Time for a new board game design attempt. This time I'm going to try documenting my design process, mostly to motivate myself to actually finish the design this time. The inspiration was the recent spate of lighter variations of the 18xx system, none of which really does what I want. If you're not familiar with 18xx, this post probably won't make much sense, and you might as well stop reading. Sorry.
Strip away everything possible from an 18xx game except for the train rush, but try to retain the feel of the rush. Lokomotive Werks tried to do something along these lines, but I didn't really care for it for various reasons.
Aim for a 90-minute game for the first play.
Fully deterministic with no hidden information.
It's a plus if the mechanisms make more sense for something that isn't train-themed. (Some people I play with have an allergy to trains).
First decision that had to be made was whether players represent companies or investors that might control several companies. The latter seemed much more likely to create interesting turn order and timing issues. Features like collusion between the same player's companies are pretty central to the way a 18xx train rush unfolds. And the possibility of controlling multiple companies also creates the delicious dilemma of companies being both assets and liabilities.
But having companies be separate entities has the problem that the game then also needs to keep at least a limited form of share dealing in the game. This is in conflict with both of the first goals.
After a couple of days of doodling around with the game I hit upon the idea of keeping the companies, but totally gutting the 18xx stock round. The only thing that you can do on a stock action would be to start a company: set the par price, buy 50-100% of the company, which will then receive 10*par price as initial capital. No selling or buying of shares. The latter would allow cutting out the stock market entirely.
For operating companies I absolutely did not want any kind of board play. Again the goal is to cut ruthlessly on the first pass. The simplest possible solution would be to have each train produce a constant income every turn until it rusts (e.g. Railroad Barons), but that felt a little too spartan for me. There should be something players can do to differentiate two identical companies.
My first stab at this is that train won't necessarily produce income every turn. Instead when you first buy one, you put 1-3 tokens on top of it. When the company operates, flip one token on each train. Any train that has all the tokens on it flipped produces income that grows non-linearly in the number of tokens. For one token $x, for two tokens $2.5x, for three tokens $4.5x. After a train has produced income, populate it with new tokens (can be different amount from original).
This seemed like a system with some potential: mechanically still very simple, but offering some interesting decision space when it came to the number of tokens you place on a train. You'd much rather prime a train with 3 tokens once than with 1 token thrice. But maybe you need the money sooner (or at least an option for having the money). And how long will the train live anyway? Somebody being too optimistic with their scheduling might even prompt other players to short-schedule their trains, withhold the money, and accelerate the train-buying.
As a bonus this also satisfied goal 4. This mechanism makes no thematic sense with trains, but the whole "small return safely and quickly" vs. "large risky payoff later" should have some resonance with other themes. The first couple of possibilities that popped into my mind were technology companies (where the "trains" would represent R&D projects) and trade fleets in the Age of Exploration. I'm sure there are more.
Other than that the structure of a single 18xx operation round seems ok. First run trains, then give a dividend or withhold the income (though no stock price effect), finally buy trains either from the bank at face value or from another company controlled by the same player. Every company must always own a train. If it can't afford one, then the controlling player will pay for one (this might not work in practice given shares can't be sold: it might be too easy to go bankrupt). There should probably also be a train limit.
I figure the game should have a constant alternation between an SR and sets of 2 ORs, and since it's explicitly a train rush game, using either the 1844 or 18FR-RCE rule for the foreigners buying a train from the bank at the end of each set of ORs.
That was enough for a first draft, so I plugged in some numbers. The train schedule and prices were taken from 1856 and made up an arbitrary payoff table for the different kinds of trains. Finally I set a par price range of $40-$100 and started the players with $300 (so the first choice of the game would be deciding which of 50-70% to go for).
I soloed half a game with these rules, and the game seemed to have the right kind of feel. Both the decisions on how many trains to buy each turn and how to run them seemed non-obvious, and there was the same kind of wheels within wheels planning.
The numbers I'd chosen were of course badly off, but that was to be expected and can be tuned. There were a number of worrying issues though:
There probably isn't a runaway winner problem, but there might be a spiral of death problem. If someone ends up running a company with fewer trains the opponents, they don't have the usual possibility of investing into the opponents' companies to boost their income. They are also unlikely to be able to found a new company very soon, and are thus much more likely to have dead cash after an SR. Possibly it should be possible to sell shares (down to 20%?) for the original price as a stock action.
A totally fixed company operating order is too restrictive and predictable, there needs to be some way to manipulate it. I think this should be done in the SR, ideally making stock priority more important. Perhaps something along these lines: there's a linear turn order track. A company starts on a space based on its par price (there's probably a couple of extra spaces between each par price space). As a stock action a player can move one token up or down on the track. This costs money: the nth token movement by any player on a SR costs fib(n).
I also suspect some company differentiation might need to be added later on. The current idea is that some companies get bonuses for running trains of a certain level.