Community Event / Creation Buckyball Racing Club presents: The Kessel Run (31.10 - 15.11)

Calculating jump-by-jump, but alas there was a bug in my program (as there are with all programs that run overnight) which meant I had to kick it off again this morning (although now with new-improved optimizations that crept into my head while I slept). Currently just over a 3rd of the way through.

And errr ... I'm not planning on sharing the results of my program (yet). Now, before you all hate me let me explain. Some races dictate the route, others leave the planning as part of the challenge. This is the race planner's prerogative so, in this case where route planning is all part of the fun ... who am I to spoil that element of the race by simply giving the game away. :D
Cool.

BTW, I didn't mean that either of you should publish your findings before the race is over, though :D at the P.S...

Anyway, it will be interesting to compare notes after the event has finished. :)
 
Alec & Jigster:
Interesting to see what you calculations come up with. :)
Are you calculating the routes jump-by-jump or just straight lines from target system to target system? (that 13! reminds me of TSP and suggests the latter)

I'm calculating straight line distances and the best output so far is only a couple of percent better than my best visual guess. Worth it? Not sure.

Ship specific jump-by-jump matrices would be better but they're a good day's work each and not for me.
 
My ship can make the Kessel Run in less than 12 kilometers :D:D:D

OK, we might as well get into this now ... there's a LOT of stuff out there on the internet about Han Solo's famous "made the Kessel Run in less than 12 parsecs" quote (the apparent problem being that a parsec is a measurement of distance and not time). Noob error or is there more to this than meets the eye?

Further reading: http://www.wired.com/2013/02/kessel-run-12-parsecs/
 
Last edited:
So my original brute-force algorithm is still going (running on a 5-year-old machine), and silly me, I'm printing every iteration so I can watch the progress, slowing it down. Right now the shortest route is 1638 LYs, but I'll be surprised if it's done running before the race starts in another 4 days.

In the meantime, I ran an algorithm with several pre-selected locations already assigned in the route and generated a route of 1604 LYs. I'll be very surprised if the raw brute force method can provide a significant improvement.
 
Last edited:
So my original brute-force algorithm is still going (running on a 5-year-old machine), and silly me, I'm printing every iteration so I can watch the progress, slowing it down. Right now the shortest route is 1638 LYs, but I'll be surprised if it's done running before the race starts in another 4 days.

In the meantime, I ran an algorithm with several pre-selected locations already assigned in the route and generated a route of 1604 LYs. I'll be very surprised if the raw brute force method can provide a significant improvement.
LYs or coordinate units? Or are they the same?
 
LYs or coordinate units? Or are they the same?

They're the same, but that distance is "as the crow flies" between our destinations, so in practice it'll be a longer journey than that. For purposes of comparing possible routes, it's good enough.

On a side note, it turns that out I was wrong about not finding a significant improvement. I reduced the number of pre-selected destinations, which will take a few hours to finish...but only 15 minutes in I've shaved off almost 100 LYs.
 
They're the same, but that distance is "as the crow flies" between our destinations, so in practice it'll be a longer journey than that. For purposes of comparing possible routes, it's good enough.

On a side note, it turns that out I was wrong about not finding a significant improvement. I reduced the number of pre-selected destinations, which will take a few hours to finish...but only 15 minutes in I've shaved off almost 100 LYs.


Whipped up a quick mockup in Maya and tallied up the distances for my best guess just eyeballing it. I got ~1535.94ly as the crow flies. How close am I?
 
Whipped up a quick mockup in Maya and tallied up the distances for my best guess just eyeballing it. I got ~1535.94ly as the crow flies. How close am I?

Mine came out at 1536.6 and that's from a definitive run through of every single possible alternative. Now all I need is a starters flag and about an hour and a half's worth of cold beer and/or hot coffee (depending on time of run).
 
Last edited:
Using rounded values (integers) resulted in 1537 coordinate units here. Thought that it might be a local optimum, but as brute force method gave same result, it is the global optimum. :)

I think I'll do at least a partial check on actual in game distances on my recon run...
 
Mine came out at 1536.6 and that's from a definitive run through of every single possible alternative. Now all I need is a starters flag and about an hour and a half's worth of cold beer and/or hot coffee (depending on time of run).

Glad we got the same result. :)
 
Well it appears my shortcuts have only gotten me close...but it's close enough that I'm going to stick with it. I need my CPU cycles back to play the Overwatch beta. :cool:
 
There's a black hole near the end point on this run in the system 9 OMEGA-1 SCORPII B if anyone is interested. It's just 15.3Ls from its nearest star.

Screenshot_0017.jpg
 
Just completed a Reg. Cobra scouting run in 2:19:15. I had some network trouble on the last leg of the journey that cost me maybe 5 minutes so I'm hoping to shave my time down to at least 2:10 once the race starts in earnest. Anyone else got a time they want to share yet?
 
Back
Top Bottom