I’ve been comparing the subsector generation of *Mongoose Traveller* (MGT) and Classic Traveller (CT) in the incarnation of The Traveller Book.
Here’s what I noticed:
The distribution of **starports** is different. Note how Classic Traveller has only a bit less D and a lot more A starports.
+-----+-----+----+ | 2d6 | MGT | CT | +-----+-----+----+ | 2 | X | A | | 3 | E | A | | 4 | E | A | | 5 | D | B | | 6 | D | B | | 7 | C | C | | 8 | C | C | | 9 | B | D | | 10 | B | E | | 11 | A | E | | 12 | A | X | +-----+-----+----+
But reddit has the answer:
E Religious Autocracy. Government by a single religious leader having absolute power over the citizenry
F Totalitarian Oligarchy. Government by an all-powerful minority which maintains absolute control through widespread coercion and oppression
– MegaTraveller Referee’s Manual, apparently
The pirate bases, the travel codes, and the routes sure make the maps generated for Mongoose Traveller more interesting to look at.
Here is an example subsector generated. Notice the limited bases created.
Traveller Subsector UWP List Generator (1619343301)
Tizita 0102 C41058A-11 NI A
Vibannare 0105 D7A38A6-3 A
Nafaxe 0204 E447301-4 NI A
Zititu 0207 B648663-7 S Ag NI
Mavifainti 0208 X3016AB-3 Na NI Va IC A
Pavima 0210 D200103-6 NI Va A
Vinxena 0301 D88A644-8 S NI Ri Wa
Zoziq 0302 B454842-8
Tituteluzi 0305 A310420-12 NI A
Rabama 0306 B545412-8 N NI
Totamavi 0310 B678563-7 N Ag NI
Zon 0401 E88A522-7 NI Wa
Qnq 0410 C87997A-11 In A
Mavota 0504 A99A977-12 In Wa A
Razutife 0505 B754867-8 S
Zomanzi 0507 D67A355-10 S NI Wa
Vizoetuvo 0508 B120510-12 NI Po De A
Lurinan 0509 C4555A8-9 Ag NI A
Zotato 0510 A66A632-9 NI Wa
Lunainrina 0601 B747564-6 N Ag NI
Rizo 0604 B77A7BF-8 S Wa A
Lutomata 0605 C9769CF-6 In A
Patoin 0607 A436784-13
Ineratu 0608 A978753-11 S Ag
Zozota 0610 C220202-10 NI Po De A
Femavita 0701 A302666-10 Na NI Va IC
Titutemana 0703 BA9A796-12 NS Wa
Titezofa 0705 C99A543-6 S NI Wa
Favorivi 0709 C130383-11 NI Po De
Qtelu 0710 AAA9000-0 NI A
Nnafema 0802 C5376B9-5 NI A
Tuzinare 0803 A564304-12 NI A
Zupaezi 0804 B210964-12 N Na In
Zoreluzu 0805 C78A589-6 NI Wa A
Tofevomaq 0806 A200756-10 N Na Va
Texebain 0807 C4428AE-7 Po A
Favifabafa 0808 B99A88C-11 N Wa A
Reinxenazi 0809 B401588-8 NI Va IC
Zufate 0810 B9C8644-8 NI A
||||||| |
Ag Agricultural ||||||| | In Industrial
As Asteroid ||||||| +- Tech Lo Low Population
Ba Barren ||||||+- Law Lt Low Technology
De Desert |||||+- Government Na Non-Agricultural
Fl Fluid Oceans ||||+- Population NI Non-Industrial
Ga Garden |||+- Hydro Po Poor
Hi High Population ||+- Atmosphere Ri Rich
Ht High Technology |+- Size Wa Water World
IC Ice-Capped +- Starport Va Vacuum
Bases: Naval – Scout �� Research – TAS – Consulate – Pirate
I’ve left the Mongoose Traveller suggestions for routes in place and this is what we get:
On G+, I got linked to TRAVELLER: Out of the Box – Interlude: Two Points Where I Prefer the 1977 Edition Over the 1981 Edition talking about the old 1977 trade routes. When I tried it, not bad! (Also note this forum thread, Default CT 1977 and CT 1981, comparing the 1977 and 1981 editions; log in to see more than just the first message of the thread.)
(Classic Traveller map, with minimal spanning tree)
I decided to drop the optimising I had been doing, to get a feel for the “raw” data. This also required some rewriting and this always messes with the random number generator. That’s why this map cannot be easily compared to the one above.
I did some rewriting and added the [minimal spanning tree algorithm](http://en.wikipedia.org/wiki/Kruskal's_algorithm) again. It looks less busy, the clusters are still clusters. But there are some drawbacks. Take a look at the top right corner: 0802 and 0703 are connected (distance 1, shorter connections go first), 0701 and 0703 are connected (distance 2), and thus 0701 and 0802 cannot be connected because that would introduce a loop.
http://en.wikipedia.org/wiki/Kruskal's_algorithm
(Classic Traveller map, with minimal spanning tree)
I think for the time being I’ll remove the minimal spanning tree optimisation again.
#RPG #Maps #Traveller
(Please contact me if you want to remove your comment.)
⁂
There should be a space lane between 0601 and 0701 by the 1977 rules (A-B J-1 is 1+ to have a space lane). Also 0504-0505 and 0508-0607, hmm and a bunch more...
– Frank Filz 2017-07-13 07:31 UTC
---
Indeed! A bug!
– Alex Schroeder 2017-07-13 09:25 UTC
---
New code produces:
– Alex Schroeder 2017-07-13 09:35 UTC
---
Highlighting jump-1 routes:
Not sure about the colour scheme. And the legend at the bottom needs fixing.
– Alex Schroeder 2017-07-13 10:59 UTC
---
Nice, I like the red to highlight the J-1. When I did my subsectors, I ended up not rolling for the J-3 and J-4 routes. I also didn’t bother with a J-2 route that was already connected by J-1 (for example 0604-0804 above, there’s already an implied J-2 route as 0604-0705-0804). Similarly you could remove the 0305-0604 J-3 route (unless there isn’t a J-2 between 0305-0505 and that blue line is actually JUST a J-4 route 0305-0705...).
– Frank Filz 2017-07-13 16:28 UTC
---
Hm... Isn’t that a bit like the minimum spanning tree optimisation? As Ian Borchardt said on G+:
One problem with your optimised paths is that you are assuming that space == time, which isn’t the case with /Traveller/ jumps, because any jump takes two weeks, regardless of distance. [IIRC]
Which means it’s not cheaper (time wise) for a starship to call in at an intermediate port. So given a choice between A-B-A and A-A most people will still choose A-A by preference because it gets there in half the time. It has a greater efficiency.
I responded with the following:
As for trade route optimisation: one might say that the trade routes indicate the most popular trade routes and thus two jump-2 segments will be more popular than a single jump-4 segment because there will be more ships that can handle it, even if that makes the route take twice the time. Well, that was my argument when I realised that too many routes just made for an ugly map.
So basically the jump-2 route 0604-0804 says that ships can and will make this jump directly – unlike the jump 0604-0703, which is also a jump-2 but somehow the dice have decided that it’s not profitable enough to do. It’s up to the referee to explain the difference. At least, that’s Ian’s argument.
For myself, I’m still undecided. All I can say is that it looks busy and messy and I don’t like it too much.
As for distance, I guess from a player perspective, at the beginning of a typical campaign all we care about are jump-1 and jump-2 routes. Perhaps I should just use a third colour for jump-2 routes.
– Alex 2017-07-13 16:52 UTC
---
Support for sectors, click to zoom. Note how I used a third color for jump-2 routes.
– Alex 2017-07-13 18:03 UTC
---
D’oh. I just noticed that my generator never generated any gas giants. Fixed!
This also means that all the random seeds will now generate different maps. Nooooo!
– Alex 2017-07-13 18:33 UTC
---
As for picking a subsector capital, Ian Borchardt says:
Definitely an A class spaceport (and it will have an attached Imperial naval and scout base). It probably should be a Rich world (Atmosphere 6-8, Population 6-8), and I’d assign it a Government Type of 8. The TL should be amongst the highest in the subsector (and a minimum of 12).
And regarding trade routes, he says:
This is the approach I would take.
First generate the main routes.
(1) Connect all Starport A within Jump 2 of each other.
(2a) Connect all Starport B within Jump 2 of a Starport A.
(2b) Connect all Starport B within Jump 1 or each other.
(2c) Connect an unconnected Starport B to another Starport B within Jump 2.
If a connection in step 2 connects Starport A to the sector capital then upgrade it to a main route.
Next comes the feeder routes.
(3a) Connect Starport C to a single Starport A within 1 jump.
(3b) If not connected, connect Starport C to a single Starport B within Jump 1.
(3c) If not connected, connect Starport C to a single Starport A within Jump 2.
(3d) If not connected, connect Starport C to a single Starport A within Jump 2.
(4a) Connect Starport D to a single Starport A within 1 Jump. (optional)
(5) Class E starports are never connected by a trade route.
If in need of direction, assume the shortest path to the subsector capital.
This defines the regular services (for various definitions of regular). Everything else relies on tramp freighters.
[In your example map the small disconnected group of systems to coreward might have better trade connections to coreward than it does with the subsector it is technically a part of. Or it may be a different polity. And yes, I cheated with the rosette of class A and C starports to rimward, because of a fondness for benzene rings...]
– Alex Schroeder 2017-07-13 22:31 UTC
---
One thing I did to simplify my route maps is that if you have three worlds, One, Two, and Three, and there is a J-1 route from One to Two, and a J-1 route from Two to Three, I also assume the J-2 route from One to Three is also regularly travelled. I also in a few places, having rolled just the J-2 route, assumed that the J-1 routes were thus also regularly travelled (this makes a world that might otherwise have had less access have more, I’m not that bothered by that, and would rather have a simpler map).
– Frank Filz 2017-07-31 16:02 UTC
---
I’ve been working on a sector and my rules of what to draw were like so: Two planets side by side - J-1 implicit. No need to represent. Two planets with a space in between, draw J-2 link between the two (green). Two planets with 2 spaces in between, draw J-3 link between the two (green). Two planets with 3 spaces in between, draw J-4 link between the two (green). If two worlds have a path between that would go through the center of an intermediary planet, then don’t draw that link.
I like your sector tool, but one advantage in ’Traveller Map’ is that you work with a file of system data and a file of metadata like routes, regions, borders, etc. So I can generate the final map with all my jump routes or I can show only one or other level of jumps (J-1s only for instance) by a simple commenting in the XML metadata. It’s not quite as nifty as having a layered file (like something from GIMP) with layers you can just switch on or off or adjust the transparency of and to control Z order, but the hard part doing this in gimp is getting consistent manual route marks (angles off, distance off). Traveller Map does that part for me (if I define the routes in the XML).
As I’m using a variant set of Jump rules lately: Small ships are easiest to send into Jump space. Bigger ships run bigger risks and damage as well as burning more fuel. Similarly, if I have a better drive (J-3 vs. J-1), if they both Jump 1, the Jump 1 ship does it slower but with more fuel efficiency. Jump timing is variable by distance and by how advanced your J-drive is. Ships need to discharge some radiations from Jump for a time which is a time consumer in multiple jumps. Typically, 4 J-1s will be notably slower than 1 J-4. But the J-4 may burn twice as much fuel as the four single jumps. That’s one of the tradeoffs.
I like the notion of having software determine likely trade routes (vs. just jump routes), but I’d want to be able to supply my own math model for that.
High tech planets seem to be powerhouses, but if they have a very small population, maybe not. If you have a lot of pop, but your TL is 9.... my point here is just that there’s a bunch of factors you’d need to include in any route finding:
– TomB 2022-09-01 06:08 UTC
---
I have a different generator in Text Mapper which generates Traveller map data in text format, which you can edit before rendering it. It has three kinds of routes, these days: “rich” is golden, “trade” is cyan, and “communication” is tomato red. For a quick impression, render a random map a few times. The manual fiddling would allow me to invest the same amount of thought into the map as what you’re describing. In my case, I doubt that it would make the game significantly more interesting, so I’m not sure I actually want to do that. Perhaps it’s good enough as it is, right in the middle between too many confusing routes and enough routes to make for some inspiration.
If I wanted to spend a bit more time on a subsector, that’s probably what I’d use, combined with a more elaborate generator like Hex Describe. It generates a lot of text that fits the map. For a quick impression, reload the random map a few times, if you’re interested. It would need a lot more detail, I think. The problem in writing the random tables is that so man of the attributes are co-dependent. Oh well. One thing on the todo list if I return to Traveller one day. 😀
– Alex 2022-09-01 12:55 UTC
