Galaxy Investment Strategy and Logistics

From: desj@ccr-p.ida.org (David desJardins)

Date: Wed, 06 Jan 1993 01:31:24 +0000

In article <28967@oasys.dt.navy.mil> curt@oasys.dt.navy.mil (Curt Welch) writes:
>> To be perfectly frank, if you are making these basic decisions in 5
>> minutes, you are doing a pretty lousy job.  How do you decide exactly
>> how much of each resource you need in each place?

> It's much better to fully develop a few planets than to partially develop
> lots of planets because a planet costs the same amount of money to defend
> no mater what size it is.  So the five minute decision is picking which
> planets to develop next.

This is so wrong I hardly know where to begin.  In Galaxy version 2.X,
the single most fundamental principle of growth is that one does much
better to colonize many different planets rather than focusing on a few.
This is of almost overwhelming importance.

In version 2.X, the primary limit on growth is shipping capacity.  A
single home planet can produce over 150 cap per turn, but it can only
produce about 10-15 cap/turn of shipping capacity per turn.  (For
example, if you build 46 transports with design [1 0 0 0 1], and are
colonizing planets about 10 light years away, then a round trip takes
three or four turns and you can ship at most 15 cap per turn.)  By the
time you have built up enough transports to be able to fully exploit
your home planet, you have new production planets coming on line, and
you also have to travel farther to new colonies, so it's a very long
time if ever before you "catch up" to all the shipping you need.

It is thus *much* more efficient to produce capital on your colony
worlds, where it converts directly to industry, than it is to ship
capital out.  Your goal should be to produce almost all of your capital
at a place where it can be used.

If you heavily colonize just a few planets, most of your capital is
being shipped.  You ship a few hundred capital to a world, and it
produces perhaps a few hundred more, and it maxes out.  Then you have to
start shipping capital to another world.  But if you spread your
investment over many planets, then you can invest just 50 or 100 capital
in each, and have them create all the remaining industry themselves.  If
you start with the same amount of investment, you end up accomplishing
much more.

It's true that you have to defend more planets.  But consider: (1) you
are probably going to defend all of your planets, big or small,
developed or not, just because you don't want someone else seizing them
and using them as bases against you; and (2) if you really do get into
an early war you are probably doomed in any case, unless your attacker
is incompetent.

>> And exactly what ship
>> designs will be most valuable in the long run?
>
> You can't calculate this because the correct answer is based on
> knowning, or guessing what the other players will do.  There are some
> stupid things you have to learn not to do, like building a drive-only
> ship with more than one drive.  But once you learn the basic correct
> stratiges, then the choice is a matter of taking a random guess at
> what the other players will do.  Making that guess only takes me 5 or
> maybe 10 minutes.

While your comments on industrialization were simply erroneous, this
comment has slightly more merit.  Nevertheless it is a gross
exaggeration.  Do you really think that all games of hidden information
are pure guesswork?  I would like to play poker with you sometime,
preferably for very large stakes.

I can't get this image out of my head, of Eisenhower saying to his
aides, "We don't know the exact positions and strengths of all of the
enemy forces, so let's just flip a coin to decide where to attack."

>> There's a reason that I have the largest industry in each of two games I
>> am in (Clarke and Alpha).  It's because I take care with these decisions.
>
> I doubt that's the entire reason.  If you have the largest industry,
> then you probably have the smallest fleet (depending on how far along
> the game is) That's just the gamble you have chosen to take.  If two of
> your neighbors decided they didn't like you having that much industry,
> and launched a joint attack, you would most likely be dead.  You have
> just decided to gamble on them not doing that.

There is absolutely no question in my mind that if you play as you have
described, and I play as I have described, that I can have both a bigger
fleet, AND more industry, AND higher tech levels than you, throughout
the entire game.  If I am using my cargo ships much more efficiently
than you, I can afford a larger investment in defenses as well.

It is indeed true that if two players team up against one fairly early
in the game, the one player is in a lot of trouble.  That is only
natural for a game where early growth is relatively slow.  In fact, I
would assert that if one player attacks another early in the game, and
both are competent and neither caught by surprise and defeated
immediately, then both are likely to be out of the game.  It just takes
too much out of your investment capability to fight an early war, even
if you "win."  (Ask Snafubar in Clarke about that!)

> And at this point of that game (move forty something), the outcome is
> going to be determined by what teams get formed, and what they decide
> to do, not by who made the best decision on what type of ship to build
> back in the first 40 turns.

Clarke is also on turn 40:

N                   D     W     S     C         P         I   #
StarVampires     1.00  3.50  2.75  1.00   6288.83   5814.34  11
Lsars_Children   1.49  2.89  2.35  1.00   4281.59   3220.55   7
WayansFamily     1.92  1.37  4.67  1.00   4478.55   4356.52  13
Olympians        1.04  1.00  1.98  1.19    990.44    990.44   1
Agoraphobics     2.45  1.71  2.29  1.00   4910.98   4244.88   9
Klingons         3.00  2.39  3.07  1.00  16934.60   9175.39  34
Clan_Ghost_Bear  1.38  1.00  1.00  1.00   3119.53   3096.45   4
Doomfarers       3.13  2.16  2.15  1.00  10990.09   8225.90  30
Draconians       2.08  2.40  2.25  1.00   9536.00   7975.63  31
Garden_People    2.55  6.48  4.36  1.00  20437.14  13529.88  49
Vorkall_Empire   3.50  2.71  2.23  1.00  10965.29   9073.14  29
Glorkians        1.00  1.00  1.00  1.00   2797.56   2359.64   5
Paezhuta         1.19  1.41  2.69  1.00   3660.43   2487.71  13
Vogons           1.75  1.00  1.00  1.00      0.00      0.00   0
Yuggoth_Fungi    2.07  1.47  1.47  1.00   2270.13   1904.24   7
Emerald_Knights  2.18  2.31  2.10  1.16  12831.96   4172.76  28
Clan_Wolf        1.00  1.00  1.00  1.00   1300.08   1009.30   2
Imperium         3.07  2.45  2.74  1.00   6747.15   5682.30  10
Horta            2.66  1.00  2.25  1.00  11658.87   4614.51  22
Macho_Flies      2.18  1.57  2.15  1.00   5427.44   4731.17  19

Draw your own conclusions about what difference the decisions in the
first 40 turns might have made.  Again, of course Galaxy is a game of
diplomacy.  But that doesn't mean that the resources available to you
make no difference at all.  If you really think that's the case, I
invite you to play an entire game with just your home planet.

> Which is why I'd like to reduce the amount of calculations that the
> game encourages you to do in the first 40 moves.  I don't mind spending
> 2 hours thinking about strategy, or 2 hours figuring out the best tactics,
> but I don't like being forced to do 2 hours of math problems and logistics
> to implement it.  At first, the math problems and logistics planing were
> interesting, but now I find them a pain.

There's a simple solution: just play Galaxy version 3.X.  There are a
lot of games going, and presumably more in the future.  This version
*drastically* reduces the importance of logistics and planning.  The
reason is that growth, instead of being limited by capital and shipping,
is limited simply by the availability of colonists.  The combination of
cheaper capital, more efficient cargo ships, and reduced population
growth, mean that all of the reasonably competent players (who don't get
embroiled in a war are something) are able to grow their industry just
as fast as their population expands.  For example, look at the status of
game Alpha after 16 turns (sorted by decreasing industry):

N                        D     W     S     C        P        I   #  R
Forest_People         1.71  1.00  1.00  1.00  2815.99  2691.99  16
Mourning              1.20  1.00  1.67  1.00  2695.47  2578.27   8  War
Sky_Net               1.54  1.00  1.00  1.00  2638.34  2467.19   7  War
WaveBand              1.40  1.00  1.00  1.00  2922.39  2398.39  10  War
Shinma                1.40  1.00  1.00  1.00  2611.83  2321.58  12  War
Vorkall_Empire        2.00  1.00  1.00  1.00  2347.87  2273.27  11  Peace
Acoma                 1.60  1.00  1.00  1.00  2345.16  2041.76   4  War
Lost&forgotten        1.40  1.00  1.00  1.00  2813.93  1967.39  11  Peace
Lovers                1.40  1.00  1.35  1.01  2378.19  1966.45   5  War
Paulaner              1.40  1.00  1.00  1.00  2889.29  1956.33  21  War
Kanuug                1.08  1.20  1.11  1.00  2510.64  1923.42   6  War
Fungi_From_Yuggoth    1.00  1.00  2.05  1.00  2166.26  1889.31   6  War
Vladolators           1.50  1.00  1.11  1.00  2539.77  1877.29   6  War
Interstellar_Aquatus  1.22  1.00  1.00  1.00  2511.15  1872.05   7  War
Starkaad              1.27  1.00  1.32  1.00  2554.46  1859.40   8  War
LogeSpawn             2.02  1.05  1.00  1.00  2848.91  1791.02   9  War
Alphatia              1.00  1.00  1.00  1.00  2172.42  1769.61   5  War
Newbies               2.28  1.00  1.20  1.00  2487.50  1728.73   6  War
Awful_Green_Things    1.40  1.00  1.23  1.00  2459.02  1708.99  10  War
Ooze                  1.00  1.00  1.00  1.00  2703.72  1688.99  16  Peace
Malorian_Empire       1.00  1.00  1.00  1.00  2363.90  1652.36  10  War
Italianos             1.00  1.00  1.00  1.00  2256.00  1624.87   3  War
Apapane               1.20  1.19  1.19  1.00  2560.91  1619.40  13  War
The_Other_Empire      1.00  1.00  1.00  1.00  1903.84  1596.37   4  Peace
Zethrin               1.40  1.00  1.00  1.00  2226.38  1587.94  15  War
Dorsai                1.00  1.00  1.00  1.00  2000.43  1578.30   3  War
The_Empire            1.00  1.00  1.00  1.00  2527.37  1474.29  11  War
Terminators           1.40  1.00  1.00  1.00  1917.65  1453.76   4  War
Termites              1.00  1.00  1.00  1.00  2291.48  1428.10   6  War
Snafu                 1.37  1.00  1.48  1.00  1894.62  1426.21   5  War
Bitches               1.42  1.00  1.00  1.00  2096.01  1383.68   6  War
Vre'Lore              1.40  1.00  1.00  1.00  2132.43  1359.48   9  War
Norns                 1.03  1.00  1.00  1.00  1863.00  1287.53   6  War
Handwerker            1.53  1.00  1.00  1.00  2752.40  1236.15  12  War
Pink                  1.17  1.00  1.00  1.00  1212.61  1212.61   2  War
Asgardians            1.40  1.20  1.00  1.00  1736.68  1171.08   6  War
Melderyn              1.00  1.00  1.00  1.00  1736.30  1091.58   6  War
Puce                  1.00  1.00  1.00  1.00  1300.07  1071.36   4  War
Stones                1.00  1.00  1.00  1.00  1367.15  1066.36   6  War
Cartesians            1.00  1.00  1.00  1.00  1172.80  1007.22   2  War
Smellies              1.00  1.00  1.40  1.00  1167.82  1004.64   2  War
Froggies              1.00  1.00  1.00  1.00  1224.92  1000.00   3  War
Hicks                 1.00  1.00  1.00  1.00  1480.00  1000.00   2  Peace
Cretians              1.00  1.00  1.00  1.00   328.29   111.45   7  War
Fremen                1.00  1.00  1.00  1.00     0.00     0.00   0  War
Mycon                 1.00  1.00  1.00  1.00     0.00     0.00   0  War
Scum                  1.00  1.00  1.00  1.00     0.00     0.00   0  War

All of the largest players are just about the same size, and have almost
as much industry as they do population.  And since there's nothing you
can do to increase that rate beyond 8% per turn, it makes much less
difference how clever or thoughtful or careful you are.  At the same
point in game Clarke, there was a *much* larger variation.

David desJardins

References

Galaxy Results Format
markhof@ls12r.informatik.uni-dortmund.de (Ingolf Markhof)
Tue, 22 Dec 1992 14:17:20 +0000
Re: Galaxy Results Format
desj@ccr-p.ida.org (David desJardins)
Thu, 24 Dec 1992 16:10:04 +0000
Re: Galaxy Results Format
curt@oasys.dt.navy.mil (Curt Welch)
Thu, 24 Dec 1992 19:04:09 +0000