SGOTM 13 - One Short Straw

[shyuhe]

Dhoom - is there an up-to-date test save (for just our empire)? I'm not really convinced on the whole currency first plan and want to try playing around with it.

 

[babybluepants]

Current
88bpt -33gpt @ 100%
7bpt +41gpt @ 0%
8.1bpt -7.4gpt per 10% slider

After iron city, we're probably looking at something like:
91bpt -40gpt @ 100%
7bpt +37gpt @ 0%
8.4bpt -7.7gpt per 10% slider

After copper city:
94bpt -47gpt @ 100%
7bpt +33gpt @ 0%
8.7bpt -8.0gpt per 10% slider

Let's take that as a base. The specialists in Mitchum's spreadsheet with Rep give us an extra 108bpt. That's up to 202bpt @ 100%, 115bpt @ 0%, 150bpt @ break-even. The question is how long to get there. We are probably assuming T+15? With Currency we get +18cpt (let's say 3cpt per top 4 cities, 2cpt elsewhere), which translates into:
115bpt -47gpt @ 100%
7bpt +54gpt @ 0%
10.8bpt -10.1gpt per 10% slider

That's up to 223bpt @ 100%, 115bpt @ 0%, 169bpt @ break-even.

If we take the break-even points, we finish Astro+CS+Construction in 37t without Currency, or 33t with Currency. Without Construction, it's 33t no Currency, 29t Currency. With a GM bulb on top of 3 GS bulbs, it's 24t Currency, 21t Currency. Now 3-4t extra is 507-676b worth, compared to 669b cost of Currency. Currency itself takes 4.4t to research @ 150bpt, compared to 4t shaved off other research. So, we're looking at practically break-even, without cash trading.

This is very, very rough, however. I don't have time to estimate turn-by-turn improvements to cities and all. Last game we found out just how much work it is to estimate research rates, even when the empire is largely set up, so I don't really trust any numbers I put together in half an hour. Also, once again, we don't know how much cash we can get from AIs, or how quickly Willem will have Currency available for trade.

Spoiler :

Math			585		488
Philosophy		1872		GS
Compass		936		780
Machinery		1638		1365
Optics		1404		1170
Astronomy	4680		2xGS
CS			1872		1560		GM+210b
--------------------------------------------------------
Subtotal				5363		4013

Currency		936		669
Construction	819		683

Conclusion: I could go either way. To confuse things more...

Edit: feel free to check my numbers and refute them. I'm not very confident in this estimate.

 

[Dhoomstriker]

Pigs City should not build a warrior since Stone or other cities can. Pigs needs granary+lh asap (if it doesn't have them yet).

Okay, I have apparently confused two people now.

I am only talking about the build item in Pig City that comes AFTER Pig City has completed both the Granary and the Lighthouse.

The test game had: Granary -> Lighthouse -> Monument

My spreadsheet has: Granary -> Lighthouse -> Warrior

I just wanted to make it clear that we're going with what my spreadsheet says.

At 1 Hammer per turn, and obsoletion at Astronomy, the Monument build item is dubious at best.

 

[Dhoomstriker]

Dhoom - is there an up-to-date test save (for just our empire)?

Not yet. I can try and put one together, but it will take some time.

Current
88bpt -33gpt @ 100%
7bpt +41gpt @ 0%
8.1bpt -7.4gpt per 10% slider

I'm not sure where these numbers come from. If I set the Science Rate to 100%, we are losing 17 Gold per Turn, not 33 Gold per Turn.

We have 345 Gold in the bank.

After copper city:

Just to be clear, there will not be a 7th Settler (i.e. a Settler for the City that gets built after Iron City is built) for a long, long time. I don't think that we'll even be able to settle a 7th City prior to getting Astronomy, but I guess it depends upon what we build where (Stone City might have the best chance of pumping-out a 7th Settler, but we may not want to invest our Hammers into a Settler).

 

[shyuhe]

A 7th settler for fur may be worth it, if only for the extra whipping capacity it gives our empire. It's not urgent though since we'll be in caste for the foreseeable future.

Don't worry about the test game for now then. I'll try futzing with the math myself.

 

[Dhoomstriker]

Another thought that I want to run by you guys:
We can either settle Iron City so that we have 2 Clams in its initial-9-square radius, as mdy suggested, or one west of there, having only one Clam in its initial-9-square radius.

In both cases, we get both the Iron and the Plains Hills square within its initial-9-square radius.

The reason to consider settling 1W of where mdy suggested is that we'd then get a Silk within our Cultural Borders if we were able to get 150 Culture in Iron City.


So, the decision comes down to:
Will we commit to building some form of Culture in Iron City?

If yes, then we can safely settle 1W and then eventually plan to get +1 Happiness from the Silk.

If no, then we won't even expand to its fat cross, and therefore, we will want both Clams within its initial-9-square radius.


My proposal:
Iron City grows to Size 2 while working a Clam and then works the Clam + Iron.

It can then grow to Size 3 and work the Plains Hills Mine.

We'd probably follow this approach regardless of which settling location we chose.

The difference would be that I propose that we build: Monument -> Granary -> Lighthouse

If we settle where mdy suggested, then we can forget about ever getting the Silk within our Cultural Borders and thus we might as well just build: Granary -> Lighthouse


Note that unlike Pig City, Iron City will have plenty of time to complete a Monument.

Another factor is that a Monument's Culture does not expire once Astronomy is learned, so it wouldn't be a wasted benefit.

We would probably still eventually need to build a second Cultural Building there (a Monastery?--probably Taoist Monastery as Taoism will likely get founded in Iron City if we aim to spread Confucianism everywhere else), but that Monastery would simply be the "cost" of "earning" +1 Happiness throughout our empire.

Given that it normally costs 120 Hammers for +1 Happiness in a SINGLE CITY, such as by building a Temple or building a Colosseum, I think that a cost of 90 Hammers (a Monastery) plus 45 Hammers (a Monument) plus a bit of time lost on being able to use a Granary and a Lighthouse in Iron City, is a very worthwhile payoff.

 

[babybluepants]

I'm not sure where these numbers come from. If I set the Science Rate to 100%, we are losing 17 Gold per Turn, not 33 Gold per Turn.

Hmmm... I guess I should've opened the save, rather than quickly glancing at your screenshots. Oh well, it should be done more carefully, but I can't get back to calc 'till late tonight.

 

[shyuhe]

150 turns is too far out (or even 75 with a double cultured monument). We'll be in full war mode by that time so it's not worth it. If we really need the silk, we can just dump a city on it. I doubt we'll have that issue though.

 

[Dhoomstriker]

A 7th settler for fur may be worth it, if only for the extra whipping capacity it gives our empire.

I don't disagree that additional Settlers could be useful.

The point is more that when we hire Specialists, our Cities will be producting 0 Hammers from Food and 1 to 2 Hammers from Hammers per turn... 1 to 2 Hammers per turn would take a very, very long time to complete a 149-Hammer Settler.

Iron City will be busy building infrastructure, but Stone City would be able to produce a Settler.

However, we might prefer to just keep growing Stone City as well as using Hammers on other build items, saving our post-Caste-System Food from other Cities for Settlers, either from manually building Settlers or from whipping them, or a combination thereof.

So, unless we decide to spend our raw Hammers from Stone City on a Settler, and we'll probably be better off spending raw Hammers on Buildings (a Forge? a Barracks?) or Military Units, then I don't foresee a Settler being completed until after we are done the Great-Person-Generation process.

I'll try futzing with the math myself.

Well, I can certainly offer some data that will help:
Turn 137: Iron City will be founded

It will take 9 turns for this City to grow while working a Clam


I created this 6th City (Iron City) in our old Turn 142 test saved game in place of Fur City.

It will cost us roughly 6 to 7 Gold per Turn empire-wide to own Iron City, so let's say 7 Gold per Turn.

At a 100% Science Rate, Iron City adds 6 raw Flasks per turn (not counting bonuses to tech research).


As of that test saved game, it looks like Math will be completed in about 5.2 turns. Once we factor in just about any other factor, such as Currency's additional Trade Routes or even growth into a couple of more Coastal squares, it is very practical to believe that Math will take us 5 turns to complete.


Currency will take us 11 turns to complete.


If we can afford to get both techs without needing to pause for any turns of 0% Science, that brings my estimate down from 17 turns to 16 turns.

So, what do we have to work with? 345 Gold.

345 Gold / 16 turns = 21 Gold per Turn.


Our current costs are 17 Gold per Turn.

As of Turn 142 where we own Iron City instead of Fur City, our costs are 28 Gold per Turn.

As of Turn 142, if we spent 1 turn at a 0% Science Rate, we'd make 49 Gold, which would be more than sufficient to cover the cost of our 16 turns of Research.


Therefore, I think that it is very reasonable to go with the "17 turns" number for Currency -> Math.


Could that date be ealier? I would say that at best, we could research Currency in 10 turns instead of 11, if we manage to get additional Commerce squares boosting us up. However, I would not count on this extra turn.

However, I think that it is very reasonable to skip Binary Research on the last turn of learning Math. The Trade Routes from Currency, overflow Flasks from Currency, and skipping Binary Research should mean that we can easily finish Math without needing that extra turn at a 0% Science Rate.


Thus, I believe that while we're stuck with 11 turns for Currency, I think that Math is very doable in 5 turns with the last turn not being Binary Science, putting us at 16 turns for Currency -> Math.


That situation reduces our lost GPP to an even smaller amount.

 

[Dhoomstriker]

Hmmm... I guess I should've opened the save, rather than quickly glancing at your screenshots. Oh well, it should be done more carefully, but I can't get back to calc 'till late tonight.

Yeah, sorry... the screenshot was done very quickly and I hadn't deleted Fur City in it.

 

[Dhoomstriker]

150 turns is too far out (or even 75 with a double cultured monument). We'll be in full war mode by that time so it's not worth it. If we really need the silk, we can just dump a city on it. I doubt we'll have that issue though.

I agree that 150 turns is too long, that's why I suggested that we'd also prioritize a Monastery there, after completing a Granary + Lighthouse, for +3 Culture per turn (+4 if we are lucky enough for Confucianism to appear there, but I won't count on that possibility).

 

[shyuhe]

You need a religion to build a monastery.

Why don't we just tech currency without math? We'd give up a slight discount but it'll let us sell some small techs for cash.

 

[Dhoomstriker]

Another (minor) benefit of building a Monument as our first build item in Iron City would be that it would reduce the threat from Barb Galleys.

 

[Dhoomstriker]

You need a religion to build a monastery.

Well, we want Confucianism in at least 4 of our Cities before founding Taoism, right?

That leaves Taoism to be founded in one of either Stone City or Iron City.

Taoism also comes with a free Missionary.

Getting a Religion in Iron City won't be an issue if we don't mind settling for Taoism.

 

[Dhoomstriker]

Why don't we just tech currency without math? We'd give up a slight discount but it'll let us sell some small techs for cash.

Yeah, I'm pretty convinced on the math, too. That is, the math of "not knowing math" before we learn how to make coins. Haha. Hopefully, we won't have our mathematically-challenged City officials bumbling away our coffers!

 

[mdy]

Another thought that I want to run by you guys:
We can either settle Iron City so that we have 2 Clams in its initial-9-square radius, as mdy suggested, or one west of there, having only one Clam in its initial-9-square radius.

Even with a monastry and monument it would still take 50+ turns to get that border pop. If we wanted that happy I think we would have to settle another city to get it. Given that we will get most of our hammers from the whip, whipping a settler would be more hammer efficient than building a monastry+monument, especially as it would allow iron to grow quicker.

Currency would effectively give us +18 base beakers/turn. Currency costs 936 beakers. Taking account of the 20% bonus for knowing a tech pre-requisite currency would need 36-37 turns to pay off if we just relied on trade routes. Assuming we are not too unlucky with GP's and trading opportunities with other AI's we have 25 turns of teching after we revolt to CS, and around 8 (I hope) before we revolt. If we go maths-currency we should get currency around 4t after revolting to CS. So if we just relied on trade route income I believe we would be around 7-8t behind if we self teched currency. We would need 120-140 gold to make up for this which looks doable, so whether currency would give us more beakers or not depends critically on when the AI would be willing to trade it to us, which is always something of a gamble. However...

One critical point that has not been mentioned is that, in the best case scenario the date we stop researching and go to war will be determined by the date at which we generate the final GP, not by our research, so even if currency-maths would give us slightly more beakers over maths-currency it would still be counterproductive if it delayed our final GP by even 1T which it would.

 

[babybluepants]

One critical point that has not been mentioned is that, in the best case scenario the date we stop researching and go to war will be determined by the date at which we generate the final GP, not by our research, so even if currency-maths would give us slightly more beakers over maths-currency it would still be counterproductive if it delayed our final GP by even 1T which it would.

Sure, which is why I'm assuming Math-Currency, which also reduces the actual cost of Currency to 669b. We def. shouldn't be delaying our civic-switching. I think the math works even better for Currency, now that I know our per-turn money loss is less, btw.

 

[mdy]

Another (minor) benefit of building a Monument as our first build item in Iron City would be that it would reduce the threat from Barb Galleys.

I think we can afford to dedicate our existing trieme to protecting both of those nets. Everything else can be protected by the trieme being built in gold city, so I don't think we need to worry about this.

 

[Dhoomstriker]

One critical point that has not been mentioned is that, in the best case scenario the date we stop researching and go to war will be determined by the date at which we generate the final GP, not by our research, so even if currency-maths would give us slightly more beakers over maths-currency it would still be counterproductive if it delayed our final GP by even 1T which it would.

Looking at things from this perspective, then Math -> Currency would take at most +1 turn to complete compared to Currency -> Math, since we may need to spend 17 turns instead of 16 turns (having 1 turn of 0% Science Rate in between Math and Currency).

However, that +1 turn to complete Math -> Currency WOULD NOT affect the GPP generation, so, as mdy points out, it's kind of a meaningless difference.


The other point is the assumption that I made in that Gold City would need to whip a Confucian Missionary. If it does not need to do so, then, with a Currency -> Math in 16 turns scenario, we'd be looking at:
- 2 turns of delay in Paris
- 4 turns of delay in Gold City
- possibly some turns of delay in Marble City, depending upon how quickly we can generate Great People in the other Cities

If we get Math early and thus get Pacifism as soon as we can possibly first use it, then Marble City CAN hire Scientists earlier, as other Cities will be able to hire Scientists earlier and thus will be able to beat Marble City in terms of generating a Great Person.



So, yes, we could say that Currency -> Math saves us potentially 1 turn of Research. However, it is 1 turn on self-teching Research, which won't matter nearly as much as Lightbulbing Research. Thus, it's a savings on the wrong item.


The gamble is that:
If Willem learns Math
Or
If Cathy and Isabella BOTH learn Math

Then delaying Math could help


The gamble with that gamble is that if we do NOT get Math in trade, then we will be very late in spreading Math around to the AIs and we will probably end up being forced to self-tech Calendar.


Self-teching Calendar would take up far more than the 1 turn saved on self-teching Research by going Currency -> Math.


So, if mdy is willing to compromise and actually research Currency, then I am on board with going:
1. Math -> Currency
AND
2. Trading/gifting around Math as soon as possible (forget the free Gold, we can get it in other ways, such as by demanding it or by trading other techs), with the hopes of actually being able to get Calendar in trade on time
AND
3. Putting our main-strategy Great-Person-Generation back as the primary goal, instead of saving 1 turn of self-teching Research as our primary goal

 

[Mitchum]

I'm fine with Math -> Currency.