[SPOILERS] scooter's Industrial Revolution

[scooter]

I was thinking a little more about the Engineer if it happens, and here's another weird idea. What if we settle it? It's worth base 3h and 3b, and that comes up to 6b with Representation. Factor in the library, and it's 7.5b. It'll take an eternity to catch up to a bulb, but it's something. Then the 3h are 5.25 immediately with Bureau + Forge, and that'll improve as we add a few more hammer modifiers. It's basically a plains workshop that you don't need to feed that also generates some beakers.

It's underwhelming, but we're dealing with a host of underwhelming options - may as well leave no stone unturned. As of right this second I think I like it better than 1-man GA and slightly less than saving for a bulb, but let's be honest, I don't particularly love any of the choices.


Also, quick programming notice for you Sullla. Turn rolled too late for me last night, and I didn't have time this morning. I will definitely have time to get it when I get home from work, but I'm not certain I'll have the time to write anything up before I'm right back out the door for the evening. If that happens, the report will either wait until late tonight or tomorrow morning, or you can just do it yourself tonight if you really feel like it. Although the only real noteworthy thing for this turn is to see if we got demos on Donovan yet, depending on how many EPs he spent on us.

[Sullla]

Go ahead and play when you can, I can cover the reporting for today's turn.

[Sullla]

Turn 264



This is our eastern scout, who will meet pindicator next turn. Pindicator is actually our eastern neighbor, even though we've had no contact thus far. Hopefully the late contact will have him focusing on other directions for early expansion. Our scout will be able to continue northeast from here and meet REM in another half dozen turns to complete the full set of contacts.



Our western border with Donovan. No, he's not especially far away either. It's a cramped map. Scooter, I was thinking that we could take two rifles and put one apiece on the yellow dotted hill tiles. That will reveal the tiles boxed in white, and we'll be able to see anything that Donovan tries to push into the border zone. I moved the rifle we drafted out of Cotton Gin along the roads in this direction; if you don't like this plan, just move him back into the city next turn.

We have Donovan's bar graphs, and he's ahead of us in pretty much every category. Of course, he's never whipped or drafted (we've sacrificed 4 total pop) and his Expansive trait holds little use after these early turns are over. But there's no mistaking that he's ahead of us right now. I remain a bit concerned that our third city location, which only claims a single food resource, is going to hurt us against all these other teams which walked into the fog and found double and triple food bonuses everywhere. There's nothing we can do about it now though, just have to play smart and get the best value we can from what we have.



Here's our core of cities. We drafted Telegraph this turn for a rifle, and Cotton Gin is working a really odd tile configuration to get the library finished at exactly 72/72 production. Yes, we're working an unimproved plains hill tile. We need the production to get the library done!



Inside our capital as we continue to grow as fast as possible. Do you want to see how you get crummy Demographics? Well, this is how you get crummy Demographics! We're not in Bureaucracy civic, we're not working the gold or the incense tile, and we're working a bunch of 2/0/2 coastal tiles. That is... not good. But the tradeoff is +11 food/turn, so we grow to size 8 at the end of this turn, size 9 at the end of next turn, and size 10 the following turn, then straight onto settler duty with a vastly stronger capital and watermills/workshops everywhere.

Is this a winning strategy? I have no idea! This could be totally idiotic. But that's the fun of this setup, as we're all playing in the dark and testing different theories to see what works best. I love how different teams are playing different openings, with Dreylin getting the early Taj Mahal and sitting in Universal Suffrage, Donovan pushing for the very early Academy in the capital, REM doing the early Imperialistic settler whip, and so on. Great stuff for comparison purposes. I hope the lurkers are doing some commentary on all this; I've been saddened at how little lurker decision there's been for recent Pitboss games. The new gameplay of an Industrial start should be fertile grounds for lurker chatter once more.



Ummm... keep on moving, don't look at the numbers, nothing to see here. Yeah, these are pretty bad. However, this really is the nadir for our Nationhood civic play. All three cities grow a size at the end of next turn, and then we're back into Bureaucracy with a very sweet capital, our other two cities Building Research to power us towards Steam Power tech. Despite what it might look like, we're going to finish Steam Power in 12 more turns, and we'll be finishing our first settler in 10 more turns. I've "looked into the future" via our sandbox, and the Demos do get vastly better in a couple more turns. You'll have to trust me on this one.

Scooter has us investing our espionage points (EP) into Dreylin next, and that's fine with me. I think we want graphs on Donovan, Dreylin, and Gaspar/Noble more than the other teams. I did change our EP allocation slightly, splitting them between Donovan and Dreylin since it looked like we would lose bar graphs next turn if I failed to spend anything. Feel free to tweak them as you see fit from here, scooter.

Noble/Gaspar have 112 production in the box at their capital, the only city with a lot of production stored up that we have vision on. Perhaps a very early settler build? They'll get one out before we do, but I don't think trying to build a settler this early at size 6 is a great idea. Even with Bureaucracy civic, that has to be close to 15 turns for the build. Despite 112 production in the box, they need 225 more production to finish a settler (!) Dem things are expensive.

That's all for this turn. I'll try to do a theory post on our economic plan later tonight for illustration purposes.

[Sullla]

Production Economy

This is a post about the economic model that we're planning on using in this game if all goes according to plan. The basic gist is that we ignore building cottages in favor of blanketing the landscape with workshops and watermills. We rely on production to accumulate science and wealth via the Build Research and Build Wealth options. This kind of setup only works for a late era start, as it relies on lategame civics and tile improvements to be effective.

Let me illustrate this with a couple of screenshots.



Here's how someone would typically set up a capital for a normal game. (For the purposes of this comparison, it would be better to use a non-capital city to take out the Bureaucracy bonus but I just used our sandbox for purposes of laziness.) You would improve the food bonuses first, then work as many cottages as possible, preferably grassland and riverside cottages. Use whipping for production wherever possible and don't work low food/high production tiles. This works well because Slavery civic is so effective in the Ancient era and growing cottages into villages/towns produces such high commerce yields. A town will produce 4 commerce even in the worst case scenario, and it's easy to boost them further from there (+1 commerce when riverside, +1 commerce with Printing Press, +1 commerce from Financial, +2 commerce from Free Speech, etc.) A core full of mature towns will always produce the most total economic power in Civ4, there's no getting around that fact.

But there are a couple of factors that make an Industrial start significantly different. For starters, we have access to much better tile improvements and civics than we do in an Ancient start. If you could build 2 food / 4 production workshops from the start of the game... well, I don't think I'd bother to put down too many cottages! Second, Slavery civic becomes vastly less effective due to the expensive costs of everything in the lategame. Rifles (88 production) and grenadiers (80 production) are the cheapest units we can build, and that's a 3 pop whip. Per unit. It's a far cry from doing a 2 pop whip and getting 2 axes in return. (The units are even more expensive when not doing an Industrial start; everything is 20% discounted for us here.) Third, cottages take a long time to grow into towns. It's 70 turns of working the tile unless running Emancipation, and that's a long time indeed. In the Ancient age, there's plenty of time to grow those cottages into mature towns, and it's a reason why it's a winning strategy. Here in this game, the whole map is going to be settled in 40-50 turns, and the whole game will be over in a little more than 100 turns unless I'm seriously misreading things. Do you want to spend 75% of the entire game working a crummy cottage or hamlet tile so that it will eventually become a town? It's not a good call.

So we intend to go for the production route instead:



That's a little different, right? Instead of cottages, we put down watermills and workshops all over the place. Our workers are in the process of doing this right now as our capital frantically grow to size 10. We can then use this production for Build Research or Build Wealth as needed to power the economy. Or we can also use it to construct buildings or train units too! One of the great advantages of this setup is its flexibility. As a Spiritual civ, we can also pick civics to complement whatever we're doing at the moment, going Representation/Mercantilism/Free Religion when we want to research for a while, then into Police State/Vassalage/Theocracy when it's time to crank the military. The other great advantage is speed: all we have to do is build the tile improvements and get the proper civics in place, and we're off and running. No need to grow cottages over the span of dozens and dozens of turns. We just won't have time for that in this late era start.

Note how the production modifiers work here. Build Research does not benefit from having a library in place (the beakers get added as a final step in the calculation and don't benefit from commerce modifiers), but Build Research DOES benefit from production modifiers. This means that we want anything that adds production in the worst possible way. Production bonuses will let us expand faster, and they'll also let us research faster, both at the same time. Our whole research strategy is based around lining up those things that boost production ASAP. So what are the key things we want?

* Steam Power (levees)
* Communism (State Property civic)
* Assembly Line (factories/power plants)

Those are our holy trinity of techs. We are pushing for them as fast as possible, since they let us do everything possible faster and better. This is why I think we have to consider popping our second Great Engineer for a Golden Age if we would get the bad dice roll. Yes, the Golden Age would be too early, but it would let us get a Great Scientist quickly (for the third Great Person) to lightbulb Scientific Method, clearing the way to fast Communism, and we need these key techs to fuel our production snowball. Hopefully it won't matter, but I think that's the play if we get unlucky. (This is going to be a short and intense game. I'm fine with getting some of our big bonuses early and up front.)

That first screenshot is what our capital would look like with no additional techs in place. Here's the same situation with a levee and a civics swap to Caste System:



Now we're talking. Production has gone up by 50% since the last picture, and all we did was build a levee. Now this is cheating a bit because our capital gets a massive bonus from the levee, but we'll have at least one other city (the red dot location) which will be equally beneficial. Another huge benefit to Steam Power is the +50% faster worker boost, which allows us to drop Serfdom and swap to Caste System at no loss. I think we'll be switching to Caste System relatively soon, probably as soon as our double drafting penalty wears off at Cotton Gin and we can grow large enough to run 3 or 4 Scientist specialists.

Note that we have no excess food in the last screenshot. A swap to State Property civic fixes this:



The +1 food on workshops and watermills is of course amazing. It allows us to go production crazy with the workshops while still growing at a fast pace. But State Property is even better than that, with its global +10% production bonus, and its "no maintenance due to distance from capital" ability, handy to have on a Toroidal map. Note that we lose the free Mercantilism specialist, but in return we gain the ability to have foreign trade routes once again - that's potentially worth a lot of commerce in and of itself.

Anyway, I don't think it's much of a shock for me to state that State Property is an amazing lategame civic. Everyone pretty much knows this already. I doubt there's any economic setup where it's better than what we're going to run here, however. The civic is simply perfect for what we want to do.

Then we add factories and power plants to the mix...



Production up another 75% from before, now close to cresting 100 production/turn. Note that this city is only size 12, and we're nowhere close to being maxed out here. We can build any infrastructure we want here very quickly; I think every single building can be done in 1 or 2 turns. Units are also quite cheap to train too, many of them doable in a single turn. And the capital is honestly a bad choice for this comparison, since Bureaucracy civic is distorting things. The factory/power plant combination takes most normal cities from +25% production to +100% production - it's an enormous jump. We might not manage 100 production/turn in every city, but most of them should be able to do 75/turn. Oh, and did I mention that we get the German unique building version of the factory, which builds at double speed? It's nice to be fake Organized trait.

The cost of all this is terrible unhealthiness; note the -7 health penalty above. Sheesh. Health resources are going to be super important on this map, just as important as happiness ones. Hopefully we can landgrab and/or trade for enough to avoid most of that penalty. State Property is key here as well, keeping cities from starving due to all that free bonus food.

We probably need about 40-50 turns to get these techs researched and our cities set up like this. Once we do have this kind of rocking production though, there's not much another team can do to match us unless they're also rocking the same system. A team with a lot of cottages everywhere simply cannot keep up; whipping is too inefficient. I found this out firsthand in Pitboss #6, where oledavy was slowly outproducing me with his State Property empire of workshops compared to my cottages + factory core. I would have lost to him eventually if both of our teams hadn't lost to Lewwyn's cultural victory. Well, I try to learn something from every game, and this time I want to be on the other side of the coin.

Of course, given a long enough span of time, the cottage economy will produce more beakers than the production economy and pull away in tech. You can see that because we get no commerce modifiers, we aren't able to pull off the 500 beaker capitals that are common in normal games. But we're pretty sure that under these settings, the cottage economy won't get enough time to reach that point... since our team or someone else will be parking dozens of tanks on their doorstep first, hehe.

So that's the overall theory behind this setup. Let's see how it works in practice.

[Dp101]

One question about tile improvements. Generally, when given a riverside tile, will you always choose to build a watermill or will you sometimes instead build a workshop? Just wondering if you think one more food and two more commerce is more valuable than an extra 2 hammers.

[scooter]

Great stuff as usual Sullla. One thing you didn't mention that may also be worth noting is that when State Property gives us that large food boost, it'll allow us to do two things. One is the obvious growth benefits, but the other is that it'll free us up to run more Representation scientists too with extra population. Rep scientists are worth a base 6b, which is still better in a vacuum at producing beakers than building research on a grassland workshop for example. Obviously you have to consider the food benefits too, but when a city is pretty capped out on population, we'll get some good value out of this. It also depends on a lot on what multipliers are present in a city. Most cities will prioritize hammer multipliers, so they'll always prefer to work a workshop unless we literally don't have any left to work.

Rep scientists being available will also allow us to specialize cities a little bit. Obviously Cotton Gin is just going to be an underwhelming city no matter what. That said, if we can complete the science boost trio in there, each scientist will effectively be worth 10.5b, not to mention the GPP benefits of spurting out scientists at a stage of the game where they are very useful. For us that may end up being a better option than workshopping it up and building the hammer multipliers in it. (I'm not 100% certain or committing to that being our plan, that's just a general statement.)

In other cities it'll be more worthwhile to build research. Take Sullla's example up above - look at the grassland river tile. Workshopped, Levee'd, Bureau'd, and factory'd that 5h workshop is worth 13b. If you take away the Bureau bonus since it isn't available in every city, it's still worth 10.5b, the same as the Cotton Gin scientist, but in this case it also feeds itself.

Just something extra for us to think about when we plan cities. Cotton Gin is a mediocre city, but with a little bit of work, it can still contribute a lot for us. We mentioned earlier that we should be able to get it to a place where it's feeding 5 scientists without too much effort, and that's before Biology. Before too long we'll be in Caste System nearly full-time, and at that point the only scientist limit for that city would be however many we could feed.

One question about tile improvements. Generally, when given a riverside tile, will you always choose to build a watermill or will you sometimes instead build a workshop? Just wondering if you think one more food and two more commerce is more valuable than an extra 2 hammers.

To answer this question, well it varies a little bit. Not sure if you know this (I didn't until I finally built them seriously in PB13, and I had played Civ4 for years by then), but Watermills have weird rules, so you can't actually build them on every riverside tile. You can't build them on river corners, and you can't build them across from another watermill. That said, early on we'll definitely prefer Watermills over workshops in cases where both are available for the food bonus. Once cities start to mature and hit their reasonable pop limits, I wouldn't be surprised if we convert some of them to workshops. It'll just depend on what the city needs. But yeah, in the short term I think we'll be putting watermills on every tile that's eligible and workshopping just about all of the rest.


New turn played, pictures coming shortly-ish.

[scooter]

Turn 265



We met pindicator this turn, so all we've got left is to meet REM. We can either scout around his borders here, or we can head straight north to meet REM. I don't have an especially strong preference here. I didn't offer a map trade, but it is an option. Maybe I'll wait and see if he does first.



Unfogged a few more island tiles this turn. The explorer will get picked up next turn. The south half of the island has that silver which will be useful, but other than that it's pretty underwhelming and probably not going to be an early settlement. I'm guessing the north half of the island will be a little better. It's going to create an interesting dilemma. Civs on this side of the island will want a city on the south half of the island for defensive purposes, but you won't want to sacrifice your growth curve to do so. It'll be kinda challenging to decide when to go for it. I think our Explorer may want to just camp on the north side of that island and sentry for awhile once we've unfogged it, that way we know when Gaspar decides to go for it. It'll have a small bonus benefit of preventing them from getting a free barb city on the north side of the island.



Dreylin is still saving gold, so at this point it seems he's planning to spend cash to speed up his settler. It's not a crazy idea, but time will tell if it's worth it. Yes I realize that someone in another thread is probably saying the exact thing about our 5-turn spurt in Nationhood. I also included our finances so that you can see the sorry shape of our cash flow right now compared to the Taj Mahal.



We're up to 3 watermills, and that number will hit 5 very soon. Our first workshop has started too. Steam Engine will hit sz9 EoT and sz10 the following turn, so the settler starts on T267. Next turn we'll do our final draft in poor Cotton Gin, and then we'll finally get ourselves back into Bureau.



So that means these pathetic numbers here will be slightly less pathetic next turn.

Finally, I did a dump of all the Donovan demos for those interested.

Only comment is that it looks like Donovan did squeak out a military unit or two himself. It's hard to conclude anything more solid than that.

[Brian Shanahan]

Just one quick question, how big is the circumnav bonus on this map? I've not enough feeling for a map like this to know for myself.

[Honk]

Um, Sullla, I think you may be forgetting something. The Forge gives +25% production, the factory gives +25%, and if you have a powered factory, you get another +25%. BUT, this is Germany, and they have the Assembly Plant!

Per the CivFanatics online Civilopedia, the Assembly Plant gives another +50% with Coal!

So, uh, are we playing something other than BtS 3.19, or what?

Also: Earlier, you mentioned that the Assembly Plant is cheaper, but, according to CivFanatics, it costs the same as the Factory. It is different in that it has the +50% from Coal, and enables the creation of 4 Engineers instead of 2.

Great game, guys, keep it up!

[Kurumi]

So it's 50% cheaper with coal and 2 engineer spots more.