I’ve been struggling with the text for the Expand action. An abortive version of the current not-so-good stab follows:

The Expand action connects cities and foreign ports to railway companies.

General restrictions on Expansion:

• A player may only choose the expand action for a company if they own a share of that company

• Expand may not be selected for a company with insufficient funds

• Only one company may build a given route (dashed line) between cities

• Companies indicate the city connections and foreign port connections they’ve built with route markers

• The first route a company or merged component-company builds must connect to its home station. Subsequent routes may connect to the home station or to any city already connected by the (component-)company’s railway track

tables of initial and secondary home stations here

• Route building costs are paid from the building company’s treasury to the bank

• Routes may be built to the home station of a not-yet-operating company

• A company may not build more than three connections from any city, and may not build all the connections from a city with more than two connections. These limits are only checked at the time of building track

• The limits on railway track building change after the third General Dividend. Before the third General Dividend:

• Only a single track link may be built each turn. The cost is $5

• When track is built the building company’s income increases by the sum of the current values of the cities at the ends of the new track

• Track may not be built to another active company’s home station

After the third General Dividend the following are also allowed:

• A player owning a plurality of shares in a company may build two track links for a cost of $15:

  • The first track build must connect the company network to a city which is not part of the entire company network and the second track build must connect that city to another city which is also not part of the entire company network. In the case of merged companies the networks of all component-companies are considered for both builds
  • All track built with a double build must be for the same (component-) company
  • Connecting to another company’s home station or to London immediately ends the Expand action (and may cause a merger, see Mergers)
  • The company income is increased for each link as if each it were built separately

• Foreign port connections may be built if the company is already connected to their city:

  • Foreign port connections cost the price listed against them
  • Foreign ports do not increase the income of the bulding company but do increase end-game share value (see [sub:Game-End] Game End)
  • A route marker is placed on the port to indicate which company built it.
  • Foreign ports may not be built as part of a double track build.
  • Immediately after a foreign port is built the building company pays a Special Dividend (see Dividends)

• When building track for a merged company, all track built must be for the same (component-) company

• Track may be built to other company’s home stations. Connecting to another active company’s home station causes a merger (see Mergers)

Not the most brilliant piece of wordcraft I’ve done, but better than the exception heavy piece of paragraphed flat prose I’d before. I still need to make the language among links, connections, and track consistent among other changes.

The idea of different merger and floatation rules is growing on me.
Proposal: 1. All companies are limited to 3 bank pool shares each (!)

1. Merger shares continue to operate as before

2. After the 3rd General Dividend shares in the secondary companies may be sold in the normal manner with a minimum bid of $5

3. After 2 shares have sold a secondary company is active and operating with an initial income of $0

4. Companies merge by building into another other company’s home station, with a special dividend etc just like before

5. Secondary companies which start already connected by track cause insta-merges as before

6. The rule-of-3 still stands but secondary companies only have to honour it for their colour track – thus the NER may continue building links from Sheffield even though its component companies already have 3 links built

I kinda sorta maybe like the feel of this. Maybe. As I wrote, it is growing on me. It may be fungus.

I’ve also been a little annoyed by how to handle the plethora of bank pool and merger shares given the current merger rules. No answer seems good: cards, slips of paper, glass bits, cubes, chits, whatever. Time to retreat back to good old charts and tables and wet erase pens to mark off shares to players from the tracks:

Which required an update to the map to remove the pool squares. I also clean up the foreign ports to something that at looks like the values make sense.

I got talking with Aliza Panitz (BGG: Morganza) last night about the merger model in Muck & Brass and in the process realised that the first or second merger is liable to be too powerful. There are a few primary early merger paths:

1. Both northerns via the loop around Sheffield
2. One of the southern companies with the L&SR via Peterborough
3. One of the two northern companies with the EUR via Peterborough
4. Two of the southern companies via Peterborough or any of the other London orbits

The problem is with the mergers with the northern companies and with anything that involves the B&GR from the southerns. If anything in the north merges, and they start the NER, then it will auto-merge, thus generating a second special dividend for those players before the General Dividend. If anything involving the B&GR merges and they start the GWR (which the B&GR has also certainly built to), then the same problem repeats to the south. That’s 3 and a bit of the four early merge cases where a small subset of players get two special dividends in quick succession without any ability for the other players to do anything about it in advance.

That may not be good. Another related problem is the turn ordering which puts the cash richest players first. As soon as the mergers hit they are liable to remain the cash richest as they steamroller as above. To give a quick sense of the problem in concrete terms from this morning’s solo play:

The first expansion after the 3rd General Dividend merged the LB&SRC with the L&SR via Peterborough. The special dividend went to the LB&SRC. The NER was then started, merging the L&SR& (which now contained the LB&SCR) and the L&MR into the NER. A second special dividend was paid to the original LB&SCR share holders. The next expansion merged the B&GR into the EUR and started the GWR which insta-merged into the B&GR for another two special dividends paid to the B&GR share holders. The GWR then merged into the NER via Peterborough and paid a third special dividend to the original B&GR shareholders and a second to the EUR shareholders. They started the CR which was won by a player with so much cash they could simply throw it away on the CR and still win. The next player, who was already low on cash, had no chance at any of the merger shares, built a foreign port as did the last expansion. The game then ended on the fourth General Dividend as there were only two operating companies: GWR & CR.

An obvious partial solution is to require that a merging player pick a secondary company which:

• is not connected by track

or if that’s not possible:

• is not connected to the company which just merged

Such a rule is unpleasantly clunky, if historic and reasonable from a financials market perspective. It pretty much guarantees that the SWR is going to be the first secondary company out with the GWR and LNWR in the next round. The CR is a knife aimed straight at the L&SR and has all the subtlety of the Wabash in Wabash Cannonball. Likewise the The LNWR is a somewhat similar knife aimed at the L&MR and by extension at the B&GR. The SWR conversely lurks between the B&GR and the L&MR and can’t insta-merge in one build like the CR and LNWR can, but it potentially controls many foreign ports and is thus interesting for that reason alone. The NER is just a pig – whenever it starts it will isnta-merge into whatever the current shape of the L&MR/L&SR pair is. The GWR is a similar inst-merge into whatever the B&GR is then.

At least there’s now some subtlety to the mergers with the ugly delay rule.

Sigh.

Another potential model struck this morning:

1. Scrap all the current merger rules
2. Starting after the third General Dividend players may start capitalising the secondary companies. As soon as a share sells (two? three?) they may start operating.
3. Should a secondary company build into another operating companies home station (using its own track) the other company is absorbed, shares trade up etc and there’s a special dividend for the secondary company

ObNote: This model may unduely protect the LB&SCR and EUR and unduly weaken the NER. It may also require adjusting the rule-of-three track building rules.

The port pricing is still a problem but the base game seems to be cleaning up well. I’ve also made the art a little more functional, noting the starting stations and the like, fixing a few small track problems (eg Blackpool) and such forth. Nothing major.

Seth Jaffee commented approvingly on the BGDF chat on my making actions in Muck & Brass mandatory, unlike the optional actions in Wabash Cannonball. I replied that while I’d have a hard time phrasing a strong argument for making them mandatory, I was convinced that it was necessary.

Later in the resulting (brief) conversation it struck me: A significant portion of good play in Wabash Cannonball centres around controlling the game length in terms of General Dividends. For instance in a 3 player game selecting Capitalisation once without auctioning a share will often add an additional General Dividend to the game, doing it twice makes that near guaranteed, thrice and you’ll sometimes get a second additional General Dividend. As a result Wabash Cannonball has control of game length as a central challenge in the game. But control of game-length is not central to Muck & Brass so supporting a strong mechanism to affect game length would distract from the actual core foci of the game (network potentials, financial leverage, emergent alliances etc) and should thus be avoided.

ObAside: The sway point in Wabash Cannonball appears to be at 3 passed Capitalisations for 3 players as past that and games start ending more frequently from track cubes than they do from shares, thus requiring passing on Expand to gain further General Dividends.

Several small changes based on simulation runs:

1. Setup vs game start rules clarified
2. As each initial share is auctioned that company builds a free link
3. Initial turn order after auction is now cash-based rather than starting with the L&MR player
4. Action rotation clarified
5. Actions are now explicitly stated as mandatory (unlike Wabash Cannonball)
6. Foreign Links may not be built until after the 3rd general dividend. (not sure this is needed)
7. Expand action requalified to limit builds to not more than three connections out from any city, and not all the connections from any city with more than two connections. This allows the dead ends (like Thurso and Penzance) to be interesting as well as the many two-connection way-stops
8. Clarified cases of merger shares auctioned without bids.

A little more playable:

Quite a few changes this week.

I spent a few hours digging through the early history of the English railways from around 1830 onward and picked a more representative set of companies for the areas I’ll be putting them in. I also ordered them by starting date so that the initial companies did in fact start earlier as railway companies and the merger companies did in fact start later and heavily focused on mergers (eg GWR, LNWR, NER etc). There are a large number of goofs in home stations, but none I hope egregiously large (eg the GWR starting in Exeter?).

The current initial companies are:

• Eastern Union Railway (EUR, started 1846, Ipswich)
• Bristol & Gloucester Railway (B&GR, started 1840, Bristol)
• Leeds & Selby Railway (L&SR, started 1830, York)
• Liverpool & Manchester Railway (L&MR, started 1830, Liverpool)
• London Brighton & South Coast Railway (LB&SCR, started 1846, Brighton)

Merger companies:

• Caledonian Railway (CR, started 1845, Edinburgh)
• Great Western Railway (GWR, started 1838, Exeter)
• North Eastern Railway (NER, started 1847, Sheffield)
• London & North West Railway (LNWR, started 1846, Birmingham)
• South Wales Railway (SWR, started 1845, Swansea)

I’ve also had a first pass at allocating share counts for the initial and merger companies. Total gut sense with initial companies running from 3-5 shares and merger companies running 8-10 shares. I’m clearly going to have to play with those numbers.

The rules are looking pretty clean now. Not final but at least playable without instant gamer-pattern baldness.

A remaining problem is what order to auction the initial company shares in? It is tempting to do them in historical opening order which would be: L&MR, L&SR, B&GR, EUR, and finally LB&SCR. I’m not so happy about the very ordered progression from north-to-south that sequence represents on the map – despite the fact that the industrial furnaces of the midlands did actually drive the industrial and railway revolution in England. My sense is that the auction ordering should pose immediate spending challenges with early risks made in spite of later temptations. The three companies near London are obviously attractive – London develops beautifully and there are many fine ports to the south (the white bars) which offer Special Dividends – but having them come together as a group seems…wrong/odd. Likewise for the L&MR and L&SR with the wonders of Sheffield, Leeds and Manchester right next door. Urk.

Proposal:

1. LB&SCR
2. L&SR
3. EUR
4. B&GR
5. L&MR

Of course that’s all higgledy-piggledy for historical dates, but it may work better. I’ll have to sleep on it.

Over the last couple days I moved the map from yEd over to Inkscape and into SVG format. Unfortunately this required a nearly from-scratch redraw as I lost patience writing the script to post-process yEd’s concept of what it should produce for SVG elements versus what I thought I wanted. Fortunately for such a simple graph this did not take long. A few colours here and there, an income track, yada yada and voial!

Note to self: Must remember to put numbers on the income track.

The bit about the active player choosing which side of the merger pays the Special Dividend could be a doozey. It allows minor shareholders with similar holdings to collude across turn order in order to cause mergers which mutually benefit them but not necessarily the plurality shareholder. Consider:

PlayerX owns a significant plurality of CompanyQ. PlayerY and PlayerZ both own but a single share in CompanyQ. CompanyQ is two single builds away from merging with CompanyR. Both PlayerY and PlayerZ have significant holdings in CompanyR. CompanyR is too far from CompanyQ’s home station to consider the merge that way. So, PlayerY builds one of the links toward CompanyR and PlayerZ builds the second and as predicted by PlayerY, selects CompanyR for the Special Dividend. PlayerX with no investment in CompanyR is of course overjoyed to be shut out of the Special Dividend. PlayerX would have far preferred to build the connection himself and declare CompanyQ for the special dividend but was outflanked by the emergent collusion among PlayerY and PLayerZ.

Other forms of possible insurrection and insurgency by minor share-holders abound. I’m not clear if this is delightful or merely fatal yet. In terms of layering the incentive mesh formed by share-ownership however, this is quite delightful.

The new more historically flavoured merger rules read as follows:

When a company connects to another active company’s home station the two companies merge. A merged company may have several home stations from its constituent component companies. The active player decides which of the two companies will be acquire merge into the other acquiring company.

The acquired company pays a Special Dividend (see Dividends)

1. Any unbuilt track markers for the acquired company are added to the acquiring company’s supply

2. Each player’s shares in the acquired company are replaced 1:1 with shares from the new parent company. Shares are initially taken from the bank pool of the acquiring company, then shares from the merger supply for that company. The shares of the acquired company are placed face-down in the bank pool of the acquiring company so that it may be clearly seen that it is now a component-company.

3. The income of the acquired company’s income is added to the acquiring company’s income, and the acquired company’s marker is removed from the income track

After the merger is resolved the active player selects an inactive merger company to start. If the new merger company’s home station is already connected by track, the connected companies will be acquired by the new merger company.

1. Any unbuilt track markers for companies that have connected track of the home station of the new merger company are added to the merger company’s supply

2. Each player’s shares in the companies that have connected track to the home station of the merger company are replaced 1:1 with shares from the new merger company. Shares are initially taken from the bank pool of the merger company, then shares from the merger supply. The shares of any acquired companies are placed face-down in the bank pool of the merger company so that it may be clearly seen that they are now component-companies

3. A share of merger company is auctioned in the normal manner (see Capitalise). The new share is taken from the merger company’s merger supply if a share is not available from the bank pool

4. If the merger company’s home station is already connected by track, the merger company pays a Special Dividend (see Dividends)

5. The winner of the newly auctioned share selects the next merger company to start, starting the merger process all over again

If no players bid on a merger auction the auctioning player discards the share into the bank pool and receives the current income of the company divided by the current number of issued shares including the just auctioned share in exchange.

If all the merger companies have been started and a merger occurs, the active player may either pass on the rest of their turn, or may auction any available share from the bank pool in the normal manner (see Capitalise).

Quite a bit cleaner. Conversations with Ben Keightley (Coca Lite) have caused me to re-examine this simplified model, though I don’t think that was his intention. In particular I’m re-considering the automatic chaining of mergers. I’m not sure it is justified. Without automatic chaining the mergers will still tend to happen in a flurry, that is where the money is, but it will not be quite as uncontrollable an orgy and allows for interesting decisions to be made between mergers (eg track blocking) which are not possible with automatic chaining. The specific relationship of turn order and share investments also becomes more interesting. In short methinks I’ll lose the automatic chaining of mergers. I see little loss and much potential game.

The other change, and this is a biggie, is allowing the active player to determine which of the sides of a merger will pay the Special Dividend. This change was done on a whim but it feels right. It allows minor share-holders to merge foreign companies into a company in which they hold more stock, thus getting the Special Dividend where they want it while also building the agglomerate.