When competitors build industries in cities

[MayoMayor]

When Don or Roger build an industry in a city, do they select an industry based on regional demand, or do they try to undermine your game (by like only building cheese factories)?

Just wondering if I should always select which industry to build in a city or if the game will select which industries are most needed. It could be helpful if the competitors chose which industry is most needed.

[Thineboot]

I'm actually running some tests. So far it seems that the highest demanded but not jet existing industry will be build by the private sector. Again, work in progress, no definitive conclusion reached so far.

Regional demand doesn't matter. See also ongoing parallel discussion.

[chaney]

MayoMayor, the game does tend to make some crazy choices when setting up a new factory that is far from the input sources it needs and far from people who want the product. It seems to place more inappropriate factories than really well-placed choices. Most of the new factories will be neutrally owned - none of the railway companies own them.

That can be frustrating to the player, making you want to buy and build better placed industries yourself on the limited space first. That can be very expensive, and the returns are low compared with expanding your rail network, so that becomes a key part of decision making in the game.


Thineboot, I'm so happy to hear you are doing some tests :) Preliminary result sounds good - it would be a simple but naively reasonable heuristic to select for the good that has the highest global demand that is unfulfilled in some sense (Global Demand - Global Production makes some sense, or the good with the maximum number of Cities that are demanding that good but out of stock, possibly weighted by demand rates.) Whatever the picking method is, there seems to be no dependence on proximity to input or output markets!!! I look forward to hearing more :)



Warning - this is really pedantic so most people should not read the rest of this message. If you love details, proceed, any further time wasted is on you.


Clarifying language as a few of us old-timers use it in this forum:

"Don" or "Roger" are competitors, as reasonably used by MayoMayor. I tend to refer to those other company entities as "AI Competitors." "AI" is too vague as there are a lot of game elements other than those companies that can be called "AI." Although "competitors" would be fine in this case because the game is single player, it can lead to confusion so "AI Competitors" is less likely to be misunderstood by a large class of readers.

"Private sector" or just "private" in terms of ownership of a factory means that some entity other than the player or AI Competitors are the owner. There are differences.

Sorry for getting far into the weeds.

[gardlt]

Originally posted by MayoMayor:

When Don or Roger build an industry in a city, do they select an industry based on regional demand, or do they try to undermine your game (by like only building cheese factories)?

From my observation, my best guess is that the AI competitors try to build a factory that has resources available. I believe they also "see" an end market, but I don't have a good guess on specifics. For example: Does it mean a city reachable by wagon, one that trains are currently connecting to, or ?????


"Resources available" theory favors the simpler single-input industries. It also favors those inputs like Milk that are both resource and consumer commodity. Train traffic will distribute that traffic to a broader region than seen in resource-only input.

In contrast to the "private" industries, AI-competitor-built industries have a good chance to make money. The rules for private industries are not followed by the AI competitors. The goal is to make money. I don't think there is much focus on whether the output is a "needed commodity" targeting higher FoD for better chance of growth. As you have seen with Cheese.

Just wondering if I should always select which industry to build in a city or if the game will select which industries are most needed. It could be helpful if the competitors chose which industry is most needed.

You should build the industries.

The primary reason is a little different than you may think: company value. Working industries are worth much more than their build price, at least double but with the possibility of going much higher. AI-competitor-built industries swell his/her company value. If you need/want to eventually take them over, it's in your interest to build the industries yourself.

Of course, as you might infer from the above, you are able to make a much better decision about which industry to build for a better growth outlook.

Also, instead of the AI competitor's company value jumping it will be yours. That gives you more collateral in order to re-finance for a bigger loan.


For me those are good enough reasons to call this a generally good strategy.

[rff1]

Like chaney, I use "the AI" to mean who or what places privately-owned factories, as distinct from those planted by "competitors". The AI seems to go for factories which are as advanced as possible, such as ones which have only just become available as a city somewhere has grown to a sufficient size, usually putting those factories many miles from suppliers and/or market, whereas competitors prefer to build factories which have some potential usefulness. But in any case it is preferable to fill a factory slot with your own carefully-chosen building if you can.

[chaney]

It is better for the player to own a useful factory than for a useless private or AI competitor owned industry to take a slot. There is, of course, the (quite high) opportunity cost of building that factory ... so it is not always the best use of your money. That decision is a key element of the game that is interesting. Secondary and tertiary values of the decision make it one that is not so easy to make well! (An impractical factory just sits there, a useful one gives you the chance to move more goods and grows Cities, which give you more opportunity to move goods - express and freight. On the other hand, the direct ROI for factories is low.) As the game moves on after a couple of years, income is very high, so the decision is easier. That's good because the opportunities start to come quickly once those Cities you have been nurturing start to take off.

[chaney]

Ok, I paid some attention to the late game in a recent run. Free map, nothing odd, time was late 1873 (fourth year or the game) and I'd connected all Cities and eaten both AI Companies. Half the Cities were over 100,000 and most were still growing, so I watched a few factory slots open up, then recorded Global Demand and Global Production for all the factory products. For the three Private factories that spawned, in each case the type with the highest ratio of Global Demand to Global Production was the selected factory. Far from proof, but it seems plausible to me. Get more data!!!

[Thineboot]

As for now I can approve rff1. [...] So far it seems that the highest demanded but not yet existing industry will be build by the private sector. [...]

Private owners don't care about resources available as gardlt believes competitors are looking for.


First tests have shown that with unfulfilled Cloth and Clothing first spot receives Tailor (Clothing) and second Weaving (Cloth) in connected cities.

With only one free spot connected only Tailor (Clothing) was privately built very soon while the other may or may not get a green city bonus for new industry.
Enforcing city growth to raise demand for Paper and offering two more free spots in connected cities the next private industry was Paper (Paper) and a map with still no Weaving (Cloth) to support the existing Tailor (Clothing).

Originally posted by chaney:

[...] For the three Private factories that spawned, in each case the type with the highest ratio of Global Demand to Global Production was the selected factory. Far from proof, but it seems plausible to me. Get more data!!!

Global demand (short for Weekly demand (all cities)) was 1.0 for Cloth and 0.5 for Clothing. With limited possibilities at the start of a game the margin of error is a bit lower than in late game. Both can be true, we need more data!!! (carefully counting the ! ;))

[Dray Prescot]

The Industrialist character (Leader) is interesting: can buy existing private Industries with NO auction in which the AI players can (often) drive up the price, can Build new Industries AND upgrade Industries for half price (which gives you an instant jump in Company value compared to the amount you actually spent), at the price of +50% cost to ALL auctions, including Techs and Resources. I think that he may get the half price upgrades on Resource sites after he owns them.

When trying to buy Industries and Resource Sites, I try to keep track of recent activity levels for them, when the activity level is low (as close to 0% as possible ) is the best time to buy it (lowest price asked) or start an auction on it. I think that the Company Value for those Resources/Industries may vary depending on activity level after you own them.

The jump in Company Value not only drives up your stock price, I think that Company Value also figures into how much you can borrow via Bonds.

So if you want to buy a lot Industries in a game think long and hard about using the Industrialist Leader. You pay a penalty on auctions, but you save a lot on the NO auctions and the half price builds/upgrades.

IF you can expand to new Cities and unowned (by you that you do not need to buy) Resource Sites, you can grow a lot faster than investing in Industries, Resources, and upgrading them.

However, later in the game, when you have expanded to all the Cities and Resources that you are going to, then Industrial and Resource upgrades are what increase how much you can deliver to your Cities that you want to keep growing. At this point you are trying to get the maximum income out of each carload of resources and manufactured goods that you can deliver. To do that requires owning the Resource Sites and Industries, and allows you to upgrade them as needed and not depend on the private investors.

Plus bigger Cities generate more Passenger and Mail Income, and bigger Cities require more resources and goods delivered to them, which requires bigger Resource Sites and Industries.

[Thineboot]

Originally posted by Dray Prescot:

The Industrialist [...] can [...] AND upgrade Industries for half price [...] I think that he may get the half price upgrades on Resource sites after he owns them.

All (General, Gangster, Industrialist, Engineer, Lady, Trickster) - why are the females at the end?

Buy Wood: same
Upgrade Wood: 100k, 200k, 300k, 300k
Downgrade Wood: 100k each
Giving back Wood: 0 for giving back

Buying private owned Meat: same

Construct Brewery: 400k
Upgrade Brewery: 400k, 800k, ???k, ???k
Downgrade Brewery: 400k each
Demolish Brewery: 240k

Industrialist: -50% for constructing new industries
Construct Brewery: 200k

Engineer: +70% for upgrading businesses
Upgrade Wood: 170k, 340k, 510k, 510k
Upgrade Brewery: 680k, 1,360k, ???k, ???k

[chaney]

Ah, more good data. Observation: when the new factories first start to spawn, there is no production for anything but Beer and Meat in most scenarios, so in that case all other products have Global Demand / Global Production = Inf. Perhaps the industry is selected from among all the Inf ratio options. More more data!!!!

[Thineboot]

If you wanna use GD / 0 = ∞ you should explain the math behind it to let's say 95% of the readers. Otherwise we should stick with undefined as I highly doubt MGS uses higher math to model demand and production ;)

[rff1]

An oddity. Germany scenario; Coblenz could easily get grain, fruit and beet to make gingerbread. A slot opened there but I did not have the €500k to buy a candy factory, and was still a bit short when the AI put in a private factory. So what is the oddity ? It was indeed a candy factory, the ideal choice.

[chaney]

rff1, over here in the less civilized part of humanity, we have a saying: even a blind chicken sometimes finds the corn. Also, a typical clock is (almost) never right, but a stopped clock (12 hour analog mechanical display) is right twice a day. I thought about a clock that runs at the normal rate but backwards ... that one is right four times a day, so my conclusion is that for better accuracy, one must build clocks that run backwards. The faster they run backwards, the more often they exhibit an exquisite level of precision. Anyone want to fund my high speed backward running most often super precise clock factory startup? Anyway, I'm glad that one worked out so nicely :)

Thineboot, eh, I'll give the readership more than 5% credit on this one. (I'm not going into set theory that *I* don't really grasp all that well in any case.) The intended set of people for the discussion (e.g. "how does this computer program work") is mostly computer savvy, and I expect most of them have understanding of notions like Inf and NaN that are common standards. I'd bet a nickel that the game uses some IEEE-like standard that gives Inf when a finite operand is divided by zero. (Note I typed "Inf", not "∞" to split hairs and emphasize the point!)
I wish a more useful heuristic was used to select among the "top" candidates, but if we understand how things do work, we can try to manipulate the setup to get the result we want :)

[Dray Prescot]

I have seen the private investors build a candy factory in Coblenz also, but only if one of My Cities demanded Gingerbread at the time, i.e 74k population.

Which is why I try to be very careful about when I let a small City grow past 40k or a big City past 90k, I want the "right" (for my plans) type of industry to be demanded when that happens, so I can build my "right" choice or hope that the private investors will build the "right" choice for me. It takes very careful planning with what resources and goods you deliver or NOT deliver to a growing City, and WHEN you deliver or refuse to deliver them. It also requires care with using Promoters and in which Cities to use them. Sometimes it is also a matter of when I will have enough cash to build that industry when the new lot becomes available.

[Thineboot]

I'll instantly sell my beloved watch I'm wearing every day since... oh, wait, I don't wear one since... I don't recall... decades. In the old days I could have made use of such a smart watch and nowadays that they exist I still have my handy to look at. And each time you look at your atomic clock synchronized radio-controlled watch you always just get the past time anyhow. Your highly sophisticated backward running clock on the other hand would be able to give you the exact time occasionally. Don't wait, start your kickstarter campaign at once. You have all the arguments to make it a success.

As for IEEE-754, it still only throws an exception you have to handle. In math you can do it, too, but not in RE.
We should start writing down - whenever we may think of it - which new industries go live in early game under what circumstances. Circumstances like connected city, what other industries are not present on the map, and whatever doesn't come to my mind at the moment.