Too many jump points?

[Rich.h]

[IanD]

I like the idea of fewer jump points. In my current game, real stars, 177 systems surveyed, 18.6% have one jump point, 25.4% have two jump points, 22.6% have three jump points, 15.2% have four jump points, and 18% have greater than four jump points, (the maximum being 18!).

My only other comment is that if you start with multiple NPR empires on Earth then while you can usually beat them to the first ring out from Sol, the second and third are very difficult to get to first with the current fuel usage rules. But under the existing jump point generation rules they usually leave sufficient unexplored chains for you to go down. Once you reach fuel consumption 0.4 litre per engine power hour you are back to parity. The NPRs should have some restriction on fuel/range with a multiple Earth start.

[Vandermeer]

The new distribution is great, and exactly what I had wished it would be like. The newer distribution seems better too, as I feel 3 JP chances should still slightly outweigh 2 JP ones. Then I especially like the interesting influence that mass has in the whole deal. Seems black holes have some chance of becoming big galactic terminals now each time?

[hyramgraff]

This is a second test with a slight change to the generation chances. It moves a little % to 3 JP from 4+ JP. Here is old vs new proportions:

OLD
1 JP: 10%
2 JPs: 36%
3 JPs: 32.4%
4 JPs: 15.12%
5 JPs: 4.54%
6 JPs: 1.36%
7 or more: 0.58%

NEW
1 JP: 10%
2 JPs: 36%
3 JPs: 37.8%
4 JPs: 11.34%
5 JPs: 3.40%
6 JPs: 1.02%
7 or more: 0.44%

So now we have about 1 system in 9 (for stars with 0.99 mass or smaller) with 4 JPs and 1 in 20 with 5 or more JPs. Almost three quarters will be 2 or 3 JPs. Junction systems will actually work out a little higher than that because of the higher mass stars making a significant difference at 4+. This feels better than the first version but again it is quite a small sample size. Here is a map generated under those rules:

What was the maximum number of systems in your test galaxy?  I ask because I tried to setup a similar scenario via SM mode exploration in a galaxy with 250 systems and I found that after around ~120 systems were explored, jump points would frequently link to an already explored system.  These "dormant" links will increase the number of jump points in a system and I don't think that they are accounted for in your probability table.

[Ostia]

This change is for initial JPs, not including dormant ones. The main change here is a curbing of massive JP amounts per star.

And that you get a lot of link back to explored systems is a given with a maximum of 250 systems. That's simply how the assignment algorithm works. (Which is basically just a random roll for the system ID, with a chance that the target ID is close to the ID of the orginal system.) So you get a lot of links to known systems because you already know half of the total galaxy.

[NihilRex]

Steve, did you ever see this one?



https://www.reddit.com/r/aurora/comments/3byjmq/87_systems_and_i_havent_reached_4_jumps_from_sol/

[Steve Walmsley (OP)]

What was the maximum number of systems in your test galaxy?  I ask because I tried to setup a similar scenario via SM mode exploration in a galaxy with 250 systems and I found that after around ~120 systems were explored, jump points would frequently link to an already explored system.  These "dormant" links will increase the number of jump points in a system and I don't think that they are accounted for in your probability table.

It was real stars so there was no universe size.

I agree about dormant JPs changing the distribution. In real stars, there is a 5% chance that a JP will link to an existing system. However, if the system being connected to has unlinked JPs, one of those will be used rather than creating a new one, so the actual variation due to dormant JPs will vary from game to game.

[Erik L]

Steve, did you ever see this one?



https://www.reddit.com/r/aurora/comments/3byjmq/87_systems_and_i_havent_reached_4_jumps_from_sol/

Technically, a couple of those systems are 4 jumps from Sol.

Technically correct. The best kind of correct.

[NihilRex]

The full title is "4 jumps from Sol in all branches."

And I think there was one 5 jump-out explored somewhere in there too.

I ended up saving it and restarting, because managing the map was becoming a bit difficult...

[linkxsc]

I had 1 game where i set sol to have 4 (good average number) 1j in 3 directions netted me systems with 1,5,6,6 and from 1 of them in the 5, i found an 8

So making it a little more tame might be nice, especially on bloat. 3-4 per system is a lot less by the 3rd jump than 5-6 per.

Though I am curious, i dunno code or even psuedo code. But is that a limit on the max? Could you possibly jump into a system with 15 jps?

[Steve Walmsley (OP)]

I had 1 game where i set sol to have 4 (good average number) 1j in 3 directions netted me systems with 1,5,6,6 and from 1 of them in the 5, i found an 8

So making it a little more tame might be nice, especially on bloat. 3-4 per system is a lot less by the 3rd jump than 5-6 per.

Though I am curious, i dunno code or even psuedo code. But is that a limit on the max? Could you possibly jump into a system with 15 jps?

Yes - it is very unlikely though. Rather than having set % for the number of JPs the code adds them one at a time, with a chance for each subsequent JPs. So (under new rules) 100% for first, 90% for second, 60% for 3rd and then 30% each for any additional JP. So the chance of getting 2+ JPs is 90%, the chance of 3+ is 54% (90% x 60%), the chance of 4+ is 16.2% (90% x 60% x 30%), etc.. The chance of getting a specific number of jump points, such as 3, is equal to the chance of getting that far minus the chance of getting further. For 3 that would be 54% - 16.2% = 37.8%.

[boggo2300]

Technically correct. The best kind of correct.

The ONLY kind of correct.

FTFY!!

 


Actually Steve what about (carrying the 2300 theme a little further) setting a maximum distance a JP can connect to, so you could implement the 7.7 ly range that 2300 uses

[Steve Walmsley (OP)]

The ONLY kind of correct.

FTFY!!

 


Actually Steve what about (carrying the 2300 theme a little further) setting a maximum distance a JP can connect to, so you could implement the 7.7 ly range that 2300 uses

That is actually possible with real stars because Aurora has 3D coordinates for each system in real space (converting right ascension and declination into 3D coordinates was fun ). Currently the JP link code creates a list of other systems in ascending order of distance then checks each one in turn with a 20% chance to select that system. I guess I could limit the list to those within 7.7 LY and then loop back to the start

[NihilRex]

That is actually possible with real stars because Aurora has 3D coordinates for each system in real space (converting right ascension and declination into 3D coordinates was fun ). Currently the JP link code creates a list of other systems in ascending order of distance then checks each one in turn with a 20% chance to select that system. I guess I could limit the list to those within 7.7 LY and then loop back to the start

Any chance that # could be configured at creation?

[Haji]

This is looking very good to me. There is just one small thing I'd like to add: with this and the significant increase in useless systems (due to your inclusion of massive amounts of brown dwarfs few patches back) you may want to rethink civilian shipping distances. Currently, if my memory serves, a civilian vessel will look for a destination only four jumps away, which is a little short at the moment, at least in my opinion. But that's just a small issue.