Not sure if I have everything correct here yet, but it's late and I need to stop writing.
Here's the spreadsheet I used to scrawl crazy copypasta etc while planning the race:
A. Initial Plotting
First I took a random sampling from the list, made sure it was an acceptable combo, and tested it for different combinations of Rod and Pole and engineering. From what I could see, the regulation bonus wasn't going to be high enough to counteract the bonus from high level engineering, but I'd have to verify that later. Most of my calculations would be largely independent of jump range, and rely on reducing overall hyperjump distance, so I could reevaluate that later.
B. Large ship viability?
I copied station chart/diagram from Furrycat's post, and deleted all the stations that a large ship couldn't land at. This left only one viable path for large ships (both the Orca and Beluga) was Wailladyan, Avaticus, Liu Baja, HR 5716, which included two planetary bases. I wanted to avoid those at almost any cost, since I hate getting eclipsed when taking off, so large ships seemed to be out. I already determined I would be sticked with the Saud kruger line, since a 6 minute bonus was huge, and the marine mammal ships are very capable racers in general. So Dolphin it is!
C. Route Set
Furrycat pushed a change to the EDTS program a bit ago that allowed one to select one or more stations from a list that had to be visited. However it didn't have the ability to select two parameters that needed to be fulfilled, so it couldn't do this rae all on its own. I ended up grouping the stations into four groups, either by fish or by last word, and having EDTS optimize those routes. If it found an illegal route (say I was sorting by fish and it found a route with two terminals), I would branch that into two different strings, and if those were illegal, so on and so on.
D. The Last Word
It seemed to me that elimination by last word worked better, as it created several smaller groups with less overlap, than by fish, although in the end they ought to converge on roughly the same route.
E. Traps!
Then when figuring out my route, I noticed that a certain fuzzy felie had been very tricksy, and a station that seemed like a good idea to travel to would take 25 minutes of supercruise. >.< So I added supercruise distance and time to the top of the sheet, and eliminated all planetary stations, and the one with 389k ls supercruise. I left in the 14, 11, 5, and 3 K supercruise ones, and figured that if they were on the optimal route, I could calculate that cost against the alternatives later.
F.
Now that I had a plausible route, I started looking at different loadouts (regulation vs cargo filled vs empty) to get some times for them. From this is seemed clear that engineered and filled with cargo would be the bonus winner.
G. Third course set
Then I tried the station routeset optimization again, with my cargo filled engineered jump range, and the bad stations struck off the list. With only two viable stations for Sturgeon and Bass, is seemed to make more sense to sort by fish name again instead of last word.
H. Bonus optimizations
Now that I had a route, I tested it with adding in Rod, Pole, Okinura, and skilled Hamlet because only Crotalus is THAT crazy. I took EDTS estimated time, subtracted the bonus time, and Found that either Rod or Rod/Fisher would be best, with Rod Fisher coming out on top so far. But I didn't know if adding in Rod and Fisher would change the other stations I needed to take, and that would take... AWHILE.
I. Abusing my poor CPU
This was most of my Sunday, since optimizing a problem with 4 routesets and 2-3 other waypoints took my computer over an hour at times, so I'd come back and plug in a new string every so often. This was allowing the optimizer to choose the best route from any non planetary station within reasonable distance, plus every permutation of bonus stations. Then redoing it and branching if it found a route with duplicates. This again seemed to indicate that a Rod/Fisher combo was the best bet, even though I really wanted a Rod/Pole/Fisher to work.
J. Changing jump distance.
Now it was time to find a fish market. One of the best things about the Rod Fisher comb is that I could have a jump range of 53 ly until I hit the first waypoint station, so by hitting Rod and Fisher first, I was able to minimize jumps taken.
K. FISH!
I now just needed to find a station to stop at for fish. I wanted it to be within one jump of Liu Baja/Ray Dock , and hope I could find one that was 2 jumps away from Rod, but try as I might, the space between them was empty, so I had to settle for LP 339-7/So-yeon Dock at 3 jumps, but it was only a few ls from the primary, for some consolation.
L. Armor
I chose not to go for the fsh in a barrel achievement, in spite of not having shield, because that did allow my to put heavy duty armor engineering on my hull, and repair at every station. I personally felt the bonus wasn't quite enough incentive to make up for all the slower more careful docks and increased danger (especially while trying to drop fish on another ship), however, it was a close choice.
M. Cargo usage and the Lighthouse
I was planning on using every available slot for cargo, since all refueling could be done at stations. But on my first run the in-game plotter too me to Jackson's lighthouse, which was beautiful and wonderful! Except for the fact that I didn't have a fuel scoop equipped.Even though Jackson's Lighthouse only saved me one jump on the way to Rod (as opposed to maybe 3 on the way to Pole) one jump is 45 seconds saved, maybe 30 when taking into account the boosting time, and the bonus for 4 tons of fish was 16 seconds, so taking the scoop made sense.
N. Alternate Routes proposed by Racers
Given that Furrycat designed the race, I figured I'd see how his route would compare.
My route in EDTS with system bonuses was estimated 27:17 flying time - 7:00 bonuses = 20:17 estimated time. (Actual flying time was 10 minutes longer, because landing at stations and whatnot I think.)
Furrycat's route in EDTS was estimated 25:42 flying time - 10:00 bonuses = 15:42 estimated time. However it did have two planetary landings, so there's a question of whether those would cost more or less than 4.5 minutes. I also did this assuming one could take two jumps from Kappa to Pole via Jackson's Lighthouse.
MostlyEvilKiwi's route in EDTS was estimated 23:31 flying time - 7:00 bonuses = 16:31 estimated time. With one planetary landing.
Both of those routes would likely need to add a stop to pick up fish, which would add some time as well, and my EDTS route included that.
However, they both seem highly viable, and would be fun to test sometime.
Here's the video that resulted from the above convulusion:
https://www.youtube.com/watch?v=3W5wr-EvKKM
37:00 Poisson davril
38:00 One of my favorite ads.
Songs were just from a search on Spotify for "April Fools."
I feel like the stuffed fish on my headset should've gotten a bonus though.
Also a note to MostlyEvilKiwi, I only stopped at four stations, must've been a transcription error.
Congrats to all participants, ust for finding a course, ANY course and finishing a ridiculously fiendish race.
And many thanks to furrycat for possibly the most evil plotting problem we've ever encountered, as well and very carefully balanced bonuses. I'd like to hear how the design went if you're interested in sharing.
Awesome planning and execution as ever! I don't feel nearly so bad now ... what I did (as hinted at earlier) was whack all the stations in as bookmarks, sort the ones with long SC's or planetary landings to the bottom, eyeball the galaxy map as I clicked on the others to find a bunch that satisfied the naming criteria and then eyeball the galaxy map some more to pick a sensible route between them.
So did nobody else simply dump fish outside the start station and then scoop them back up on the way out? Having to actually go somewhere to buy fish on the way seemed far too obvious!
P.S. anyone else going to Lavecon this year?
