The delicious irony of Fastest Route plotting in the gal map

[Skurrock]

I am sitting here with a set of interesting systems close to the core and it struck me, why does plotting a Fastest Route take longer than actually flying there.
The algorithm is too precise, I want a fast route not the fastest one.

And yes, I know how to get fast plots near the core but if I actually want to fly outside the box the Fastest Route algorithm is useless.

Let me plot a fast route, not necessarily the Fastest one and remove the EST (estimated smoking time) out of plotting.

Brought to you courtesy of the Candy Crew DWE Think Tank.

 

[Shanaeri]

Going to and on my way back from Beagle point I noticed the route planner has a built in "call the fuel rats" mode. Sometimes it would plot me a route that carefully avoided as many scoopable stars as possible while winding the plot around them.

Luckily, I always check the route for refuel stops and adjust as needed, but I think people do need to be aware of this new "Call the fuel rats" routing mode

see what I mean?

 

[Hanekura Shizuka]

I am sitting here with a set of interesting systems close to the core and it struck me, why does plotting a Fastest Route take longer than actually flying there.
The algorithm is too precise, I want a fast route not the fastest one.

And yes, I know how to get fast plots near the core but if I actually want to fly outside the box the Fastest Route algorithm is useless.

Let me plot a fast route, not necessarily the Fastest one and remove the EST (estimated smoking time) out of plotting.

Brought to you courtesy of the Candy Crew DWE Think Tank.

Route plotting in the core can be remarkably fast - if you know how to do it right. I would recommend this as required reading for anyone wishing to travel through the core regions.

Beyond that... if you find the current method not to your liking, I would recommend a bit of wisdom from a 20th century polymath:

"So the universe is not quite as you thought it was.
You'd better rearrange your beliefs, then.
Because you certainly can't rearrange the universe."
-- Isaac Asimov, "Nightfall"

 

[Engadine]

I wouldn't be surprised that once out in the deep black, that no 2 routes would ever be the same. That is to say plotting the same route twice would result in slightly different routes each time.

 

[superflyguy]

what Hanekura said. using the 'magic number' makes plotting super fast.

 

[Shanaeri]

I wouldn't be surprised that once out in the deep black, that no 2 routes would ever be the same. That is to say plotting the same route twice would result in slightly different routes each time.

In my experience you are correct. To help me navigate my way back I took a note of each stop to plot a new route I did on the way there. On the way back I'm often being routed to a totally new system even though both the start and end points are the same and the amount of fuel I had was the same also.

 

[Cdr Voorheez]

Magic number. Mines 919.

 

[Fox White]

Going to and on my way back from Beagle point I noticed the route planner has a built in "call the fuel rats" mode. Sometimes it would plot me a route that carefully avoided as many scoopable stars as possible while winding the plot around them.

Luckily, I always check the route for refuel stops and adjust as needed, but I think people do need to be aware of this new "Call the fuel rats" routing mode

see what I mean?

http://i.imgur.com/N2R49TU.jpg

The last 1k LY into the bubble was really difficult because of this mode.. so many scoopable stars close by but oh no those arent fast enough

 

[Hanekura Shizuka]

The last 1k LY into the bubble was really difficult because of this mode.. so many scoopable stars close by but oh no those arent fast enough

There's a reason why the word "Aucoks" has become the curse it has. Most Buckyrunners headed to A* either plan the route through Aucoks in advance, or just avoid that area. Advice in this regard: Go about 125 ly above or below for the area about 1000 ly north of Sol; you'll find far more scoopables.

 

[Skurrock]

Route plotting in the core can be remarkably fast - if you know how to do it right. I would recommend this as required reading for anyone wishing to travel through the core regions.

Beyond that... if you find the current method not to your liking, I would recommend a bit of wisdom from a 20th century polymath:

"So the universe is not quite as you thought it was.
You'd better rearrange your beliefs, then.
Because you certainly can't rearrange the universe."
-- Isaac Asimov, "Nightfall"

As I mentioned in the OP I know how to get fast plots, indeed I use the DW Discord bot to give me the magic number. It is still a pain to have to search for a system within the range recommended to search in and is no guarantee that you will get a fast plot at all.
My suggestion was made so that we are not forced to use magic numbers to get a fast plot at all. The Fastest route algorithm is almost certainly too precise and I fail to see any benefit from it in practical terms.

 

[ASC]

Beyond that... if you find the current method not to your liking,

...

"Because you certainly can't rearrange the universe."
-- Isaac Asimov, "Nightfall"

Ah, but we can!

If everyone on DW submits a bug report, or posts a video to Youtube showing how unusable the route plotter is, then FD might just have to do something about it.

 

[SvennoJ]

The planner isn't even that precise, it only does that insane fuel balancing phase for the last few jumps. Plotting 100 or 1000 doesn't make a difference as it's the last few jumps that cause the break down.

The magic number doesn't get you to the system you want to go to. Another work around is to plot a route to about a jump's distance beyond your destination and then while it's stuck on 99%, select your actual destination and chances are that you get a route instantly. (If that system is in memory, part of the route tree, you get the route there immediately bypassing the fuel phase) However very close to the core this might not be an option. My laptop drops below 1fps when plotting within 1000ly of the core which makes it impossible to select different systems without cancelling the route plan first.

The dumb thing is that planning from the core has no trouble at all. The result is the same, why not change the planner so it plans backwards when going to the core.

Routing seems completely broken in low density areas. The visual route tree is not showing up at all 95% of the time. FSD boost lines don't show up either. And even when it does it still says no route possible most of the time for a 2 jump route. At least you still get routes in the core, eventually. A bidirectional search would be best. That way any end point in a low density area can always keep expanding without being unreachable by aggressive culling. It's the most efficient for the routing problem ED is dealing with.

 

[Hanekura Shizuka]

I'll be simple and I'll be blunt.

After a certain point, mathematics just looks back at you and laughs. That's what's happening here. If you want an idea as to what I'm talking about, look at this discussion here. At some point, the problems just become too big. Too many stars, and the router is going to take time no matter what algorithm it uses.

I don't think complaining is going to work. They know every in and out of that route plotter, and - given the noise that's already been made about it - have probably done everything they know how to do to make it work as well as possible. So complaining is basically complaining that the sky is blue.

Do I think asking that the algorithm be released as well as the underlying theory and purpose behind it can help? Yes. The reason for this... I think Frontier have done everything they know how to do with regard to the route plotter. However, I don't think they've done everything that everyone playing the game knows how to do.

Do I think that the route plotter can - and probably should - be tailored to different situations? Yes. That said, the selection on which router to use should probably be manual; also, to do that, we probably need to know the algorithm(s) currently being used, so that we can make educated recommendations on how it can be improved. Right now we have guesswork; we need more concrete information on the route plotting algorithm before we can address the problem. I personally would love to see a 'heading' type router - basically goes as far as it can in a particular direction, without worrying so much about a particular star.

 

[Skurrock]

I'll be simple and I'll be blunt.

After a certain point, mathematics just looks back at you and laughs. That's what's happening here. If you want an idea as to what I'm talking about, look at this discussion here. At some point, the problems just become too big. Too many stars, and the router is going to take time no matter what algorithm it uses.

I don't think complaining is going to work. They know every in and out of that route plotter, and - given the noise that's already been made about it - have probably done everything they know how to do to make it work as well as possible. So complaining is basically complaining that the sky is blue.

Do I think asking that the algorithm be released as well as the underlying theory and purpose behind it can help? Yes. The reason for this... I think Frontier have done everything they know how to do with regard to the route plotter. However, I don't think they've done everything that everyone playing the game knows how to do.

Do I think that the route plotter can - and probably should - be tailored to different situations? Yes. That said, the selection on which router to use should probably be manual; also, to do that, we probably need to know the algorithm(s) currently being used, so that we can make educated recommendations on how it can be improved. Right now we have guesswork; we need more concrete information on the route plotting algorithm before we can address the problem. I personally would love to see a 'heading' type router - basically goes as far as it can in a particular direction, without worrying so much about a particular star.

I have a degree in CompSci & AI, I know the problems involved in implementing algorithms of this nature.
You are wrong though: the current algorithm is constrained to find the "fastest" route so it spends way too much time optimising the last few jumps.
If you remove that constraint and constrain it to return the first route that works then it doesn't need to optimise anything, you prune all the alternative routes after finding the first one.

 

[ASC]

I'll be simple and I'll be blunt.

After a certain point, mathematics just looks back at you and laughs. That's what's happening here. If you want an idea as to what I'm talking about, look at this discussion here. At some point, the problems just become too big. Too many stars, and the router is going to take time no matter what algorithm it uses.

I don't think complaining is going to work. They know every in and out of that route plotter, and - given the noise that's already been made about it - have probably done everything they know how to do to make it work as well as possible. So complaining is basically complaining that the sky is blue.

You might be right, but I don't think you are. There definitely seems to be a fuel optimisation phase for the last 3 jumps. I can't remember ever getting a route with long jumps all the way but for a tiny final jump; it's always several jumps that have reduced range, which is consistent with them being optimised for fuel use.

I think the reason that nothing has been done is the relative shortage of complaints compared to other issues. Most people seem to know about the workarounds and don't make any noise. FD seem to design and test with travel around the bubble in mind, which is understandable as it is where the vast majority of players are, but it means that problems like this aren't found or considered important.