When I went to college, I got my first Windows PC, and one of the first games I played on it was Railroad Tycoon II.
I tried the campaign, which was a collection of scenarios basically, but most of the scenarios were time limited, with small goals and a local area where you were building your railroad network up. No, after playing the early campaign scenarios a few times, I spent vast hours playing one scenario in particular. I call it a sandbox scenario, because it was a scenario that shows what the main campaign of the game should have been like.
The scenario was called North America, and it covered Mexico, Cuba, the US and Canada. It was as if they had all been unified into one country as you didn’t have to worry about gaining the right to build railroads and stations in each of these countries. The European Scenarios seemed to focus much more on negotiating with, say France, to have the right to extend your german railroad company’s line into French territory.
But in the North American scenario, you could run your own railroad company over this huge landscape, all the way from 1830 to shortly after 2000. I never made it further than about 1920. This sandbox scenario was basically the nirvana for someone interested in operating a rapidly growing business sprawling out of control.
Railroad Tycoon II Platinum became available on Steam a few years ago. I had bought Railroad Tycoon III when it came out, and promptly stopped playing it when I discovered you couldn’t play on a big map with a long scenario length. I think the scenarios were limited to something like 20 years. It just lacked the pure thrill of the North America map in Tycoon II.
The thing I liked about North America scenario is it was a wide open sandbox, but with financial constraints, and also performance evaluation. It really did a great job in simulating “What have you done for me lately” shareholders and financial analysts. You could build up a monstrously wealthy corporation that was basically printing money by the bushel, but eventually it became hard to keep up the growth rate when you already had so many lines laced throughout the continent.
Immediately after I purchased Railroad Tycoon II on Steam, I begin building the most grandiose railroad corporation I had ever built in the game. I started my corporation on the basis of an Jacksonville to Tampa, Florida railroad line.
Before long I had connected to Miami, and then into Louisiana, and so on. Previously, when I had played this scenario, I tended to start in the midwest, and I never really built anything in Mexico or in Cuba. This time, I became a Cuban, Mexican and Southern united states powerhouse.
I always used to keep all of my lines connected to each other. But this time, I kept looking for opportunities elsewhere, and building independent networks there. This helped to keep congestion on my lines down, and let to great profits.
I also would never build a line to a city that was already occupied by one of my competitors. But this time, I decided that big cities were lucrative enough to share. Having a large population center as an origin and destination for passenger service was hugely helpful. The main problem with a small rail network is that passengers can’t get to popular destinations, and short routes don’t pay much.
Eventually, I expanded all over the map in a spidery way, sharing rail service with my competitors in many ways, and setting up San Francisco to Miami routes, or even Mexico City (all mine) to New York City (competitive).
Making money from passenger service, and providing restaurants and bars at stations, as well as eventually putting dining cars on your passenger trains was the easy part. The difficult, even baroque part of making money in the North American scenario is the only way real railroad companies make money now: hauling freight.
There’s a lot of considerations when hauling freight. You have to study and understand the entire industrial model in the game, and try to figure out how much demand there is for ingredients, and how degraded your rail network will be from traffic.
You will end up with main interstate rail routes which get bogged down. After all, you can’t really make a highway for trains, and your freight routes will be competing with and slowing down your time critical passenger routes.
Milk can’t sit around in a train car for very long, but steel can, and eventually, making stuff gets vastly complicated as steel, canning facilities, bauxite mines and many more enterprises have to be connected.
A steel mill needs coal and iron, and often the nearest coal is in the appalachians while the mill is in mexico city and the oil is in texas. This can take a lot of time and coordination, and after steel mills show up, the complexity begins to ramp up even more (automobile plants, rubber, aluminum, etc.
All of this complexity gives this game a lot of sandboxy realism, but it lead to the collapse in my ability to optimize or even maintain my railroad empire. At one point, I had over 160 trains to keep track of, repair, maintain and replace, and in addition I had to determine where else to expand, and what freight or passenger areas were underserved, and where line congestion was leading to big problems.
I would find out that a station was not sending out its raw materials in a timely fashion, and then order two more trains to make deliveries, only to find out that the original problem was that the trains I had purchased to make deliveries in the past were getting stuck in the middle of massive traffic snarls. Added more trains just added to the problem.
Eventually I sat down, paused the game for thirty minutes and managed to pare down the number of trains to around 120. I began to act like a modern manager, getting rid of trains who showed a negative lifetime profit. But this had knock on effects, in that often the unprofitable trains were supplying intermediate goods for more profitable manufactured goods transporters.
It did help overall though, as the huge snarls around my main hub in mexico city started to easy. I had so many steel mills in mexico, and so much need for coal and oil from elsewhere, as well as a steady thrum of produce, coffee and passenger travel, that the whole Mexican region was in a state of lockdown, as more trains kept pouring into the region, while those that needed to get out struggled in the traffic jam.
One thing my massive corporate empire did show me is that Mexico City can be an excellent hub for all kinds of railroad traffic. The main problem is that a lot of the routes through mexico have to deal with steep grades. Steep grades can cripple travel speed for all of the 1800s trains, especially the earliest ones.
One of my classic tricks, which I used in this case as well, is to take all of the money your player has collected from CEO pay, sell all of your shares in the corporation you’ve been running, and with that jumbo pile of money, start over with a new corporation which you own outright.
I started a new corporation, but this time, I had already exploited most of the easy opportunities on the map. I really didn’t have enough opportunities for the capital I had invested into the new company, so I doubt I would get a good return. On the other hand, as soon as the computer took over my old corporation, they started running routes like computers always do, and the bad AI shined through. They were losing money hand over fist by sticking to the basic routes that the AI used.
I think the lesson that the N/A scenario has is that as any corporation grows, the amount of information that people at the top need in order to make wise decisions increases. But no one can handle an unlimited amount of information, and you can only delegate to a certain extent. Obviously, as a single person playing the game, you hit the limit fairly quick. I often set up a few stations and some new train routes, and then let them be, moving onto the next opportunity. I was the growth guy, but I couldn’t also pay attention to how well the current lines were working. As you delegate, you gain a more and more filtered view of actual events, but you have to delegate because there is simply too much information to consider.
It makes me wonder if an AI could be built to design and maintain these railroad networks in the artificial, simplified case of a video game. I tend to think it would take a highly evolutionary, experimental sort of AI to take on the challenge. But maybe something in graph theory would shed light on the trouble. Also, I’m sure today’s actual railroad companies have insight on this issue.
Anyway, I started thinking today of giving the N/A scenario another shot, but this time thinking about the optimal way to build and maintain a continent wide passenger-only railroad corporation. By constraining the design to just passengers and postal mail, I feel like I can test things more thorougly, and I believe with Railroad Tycoon’s design, avoiding low profit freight in pursuit of 19th century high profit passenger travel will give a network that is both less congested and more profitable than a freight and passenger combined network. And I think I can give the corporation a faster rate of return by not holding onto so much cash, like I did last time. I will borrow aggressively as well, because of the great rate of return on passenger line investments.
One of the things I want to discover on this blog in future posts, in the optimal way to design a passenger network, and the optimal set of train routes and station stops. How do you balance short distance trips with long distance trips. I’d like to find some information on the Railroad Tycoon II model to see if it seems at least semi-realistic in terms of where passengers want to go.
Some of the corporate management bits of the game are rusty. When my last corporation was making money hand over fist, the game wouldn’t let me provide a special dividend to get rid of some of my excess cash, like Microsoft has done. I was holding onto about 10 million in cash that I hadn’t found an investment opportunity for. 10 million dollars back in 1895 was a lot of money to be holding onto. I was making more money than I could plough into reinvestments or provide as dividends because of the too strict rules on dividend payouts.
So when the network gets to that stable point, I won’t be able to actually provide the kind of dividends you would expect a railroad to provide. I’ll just have to calclulate them afterwards based on the level of cash accumulation I obtain.
There is also a whole stock market side to the game that I’d like to play around with sometime, and if I master the passenger rail corporation, I’d like to tackle the much tougher freight only side of things. Maybe once I get a perfect passenger rail network, I’ll start a new freight only corp with a massive bundle of cash to see what I can do. The capital you get at the beginning of a new game isn’t enough to build a proper freight network.