SGOTM 16 - U'nu'sual S'us'pec'ts

[ChrisShaffer]

I don't think it's that simple, Brian. If we settle a Great Merchant, we get +1 food and +6 gold until the end of the game. Say we get the great merchant on turn 100 and the game lasts until turn 400 (of 500 total possible turns). Settling would yield 300 food and 1,800 gold.

If we use the Great Merchant to bulb, we get 1,000 + (3 * total population). Assume our population is 15 when we bulb, that yields 1,045 beakers. Currency costs 624 beakers, so 421 beakers are wasted, yielding 624 beakers. We can also count as a benefit the earlier trade routes from getting Currency faster. Because we will eventually get those trade routes through manual research, we can only count them as a benefit for the earlier turns, not for the turns after it would be manually researched. Say we are generating 20 bpt at break even, we could research Currency manually in 31 turns. During those turns, we're unlikely to have any multipliers (libraries, markets, etc.) to calculate. If we are getting +5 trade from the extra trade routes, that's a benefit of 155 trade. We also get the less quantifiable benefit of researching something else during those 31 turns, thus getting other technology faster. So the total benefit is 624 beakers + 155 trade + 31 turns faster research on another tech.

So, we're comparing settling (300 food and 1,800 gold spread over 300 turns) vs. bulbing (624 beakers + 155 trade spread over 31 turns + faster research on another tech). I personally think bulbing is the better choice, because earlier stuff yields greater returns over time.

Edit to add that trade_mission_gold = 500 + 200 * trade_profit_between_target_city_and_our_capital, so early it's not great but later it can be quite profitable and better than bulbing.

 

[ChrisShaffer]

1) I like option (2) as well. We won't be building (full-price) courthouses for a while and I imagine we'll be whipping enough that we won't need to run more than two specialists at a time. And getting MC asap seems like a good idea for both the (cheap) forges and excellent trade bait.

2) I'm not convinced that we should spend any hammers trying to get the barb city now. I'd rather build our own cities for a while and capture the barb city later with a few melee units. Super especially if it's on a hill.

Agree to both of the above.

4) Good point Revent about the wonder building. We should try and focus these somewhere so as to focus our culture, just in case we go for that vc.

Whether we go for that victory condition or not, we must have two legendary cities, so Revent's advice is on target.

 

[Deckhand]

I agree that we need a plan for great people that covers the entire game. And a plan for where /what wonders will be built.

Edit:
You own at least 2 Legendary culture cities.
You have learned Future Tech 1.
You own at least three Holy Shrines
You own at least three Corporate Headquarters

 

[SteelHorse]

For the record, I will join all those in favor of option 2 and leaving the barb city mostly alone for now. Would be good to get one unit over to keep an eye on things just in case Brennus thinks it would be a nice addition to the Celtic empire.

And other than #2 and getting our cottage site (capital?) up quickly there aren't any obvious next steps until we see more of the map.

I agree with Deckhand that we map out a tentative plan for how we achieve the rest of the goals. Since many are GP related (shrines, Corps, Cult Victory?), we need to think those through as best we can, starting with the Cult Victory and working backward.

So a few Q's.

How many GP do we expect to generate over the course of the game?
How many will be needed for the final cult bombs?
Can/should we rely on the AI to build Shrines for us?
Ditto for Corps?
Will the shrine cities be our Cult cities?

 

[neilmeister]

OK, great discussion guys.

@ Barb city
Most of the team don't want to worry about the barb city just yet, that is probably sensible. We haven't got melee units, and we aren't about to get them anytime soon, so when we do take it, it will probably be with Chariots. Our thoughts may change once we see what it has. Copper

I think we should still build a Chariot or two out of Paris, just so we don't get caught with our pants down by barb Archers or worse, Axes.


@ Grand Strategy

Shrines
I personally don't think we will need to worry about shrines at all. We will conquer AI cities with shrines. We will get the occasional Prophet (we are building Oracle), so we can build our own if the AI is too lazy to do it for us.
Brennus did us a big favour by both founding a religion and building a GP generating wonder. He will build a shrine. AI's shrine Prophets if they can.
Anyway until we found a religion or get a Prophet, it is probably not worth worrying about.

Corps
We will need Sushi, any other corp is really just to satisfy the 3 corps requirement. We will hopefully get the free GM from Economics, but we will get at least 1 more from GP Farm by then anyway.
The other 2 corps just require any great person, Scientists are so easy to produce so it doesn't matter which we get. A Culture generating corp would be ideal though.

Culture
I would expect that we will need at most 1-2 artist bombs. We might even get away without any. Sushi/Sistines/Representation/Cathedrals will give us crazy cpt, at 0% slider, while researching at a super fast pace.

 

[neilmeister]

Guys, I probably won't be able to do any more testing for a day or 2, as I have to create and play test GOTM 83. (its already a day late)

 

[Revent]

I don't think it's that simple, Brian. If we settle a Great Merchant, we get +1 food and +6 gold until the end of the game. Say we get the great merchant on turn 100 and the game lasts until turn 400 (of 500 total possible turns). Settling would yield 300 food and 1,800 gold.

If we use the Great Merchant to bulb, we get 1,000 + (3 * total population). Assume our population is 15 when we bulb, that yields 1,045 beakers. Currency costs 624 beakers, so 421 beakers are wasted, yielding 624 beakers. We can also count as a benefit the earlier trade routes from getting Currency faster. Because we will eventually get those trade routes through manual research, we can only count them as a benefit for the earlier turns, not for the turns after it would be manually researched. Say we are generating 20 bpt at break even, we could research Currency manually in 31 turns. During those turns, we're unlikely to have any multipliers (libraries, markets, etc.) to calculate. If we are getting +5 trade from the extra trade routes, that's a benefit of 155 trade. We also get the less quantifiable benefit of researching something else during those 31 turns, thus getting other technology faster. So the total benefit is 624 beakers + 155 trade + 31 turns faster research on another tech.

So, we're comparing settling (300 food and 1,800 gold spread over 300 turns) vs. bulbing (624 beakers + 155 trade spread over 31 turns + faster research on another tech). I personally think bulbing is the better choice, because earlier stuff yields greater returns over time.

Edit to add that trade_mission_gold = 500 + 200 * trade_profit_between_target_city_and_our_capital, so early it's not great but later it can be quite profitable and better than bulbing.

20bpt is a bit pessimistic I think. To bear in mind, it's not just 300 food, it may allow the city to grow one size larger if we were to settle 2 whilst giving 12 gold, post multipliers, even more, One size larger means we can run an extra specialist for free as well so not as easy as just 300 food and 1800 gold. Extra specialist could get us a great person which could end up saving lots of turns in the long game. That one extra food per turn could mean our city grows faster to work the next improved tile which also results in a snowball effect. The extra gold also means that we can push our slider up a bit as well therefore resulting in lots of technologies being researched faster all game long probably saving us more than 31 turns on currency.

Besides, I'm sure we'll be doing more than 20bpt breaking even so 31 turns does seem a bit extreme. I just have a thing against merchants being used for bulbing if they're not being utilised to the max Because food is power and if we are going to bulb, we should bulb a tech that will be worth the full GM. I propose a test to see if bulbing currency significantly helps our game. The true power of bulbing lies in tech brokering which will right now be severely limited or non existent and I don't think currency is THAT crucial to the game because we will have GLH anyway.

Re: Grand Strategy

I vote against building a single shrine. Does not matter how much gold it will net us because there are alternative methods to getting this gold, but if we decide to go space race, we will need to squeeze out EVERY GP we can get for golden ages. Late game, they are very powerful. Speaking of, we need MoM as well. Not sure how we will get it, it will be a problem if it ends up on the other continent! So we may even need to focus a whole war around it

I think Astro should come before Lib actually; some may find it a bit of a gamble, but I think as we could risk losing Lib, but

What is our Lib target? When are we aiming to get astronomy? Can we optimise our game to win the same turn we finish researching Future Tech 1? (By the way, by future tech 1, does it mean we research FT and then research it again to get FT1 or just FT once should do the trick?)


Great People we SHOULD be getting:
-GM (Generated)
-GA (Music)
-GS (Academy)
-GS (Found Taoism)
-GS (Bulb Edu)
-GS (Bulb Lib, Might not be possible)
-GM (Economics hopefully for Sushi)
-GS (Physics)
-GSpy (Communism)
-GEngineer (Fusion)
-Can't remember if there's anything else we get great people for. GG doesn't count.

With Caste, the problem of getting GA's, GS's and GM's is not an issue. EVERY great person is crucial however, so going down the music route for the GA should mean we are going for GLib as well which does wonders in this game! Also, late stonehnge suggests to me that Mids will be built late so we could even slow build it, if we get lucky, good for us, otherwise lots of fail gold.

Our issue now is getting Great Prophets, Great Engineers, Great Spies. Great Engineers are the hardest to get so if possible we should try to build the hanging gardens in one of our culture cities meeting target of wonder spam and some GE points. The health will be a game long bonus (late game, health is a big problem) as well so it's a pretty powerful wonder to want but not crucial. Once we get our great scientists for bulbing our techs, we can add a combination of engineers, spies and prophets to pollute the pool and get lots of different GP's (and adjust the specialists accordingly to make one person more likely). Spies become less of an issue once we get to Constitution (or democracy, can't remember) We can just add lots of them.

We should not need to culture bomb. That should honestly be a last resort. We're founding Taoism, we will have Budhist shrine, we should hopefully have Conf as well. That should be enough with Cathedrals to get us to our culture by 1800 AD.

 

[neilmeister]

I agree with most of that. The exceptions:

- I doubt we will found Taoism. Not because we don't want to, just because the AI loves bulbing Philo and we have so much else to do.
- The extra food in Paris is not a positive for settling a GM. We have so much food in Paris (actually in all 3 cities so far) that we will have problems with unhappiness. These can be solved easily enough with 3 and even 4 pop whips, but until we have some way of raising our happy limit, the extra food is hardly desirable.
- I also believe the argument about the 'full' power of a GM is not convincing. The real power of a bulb, any bulb, is to give you an instant advantage. There are no techs for a long time that will utilise the 1000 beakers from the bulb more than Currency. Currency is also a very popular Oracle target for Deity players (which I am not, btw) because of the turbo boost to the economy.

Rapid expansion is easily affordable with 4 trade routes per coastal city and the ability to build Markets & wealth.

If we do manage to Oracle Metal Casting and bulb Currency, we will have 2 of the best trade techs. Our hemisphere contains at least one more AI, hopefully two, so there will be tech trading opportunities.

 

[Revent]

I guess it depends on how many cities we can settle that are not coastal because as it is, I see the benefit of currency being marginal with the empire as it is. I've seen successful deity games with currency still not researched at the point of Lib. (by none other than obsolete of course ). If our empire is small, getting currency has next to no immediate benefit! Oh and initially, there may be happiness problems, but that just means with caste, we can run a lot more specialists and once we get to HR (or Rep if we decide for Mids/capture them), our cities can grow really large. This ALSO gives more commerce from trade routes so is a win-win situation in my opinion. In a normal Cuirs sweep game, I would bulb without a second thought, but I think settling in this instance pays off. We don't need to build wealth straight away if we have settled our specialist. Rather than building wealth, we should be spamming culture buildings in at least one city. (Another snowball effect, more time focused on building culture whilst the settled GMs gold pays for expansion whilst waiting for currency.)


This is Emperor so I think founding Taoism shouldn't be a problem at all! We just need to open up the path to it which we will be doing anyway. It should honestly be a priority because we want to have lots of great people and pacifism is magical without Phi (and almost magical with it). Plus, if we found taoism, that means the AI put less priority on it. The earlier we get it, the less likely an AI has started researching it meaning we can delay Lib and save it for something better.

 

[neilmeister]

Happy that we are having the discussion, as it makes everyone think about where we are going. If I can't convince myself, how can I convince anyone else.

Bulbing Currency vs Settling GM.

I think the argument can be summarised as:

Bulb
5gpt initially,
Excellent trade bait.
Extra 20% discount on CoL (which would be about 5 bpt for the ~10-20 turns we are researching CoL), or in other words, about 50-100 beakers.

Settle
6gpt, permanently.
1fpt,
No trade bait.
In Representation, additional raw 3 bpt.

So, the raw gold is almost identical immediately (we will have at least 5 cities by then), the beakers and trade opportunites heavily favour bulbing.

There is no plan to build Wealth, but it is a wonderful option to have when/if you run out of cash, and we should almost run out of cash, because if we don't then we haven't expanded fast enough. (Though having both GLH and early Currency mean we should never even go close).

Maybe Obsolete skipped Currency because he didn;t want to waste the turns researching it and he didn't have a GM.


I would love us to build 4-5 Chariots immediately after connecting Horses to take that barb city, but I am clearly in the minority on that one.

 

[Revent]

Happy that we are having the discussion, as it makes everyone think about where we are going. If I can't convince myself, how can I convince anyone else.

Bulbing Currency vs Settling GM.

I think the argument can be summarised as:

Bulb
5gpt initially,
Excellent trade bait.
Extra 20% discount on CoL (which would be about 5 bpt for the ~10-20 turns we are researching CoL), or in other words, about 50-100 beakers.

Settle
6gpt, permanently.
1fpt,
No trade bait.
In Representation, additional raw 3 bpt.

So, the raw gold is almost identical immediately (we will have at least 5 cities by then), the beakers and trade opportunites heavily favour bulbing.

There is no plan to build Wealth, but it is a wonderful option to have when/if you run out of cash, and we should almost run out of cash, because if we don't then we haven't expanded fast enough. (Though having both GLH and early Currency mean we should never even go close).

Maybe Obsolete skipped Currency because he didn;t want to waste the turns researching it and he didn't have a GM.


I would love us to build 4-5 Chariots immediately after connecting Horses to take that barb city, but I am clearly in the minority on that one.

Fair enough I am obviously in the minority (only one actually ) here of course so I'm sure bulb will happen Just going to fix that a bit though

Bulb
-Extra 5 commerce per turn until currency is eventually researched anyway. @ 20 turns for currency only 100 commerce gained.
-Trade bait for two AI or three at best, techs to be gained are limited because this is Emperor after all.
-We will be getting the 20% discount anyway once we research currency.
-Currency comes x turns earlier meaning everything else comes x turns earlier saving x turns of the game

Settle
6 raw gpt, permanently on top of the commerce that is gained from currency x turns later. Lose 5x commerce to gain 6 raw gpt.
1fpt allows cities to eventually grow larger so commerce from TR's kicks in even more.
No trade bait.
Game can be considered x turns slower but it allows research slider to go up as well allowing higher research anyway so the 'loss' of x turns is offset by all techs going faster speeding up overall game.
In Representation, additional raw 3 bpt.

 

[neilmeister]

Bulbing Currency might not happen. No guarantee we will Oracle MC, so we might prefer to bulb that instead.
We might also get a library out a few turns before the GM is popped and get a GS at 2% odds.

What I find interesting is that it sounds like we are both using the same points to back our point of view.

Maybe we should have this debate when we actually get the GM. By then we will know a lot more about the map, potential trading partners.... but hey, I like making uninformed decisions hundreds of years before we have to.

 

[neilmeister]

Interesting reading kcd?

 

[kcd_swede]

Interesting reading kcd?

Yes! Hard part is keeping my mouth shut.

It seems universal that folks play a burst, talk until they confuse themselves, wash rinse repeat.

 

[neilmeister]

Yes! Hard part is keeping my mouth shut.

It seems universal that folks play a burst, talk until they confuse themselves, wash rinse repeat.

Thats how we roll !!

 

[Revent]

Fair enough. Let's just wait to see when it happens If we are unable to oracle MC, bulbing that definitely seems a lot more beneficial however!

 

[SteelHorse]

Bulb
-Extra 5 commerce per turn until currency is eventually researched anyway. @ 20 turns for currency only 100 commerce gained.
-Trade bait for two AI or three at best, techs to be gained are limited because this is Emperor after all.
-We will be getting the 20% discount anyway once we research currency.
-Currency comes x turns earlier meaning everything else comes x turns earlier saving x turns of the game

Settle
6 raw gpt, permanently on top of the commerce that is gained from currency x turns later. Lose 5x commerce to gain 6 raw gpt.
1fpt allows cities to eventually grow larger so commerce from TR's kicks in even more.
No trade bait.
Game can be considered x turns slower but it allows research slider to go up as well allowing higher research anyway so the 'loss' of x turns is offset by all techs going faster speeding up overall game.
In Representation, additional raw 3 bpt.

I like this analysis. Given that the game objective is to complete in the fewest # of turns. The question is which will reduce the number of turns the most over the course of the game. What assumption should we make for time to self-research currency - 10 turns. How many turns do we expect the game to go - target date 1700? - 250 turns (with 57 played).

If we assume we get Currency in 20 turns instead of 30 from now. Then the difference can be calculated as 1 settled GM for 170 turns versus 10 overall turns + 10 turns of Currency benefit + tech trades by getting 10 turns earlier. Do we think that 170 turns of a settled GM covers that?

 

[ChrisShaffer]

We can get currency in 20 turns, but not from "now." We can't currently research currency. The analysis needs to begin on the turn when we can start researching currency.

 

[SteelHorse]

The assumption is that we could start researching (or bulb currency) in 20 turns after writing and alpha. That may be too aggressive (as may be the 1700 end date).

 

[Revent]

I like this analysis. Given that the game objective is to complete in the fewest # of turns. The question is which will reduce the number of turns the most over the course of the game. What assumption should we make for time to self-research currency - 10 turns. How many turns do we expect the game to go - target date 1700? - 250 turns (with 57 played).

If we assume we get Currency in 20 turns instead of 30 from now. Then the difference can be calculated as 1 settled GM for 170 turns versus 10 overall turns + 10 turns of Currency benefit + tech trades by getting 10 turns earlier. Do we think that 170 turns of a settled GM covers that?

If we take the beaker cost of currency and add that to the commerce gained via trade routes for x turns and divide it by the number of turns we think the game will last, we will get a net gain of beakers/commerce per turn throughout the whole game.
If that offsets the gain in

Currency=624 beakers
At 30 beakers a turn, it will take us 21 turns to research it.
We research Alpha or Maths, it will only take us 17 turns to research it
We research both Alpha and Maths, it will only take us 15 turns.

We will definitely be researching one before currency so let's say it saves us 17 turns in the game overall. So our game is speeded up by 17 turns plus the commerce gained in those 17 turns which we can roughly say is 85 (5*17). But the actual game wide benefit of game being 17 turns faster is a bit too vague and not necessarily true. What we can say is that if the game lasts another 150 turns, the benefit of bulbing currency becomes a net gain of 85 commerce + tech trade bait with two or three AI (or even none of nothing is worth trading for!) + 3.33 total (multipliers have no effect) beakers per turn (500 beakers/ 150 turns). The longer the game lasts, the smaller the effects of bulbing become! The only effect of how many turns we save on currency is 5*n where n is the number of turns it takes us and in the big scheme of things, this figure is almost negligible. Also bear in mind, this strategy becomes weaker if we have maths and alpha before currency because the net gain in beakers throughout the game is even less!

If we were to get Rep, settling would clearly be significantly better than bulbing purely from the science gained because the 3 beakers are raw and are subject to multipliers which we will build a lot of considering we are Cre after all!
I can't remember the beakers:gold ratio but clearly 6gpt all game long (subject to multipliers) clearly outweighs bulbing currency! Add the food to the equation and that's a completely new benefit on its own!