, no slavery]
Don't switch to slavery. This will be too slow.​

[Option 2: irrigate corn first]
worker moves to irrigate corn

[option 2a: irrigate corn, worker @ size 2]
3300 grow to size 2, build worker.
3100 complete BW, begin writing. Corn irrigated.

[option 2aa: irrigate corn, worker size 2, chop settler size 2]
3100 worker moves to chop riverside plains forest
3050 complete worker2, begin settler. worker2 to chop
3025 chop forest1, begin chopping forest2
2950 chop complete, Move to mine silver
2875 mine complete. Settler complete. Wheel 114/139. Have 1 settler, 2 workers, 1 warrior, 1 warrior 18/22 and 4 hammers overflow. Delhi size 2 with 0/36 food. Not in slavery.

[option 2ab: irrigate corn, worker size 2, mine next]
3100 worker moves to chop riverside plains forest (1 turn)
3075 worker moves to mine
3050 complete worker2, moves to mine
3025 complete warrior, begin warrior
2975 complete mine. Wheel 66/139. Have 2 workers, 2 warriors, 1 warriors 11/22. Delhi size 2 with 30/36 food (growth next turn). Not in slavery.​

[option 2b: irrigate corn, chop next then mine, grow, slavery]
3250 complete warrior, begin warrior
3150 Delhi grows to size 3, works riverside grassland
3100 irrigation complete, BW complete. Worker to chop riverside plains forest.
3050 Delhi to size 4. Switch to slavery, begin worker (work 2xgrassland forests)
3025 rush worker

[option 2ba: irrigate corn, chop next then mine, grow, slavery, warriors]
3000 continue building warrior. worker2 to mine. worker1 completes chop then moves to mine
2900 complete mine. Wheel 106/139. Have 2 workers, 4 warriors, 1 warriors 6/22. Delhi size 3 with 8/39 food. In slavery.

[option 2bb: irrigate corn, chop twice then mine, grow, slavery, chop settler]
3000 switch to settler. worker2 to chop. worker1 completes chop then moves to chop
2925 chops complete, move to mine
2850 mine complete, settler complete. Wheel 130/139. Have 1 settler, 2 workers, 2 warriors, 1 warriors 16/22 and 9 overflow. Delhi size 2 with 4/36 food. In slavery.​

[Option 3a: irrigate corn, chop worker, mine, no slavery]
move worker to corn
3150 Delhi grows to size 3, begins worker.
3100 complete farm, move to chop riverside plains forest. Complete BW.
2975 chop complete, move to mine. worker2 complete (28 overflow...), move to mine. Begin warrior. [could put overflow into settler?]
2900 complete mine. Wheel 108/139. Have 2 workers, 4 warriors, 1 warriors 4/22 and 9 overflow. Delhi size 3 with 30/39 food. Not in slavery.

[Option 3b: irrigate corn, chop worker, chop settler, mine, no slavery]
move worker to corn
3150 Delhi grows to size 3, begins worker.
3100 complete farm and BW. Worker to chop plains grassland forest
2975 Complete worker2. both workers to chop forests
2850 complete settler, workers move to mine silver.
2800 (turn 48) complete mine. Pottery 17/186. Have 2 workers, 2 warriors, 1 warriors 20/22. Delhi size 3 with 22/39 food. Not in slavery.


Attached save is the test save that I'm using, based of Fluro's initial test save.

 

[pnp_dredd]

Notes:
Mine first gives a crapload more commerce. Mine first gains one worker turn:
mine-chop-irrigate (option 1a) completes the three improvements on turn 43, and the chop goes towards completing the worker with 0 overflow.
irrigate-chop-mine (option 3) completes the three improvements on turn 44, and the chop goes mainly towards warriors.

Option 1a is strong:
2925 (Turn 43): 3 improvements. Wheel complete, 9 beakers into pottery. Have 2 workers, 3 warriors and 4 hammers overflow. Delhi size 3 with 18/39 food. Not in slavery.

1aa and 1ab only differ after this point - 1aa gives a faster 3rd worker, while 1ab gives a faster settler but hasn't switched to slavery. 1ab has the option to use farms and grow to size 6 before the first whip.

Irrigate-first gives a faster settler (turn 45 rather than turn 49 for mine-first) at the cost of research:

option 2aa: irrigate corn, worker size 2, chop settler size 2, then mine
2875 (Turn 45): Wheel 114/139. Have 1 settler, 2 workers, 1 warrior, 1 warrior 18/22 and 4 hammers overflow. Delhi size 2 with 0/36 food. Not in slavery.

Option 2bb (irrigate first, grow to size 4 then whip/chop settler) makes the settler one turn later but is otherwise superior:
2850 (Turn 46): Wheel 130/139. Have 1 settler, 2 workers, 2 warriors, 1 warriors 16/22 and 9 overflow. Delhi size 2 with 4/36 food. In slavery.

 

[pnp_dredd]

Hmm, that all looks pretty incomprehensible. It's only going to make sense if you go in game and see what actually happens.

I've chopped a 2nd worker in every scenario. With BW this just seems to make sense. Please feel free to run tests with 1 worker.

The important decisions are: mine first or farm first? When to make a settler? When to switch to slavery?

We also need to decide what to tech. Wheel is useful but Pottery may not be required early on, because farms in Delhi will be strong anyway under the whip. AH may be better if we are going for an early settler near the cows. Delhi wants to build a library rather than too many warriors, so writing next, then PH because we'll need to begin the oracle.

That would mean wheel-AH-writing-poly-priesthood?

If we hold off on Writing then Delhi is going to produce a crapload of warriors (5 base hammers/turn from PH/city and silver)

 

[Fluroscent]

I think the best results come from irrigating first, then mine, then switch to settler at size3 immediately, and afterwards build warriors while sinking chops into workers/settlers.

Tech wise, if we want to settle the blocking city to the NE, then we should tech AH next. The cows in the inner ring will help that city greatly. AH teched just after BW, and running the farm-mine-size3settler scheme, will be completed 3-4 turns after the settler. Just right to start the pasture right away (need to divert the worker correctly!)
We also need to test how quickly we can build Oracle there depending on when we tech PH. Of course, let's not get ahead of ourselves if we chose another site. But we may have to make a choice when BW is complete.

So, my suggestion (to orb): Play a part turnset as proposed - Until BW is in, then reevaluate, but more to make sure we choose the correct next tech. All attention on the warrior scouting. We NEED to know the layout of the NE coastline.

My immediate priority for the first settler: Blocking Zara > Able to build Oracle in time > Bronze/Horses > Stone.

@cas, agree, but we may not have the option since we should bulb Philo and generate a GE around that time.

 

[beestar]

In general I would suggest warriors rather than library, to fogbust and scout. Fogbusting now is much better than facing a barb invasion later and having to whip a warrior. Ice and tundra to the south tends to generate barb as well.

Part turnset sounds like a good idea.

Otherwise I'm not a great micro person, and I leave it in the capable if microscopic hands of others

Part turnset to BW is good.

 

[Fluroscent]

The 2nd worker chop may be the way to go after all. I'll have to try it out. Maybe we should focus on determining whether to mine or irrigate first, as that will let orb start the turnset.
It is basically a science vs rex question. I find that the science bonus is great, but as we seem to be able to get the worker tech when needed anyway, the science bonus is strictly beneficial to speed up writing/PH for quicker library/Oracle. Rex seems like a more prudent choice at this point (but I think we should try to work all three good tiles asap and constantly)

 

[pnp_dredd]

I think the best results come from irrigating first, then mine, then switch to settler at size3 immediately, and afterwards build warriors while sinking chops into workers/settlers.

I just added 3b which attempts this.

It gets a settler at 2850 (turn 46) but the mine isn't finished until 2800 (turn 48). In terms of tech, 3b has around the same tech at turn 46 that 1a has at turn 43 (and 1a has 6 more commerce/turn from turn 43-46).

 

[pnp_dredd]

I prefer mining the silver first. Here's my favorite for each option:

Option 1: Mine first
3175 complete mine. Delhi can work mine this turn and still grow to size 3 next turn.
3150 Delhi grows to size 3, begins worker2
worker1 moves north to riverside plains forest to chop
3125 BW in, begin wheel-pottery [not necessarily the optimal path]. No switch to slavery just yet.
3025 chop complete
3000 worker2 complete (with 0 overflow), switch to settler. Both workers to irrigate corn.
2925 Corn irrigated. Chop 2 forests.
2775 Settler complete. Worker free to move with settler.

2750 (turn 50) stats: 165/186 pottery. 1 settler, 2 workers, 3 warriors, 18 overflow. Delhi size 3 and 11/39. Not in slavery.

Option 2: farm first
3150 Delhi grows to size 3, begins worker.
3100 complete farm and BW. Worker to chop plains grassland forest
2975 Complete worker2. Chop 2 forests (1W then 1S of Delhi)
2850 complete settler, workers move to mine silver [no worker available to send with settler]
2800 (turn 48) complete mine.

2750 (turn 50) stats: 61/186 pottery. 1 settler, 2 workers, 3 warriors, 8/22 warrior4. Delhi size 3 and 38/39. Not in slavery.

Notes:
Both of these are a bit light on early warriors. They only have 2 warriors "out there" who must both explore and protect the settler/new city site. Both options will create a pile of warriors after they finish the first settler.
Both build worker-settler at size 3
Both chop 3 forests.
Neither uses Slavery
One difference I just noticed is that the workers in option1 do mine/chop1/farm/chop2 while in option2 they do farm/chop1/chop2/mine.

Mine-first can do wheel-AH and still have AH ready when we settle city2. [but there's no real reason to do so - AH first FTW]
Farm-first requires AH first, and basically we are behind one tech.

 

[Fluroscent]

That (3b) was one I wanted to test. I am beginning to lean towards the mine first approach too. With the worker chop, it seems slightly better than farm first. Without worker chop, the farm first wins. I guess that stems from the added production from the extra chop that offsets the tile yield difference and makes the commerce bonus comparably stronger in that case.

But a few notes though:
Slavery is free, being that we are SPI, so I don't know why this is emphasized.
Chopped forests are also important in the final picture when comparing the methods. Chopping two rather than one is a disadvantage.

Additional thought: We could consider delaying pottery-writing in favor of (AH)-medi-PH, and use the Oracle to pop borders in the 2nd city. We could spare the monument, and ensure a safer Oracle date(and of course, faster religion bonus -> more autospread, more blocking options and quicker friendship bonus with Zara)
Downside is that wheel/pottery will have to wait until after writing (needs to be done before Oracle complete obviously), but I think that is acceptable given that the workers will be busy chopping for a while.

 

[pnp_dredd]

bugger, I forgot that we were spiritual. However if we whip 3->2 then we aren't working a nice tile while we regrow. Whipping 4-2 has the same problem but delays everything as it takes too long to get to size 4.

Lemme play a bit further and I'll see how many warriors we end up with. I'll worldbuilder the second city site too.

Also: make sure that we don't do anything before cas chimes in - he was the one who got us earliest Oracle in sgotm9.

 

[pnp_dredd]

OK here's a strong option:
mine first, worker/settler at size 3, tech AH-medi-PH, city2 between cows/oasis, first whip at size 5.
After about 2700 the unknowns start adding up.

3175 complete mine. Delhi can work mine this turn and still grow to size 3 next turn.
3150 Delhi grows to size 3, begins worker2
worker1 moves north to riverside plains forest to chop
3125 BW in, begin AH. Switch to slavery.
3025 chop complete
3000 worker2 complete (with 0 overflow), switch to settler. Both workers to irrigate corn.
2850 complete AH, begin meditation
2925 Corn irrigated. Chop 2 forests (N and W of Delhi)
2775 Settler complete. Both workers farm riverside grassland, one follows settler after 1 turn. build warriors
2675 settle city2. begin warrior. pasture cows.
2650 Delhi grows to size 4, works farmed riverside grassland. worker makes another farm for Delhi
2625 complete meditation, begin PH.
2525 Delhi grows to size 5, begin worker. [this is extrapolating too far - this is probably a settler, or build worker/whip settler]
2500 whip worker (5->3 with 28 overflow). city2 completes pasture, begins mining forested plains hill. All workers to help City2
2425 (Turn 63) Complete PH. Writing in 13 turns. city2 begins Oracle. Have built 8 warriors and Delhi will building more as it grows.


Here's a test save with the cows site worldbuildered.