Re: thoughts on 1820

[J C Lawrence]

This has been sitting in various stages of response and thought and
rewrite in my drafts folder for too long.

The original premises of the game included (not in priority order):

- Infinite companies.
- Full capitalisation.
- No privates.
- No train rusting, but train maintenance fees.
- Companies start as non-viable.
- Continuous take-overs, hostile/forced, friendly etc to make
them viable.
- Changing company issue sizes both up and down, including
going private.
- Dividends smaller than their percentage of income (money in
treasury is worth less than money in hand).
- The game starts rich and becomes increasingly poor.
- Only a small few companies will be profitable in the
end-game (less than one per player).

Stylistically the assumption was a game which married the primary
qualities of Rolling Stock with the primary qualities of 6-player 1830,
but at smaller player counts.

http://rabenste.in/rollingstock/

In this line I wanted to support the following core strategies:

- Solely liquidating companies and never actually running a
public company, just buy and flip, buy and flip.
- Taking your public company private and focusing on
portfolio building.
- Running an acquisition monster with an ever-increasing issue
that simply buys/takes-over other player's companies, has
ever more tokens clogging the board and thus dominates the
market by volume and control, aka the Cisco monopoly model.
- Briefly running loads of companies and crashing them into
each other, liquidating the failures and constantly
recycling, taking a vig on every transaction and rarely
keeping anything; constantly living in the margin on the
current float.
- Or just playing a standard 18xx-style mix between dividends,
portfolio game and managed assets (presidencies).
- Mixtures of the above.

Also in there were ideas around requiring and rewarding accurate
player-incentive manipulation and prediction and that much of a
player's position wouldn't be their own creation but would be the
product of their management of the other players through the choices
they presented them (not jaw-flapping diplomacy but moves-as-offers).

Not all of those initial intentions have worked out, but that's where I
started.

Parliament rounds
=================

The parliament rounds are noisy and marginally unfair by specific
intent. Their purpose is six-fold:

- To be coarsely unpredictable.
- To be a significant drain on the game's money.
- To strongly differentiate positions,
- To strongly reward accurate player prediction.
- To strongly reward accurate player incentive management
leading up to the Parliament round.
- To force players to commit where they don't want to (and to
thus be inefficient).

Price-control and price-finding are not among the goals. While I
agree that players may find them important, I don't see them as
necessary to the game. If players wish to price-set, then they pay for
that and become part of the same system they're setting prices for. If
they don't wish to do that, then they also forgo price-setting.

Cost & Timing of grow-ups
=========================

The auctioned cost of growing a company is mostly incidental.
Functionally it ranges between $20 and $15+($5*N), so a $10-$20
variance depending on player-count based on the player's position in
priority (which they can significantly control). That's not a
significant sum of money in the game, the variance is functionally
irrelevant, and it keeps the company grow-ups out of the stock round and
thus the efficiency and commitment gains of the Parliament Rounds
actually in the Parliament Rounds. More specifically, allowing company
grow-ups in the Stock Rounds allows for a raft of abusive/degenerate
tactics that over-power the rest of the game.

Allowing companies to both grown and shrink in I&T rounds is where the
game started. In short, it is far too efficient, rewards early money
too strongly. almost completely removes the relevance of the parliament
and stock rounds and is generally a Bad Thing. Coarsely, it means that
the early successful companies just go from success to success and take
over the game.

Efficiency, especially player efficiency, is not a design goal. I
understand that player's may wish to be efficient, and that's fine, but
that's not my interest. If the player's wish to succeed, then they are
going to have to do so despite the many inefficiencies and compromises
forced on them by the game. Managing that is part of the game.

Limiting grow-ups has been on the table for a long time. Initially
companies could only grow by one step per round -- and that was so slow
as to force a very narrow set of viable strategies: public companies
had to grow in every PR or die. The current pattern of limiting only
private grow-ups keeps the privates from being entirely dominant and it
allows players to play cash-flow and timing games in the mid-game to
good effect. And there's some interesting decisions and viability
between going 10-20-50 or jumping say straight from 10 to 50 shares.
My current impression is that making the OR-sets shorter will increase
the viability of that variability, and that seems a Good Thing.

Train maintenance costs
=======================

Yes, maintenance costs are harsh and maintenance costs jumping up
in the middle of an OR are especially harsh. By the end of the game
maintenance costs dominate the entire game to the point of making some
(many?) companies unable to cover their own maintenance even if they
withhold all dividends. Maintenance costs are the central challenge of
surviving the mid and end-game and thus predicting and managing them
accurately is a core problem.

As to the specific of the maintenance costs leaping forward: If you
didn't want to incur the risk you could have discarded the train when
the company ran. But you didn't and so now you pay. That's one of the
costs of getting your predictions wrong (and of playing with the
efficiencies of lower stock prices).

2-Share companies
=================

2-share companies are mostly, but only mostly, irrelevant. What
2-share companies do is present an unpleasant way-stop that nobody
likes, but which exerts pressure and cost. It means that shrinking a
company to private takes a bit longer as you have to go through
2-share, and that when growing a private into a public company, the
2-share way-stop means your was-private is unavoidably available for
everybody to take-over at a very modest cost.

You don't think someone will try to take-over a nice 2-share company
with great tokens/board-visibility and a semi-decent train for a
minimum bid of 2x a modest stock price? We've seen it several
times...and we've seen Shelby manage to take his privates public in
just such a way that nobody wanted them...which was a neat trick.

That said, I only partly agree on the 2-share-and-a
single-certificate bit. The problem is that adding the extra piece of
paper has no effect when shrinking the company, and only presents a
marginal-utility and mostly-false choice ("Do I buy the second share?")
while also making it cheaper to grow up a private. That optional
extra efficiency and adding a marginal utility choice which is only
actually relevant in less than 10% of cases just isn't compelling for
adding another bit of paper.

Income tax
==========

Income tax was part of the initial conception of the game: that it
would be much easier to get money in treasuries than it would be to get
money in-hand. Or, in a specific (imagine 50% income tax for this
example), that using player cash to pay for trains would be much more
expensive than using treasury, thus adding an additional problem to the
game of continuously managing cash reserves and cash-flow through
treasuries.

But, while I think the idea is sound, its presence in a game in which
train limits are high and where most companies spend most of the time
at train limit is...not. If this were a fast-trains game in which
companies mostly owned few trains and keeping a train, any train in a
company was a major question, then income tax would make more sense, but
1820 is not that game.

So I'm taking out income tax. That will significantly alter the
relative strength of privates, and in part the move from 3-OR sets to
2-OR sets is to offset that. Which is a really big change...and
one that worries me.

Cost of growing up
==================

Adding friction to the grow-up process has also been on the table for a
long time...and I'm mostly of the mind that there is already too
little movement in company sizes already, and that's given as easy and
cheap and rewarding as it is now. Adding more friction doesn't seem
like it would create better behaviour.

BtB The current growth-system is vaguely consistent with upsells by
startups in the modern market: all the current share-holders have the
effective size of their shares shrink, but the company raises cash so
the net-value of the shareholder's shares stays the same (they own a
smaller fraction of a company which is now worth more). In this line
I've long considered having company treasuries paid out of
shareholders at the end of the game, but haven't as it encourages rather
weird/degenerate end-games.

More broadly, yes, in a different game based on small money finesses
adding such marginal costs and trade-offs (10%, stock price, etc)
could be interesting. Especially in a game where stock price was more
valuable and where fine-grained player cash was more constrained. But
this is a game where players end OR-sets with a thousand-plus
in their hands... 10% or a stock bump ...really doesn't make much
difference in this game, and adding such a constraint would seem mostly
make-work noise. Maybe in a different game...

Cross-investment
================

Cross-investment is a funny thing. We see it in most 18xx as
there's a limited number of things to invest in, the money avilable to
invest is even more tight, and the game requires large block capital
costs for floating new companies (and there's a limited number of new
companies): so left over money that doesn't fit into such block
investment of course goes to cross-shares. Both those things are not
true in this game. Here there's affectively an unlimited supply of
things to invest into, an almost equally unlimited ability to spend
that same money improving your own assets, and take-overs could just
remove any cross-invested shares and quite possibly at a significant
loss at any moment. With or without income tax, cross-investing looks
pretty dubious for much of the current game.

I'm not surprised there's so little cross-investing.

That said, there's a fairly rich space of strategies in 1820 which
involve the player never floating and running public companies (or only
floating them to then sell/dump them into some other player's public),
which we've barely investigated yet (and so far its only been me). I
expect there's a really rich space of viable strategies there.

The basic premise of such strategies roughly falls to building a
killer private, running it at breakneck speed and constantly buying
and discarding trains while using your side-cash to buy up side shares
in player's public companies. The result is that the trains move
so fast that other player's companies are constantly off-balance
and cash-starved and your private cash-flow keeps you ahead of the
curve. Then when brown hits you make your last lump of money and close
the private (rather than selling it and letting some other player
get even more benefit from your tokens). And then you throw all your
cash into the best public companies and ride into the sunset. This can
be a notably strong, albeit untraditional, strategy.

BtB Now that I've removed income tax and made OR-sets shorter...I'm a
bit worried that too many of those no-publics strategies are now
unviable. Having those strategies be viable, or a reasonably large
number of them, is critical for me.

The backing inspiration for the strategy family BTW is the trader versus
builder roles in Rolling Stock:

http://rabenste.in/rollingstock/players_guide.pdf (section
4.2.5)

Loans
=====

Loans are one of the ways that players can make their companies
difficult to take-over. They're not the only way, but they're one
way...and it comes with a cost. They also allow for
hostile take-overs and thus increased dynamism and volatility which
would otherwise not be possible. This latter won't happen in every
game, but it can/will in many (and has in ours) as players go for the
efficiency advantages of running the edge of take-over territory while
trying to only just keep safe.

More broadly loans remove the arbitrary precise edges in pricing while
providing a more fluid system for modest sums than recapitalisation
($100 really isn't much). Loans move blocks of $100.
Recapitalisation moves blocks of $thousands. Now the shorter OR-sets
will decrease that utility for loans, but I don't think by a lot.


As for interest rates, in the period the game covers there were two
major recessions (really three, but the third was not such a huge deal),
with one of them large enough to result in people starving to death in
the streets. Meanwhile the railway industry progressively collapsed,
failed, collapsed and failed, and was rescued (and that only barely) by
the government doing forced bailouts and mergers of companies into ever
larger and larger (and more troubled) companies, finally resulting in
the great merger that formed British Rail at the start of the last
century.

Yeah, having interest rates climb (and more importantly having
maintenance costs start to dominate and then exceed company runs) during
that period is off in detail but doesn't seem too far off as an
expression of the capital costs of the time.

Future trains
=============

The future trains...are interesting. I'm a wee bit biased, but I
think they're brilliant. Buying them reduces your ability to influence
their value (and provides the tools and incentives for the other players
to make them less valuable). Not buying them gives you more control
over the game's progression, but less security in your position. That
seems a pretty interesting choice.

Daniel has recently been experimenting with buying both future trains,
and I think has generally concluded that that is often a Bad Idea. Prior to Daniel's experiment, we mostly saw them going to
different players.

More broadly future trains drain significant sums of money
from the game at key early-game junctures, thus putting a damper on the
explosive value of early money and providing time for players to
establish more viable positions via differing strategies before the
game flies off the handle.

-- JCL

On Mon, 28 Mar 2016 18:02:55 -0700, Jacob Butcher <jb2...@nowhen.com>
wrote:

> A very interesting game. All worthy ideas. I suggest discarding
> some of them to streamline the game. IMHO corporate growth &
> shrinkage is the most interesting part of the game. It doesn't seem
> quite right. I would focus on improving it.
>
> The following notes are only lightly edited. Only constructive
> criticism intended.
>
> parliament phase:
> - I suggest auctioning startings spots directly, similar to 1830
> privates.
> - I see no reason to auction off the right to grow a company.
> - Just let players do it or perhaps charge a flat fee or percentage.
> - Consider letting players grow companies at the same time they may
> choose to shrink them. Perhaps only capitalize the additional shares
> in multiples of 5. Actually, that sounds awesome. Needs details for
> smaller companies, but getting up to the next multiple of 5 (once)
> immediately rather than waiting for the SR seems workable.
> - Consider limiting growth to fewer steps at a time?
>
> maintenance:
> - Consider choosing between the maintenance cost and discarding the
> train.
> - Increasing the maintenance charge on companies that haven't run yet
> is pretty harsh.
> - Maybe a separate maintenance phase at the beginning or end of the
> OR?
> - Maybe grandfather the old maintenance price until the end of the
> OR?
>
> cost of business chart:
> - The big jumps are weird. Consider ensuring a steadier progression.
> - Tossing all 2 and 2+ trains at certain points feels arbitrary.
>
> takeover:
> - Consider eliminating the bidding rule.
> - It discourages interesting shenenigans because of possible loss
> of control.
> - It basically emphasizes that the 18xx stock valuations are too
> simple.
> - Consider dropping paying for "extra" tokens. That just seems
> arbitrary.
> - I think the token cost rule could be described as simply: First
> token free, second token $40, all other tokens $100.
> - I suggest either allowing a company to take over another ccmpany
> regardless of connectivity or always requiring connectivity.
> Requiring it depending on ownership seems weird.
>
> tax:
> - Omit? I don't understand the motivation for taxes.
> - Consider charging on all runs, before withholding or dividends.
> - Deduct interest or maintenance or both?
>
> private companies:
> - I think having private companies avoid taxes is probably a mistake.
> - Similarly, having 2 share companies that can be solely owned by 1
> players is weird. In practice a 2 share company is an odd mix of a
> private company and a 5 share company. I think it would be better
> for public companies to always have shares that cannot be controlled
> by the president.
> - I suspect private companies should not have shares that change in
> value.
>
> growing/shrinking companies:
> - I suggest that issuing additional shares should have consequences.
> - Drop in share price?
> - Discount to nominal share price? Or possibly 10% loss to
> capitalization? (The latter idea seems really interesting.)
> - I suggest that reducing the outstanding shares is just weird.
> - Maybe just eliminate the concept.
> - Alternately, acknowledge that it's just a way to forcibly buy
> shares. Pay a premium for them?
> - Maybe it should be possible to change a company's share value while
> growing/shrinking it, or perhaps independently?
> - Increasing the share value requires adding capital to make up the
> difference on outstanding shares? (Combines well with the 10% loss
> to capitalization idea.)
> - Maybe a company should have 2 independent abilities:
> - Change the number of shares without changing the company's market
> cap.
> - More shares means less value per share.
> - Fewer shares means more value per share.
> - Change the company's market cap:
> - More cap requires adding cash to the treasury, increases share
> value.
> - Less cap allows taking cash from the treasur, decreases share
> value.
> - I like the increase shares/cap idea, I don't think I like the
> decrease shares/cap idea.
>
> share locations:
> - Nobody ever sold a share.
> - Nobody ever bought a share of someone else's company.
> - The difference between IPO shares and market shares never came up.
> I prefer that shares that have capitalized never pay dividends to the
> company, regardless of location.
>
> loan:
> - Omit? Loans and additional capitalization serve much the same
> purpose. Having both in one game really may not be necessary.
> - I believe that historically loan interest should decrease over
> time, not increase. What is the game purpose for having it increase?
>
> future train purchase:
> - Omit?
> - Don't give both to the same player, certainly not if that's what
> typically happens. I suggest just selling them to the 2 highest
> bidders for the second highest bid.
>
> ~jacob
>



--
--
J C Lawrence They said, "You have a blue guitar,
---------(*) You do not play things as they are."
cl...@kanga.nu The man replied, "Things as they are
http://www.kanga.nu/~claw/ Are changed upon the blue guitar."