Distances and travelling time

Although why they would choose to walk in the first place is anyone's guess.

Now I think it's time we think of a scale for distances, since we've established the average human has to travel ten times the distance as giants. We could use the Chordoni - Giant Tree as the base reference.

Travel chart's a neat idea, but has the catch in that you need to attribute for "prey" travel times as well.

Furthermore, mountains, well, seem low. Mountains aren't just bad because the terrain's uneven (and thus a horizontal mile could be a vertical travel of nearly twice or thrice that whilst following the path), but because there's explicit limitations on how fast one may go in certain places along mountains, areas that one just can't cross (about equally common for Prey vs Predator, I'd imagine, as whereas a Predator can reach further and stride further, it too needs more stable support and cannot use some gap obstacles (rickety bridges, vine descending a several hundred meter sheer face, etcetera) that Prey could).

Nice place to start, but those two things (Prey-version, and overall mountain change) I feel need to be modified.

woah ! Holy shit that's a super idea you had there

I find it very very interesting ! for each locations we could have a part of the map broken down like this with travelling distance to important locations ( mainly Negav and the giant tree )

Awesome job !

Indeed, this is a great idea.

It's even be possible to make a flash (or something else) application that allows people to make a route by clicking on a hexagon divided picture of the map. The application would then automatically calculate the travelling time.

Is the 10 days travel time assuming the Prey doesn't set-up camp, or if they do? I mean, since humans need to sleep, that would make the trip longer.

Militant-Prey wrote:

Is the 10 days travel time assuming the Prey doesn't set-up camp, or if they do? I mean, since humans need to sleep, that would make the trip longer.

Depending what kind of prey we're talking about. Some of them do still have the same characteristics as humans. so yeah they get tired Nap, and possibly camp also.

No, the 10 days would account for sleeping.

Crisis's "one day" to Vivians was clearly laid out as a 10-11 hour thing, so she would have probably spent the night.

Therefore, a group of humans traveling at a tenth of her speed would also have similar days.

Militant-Prey wrote:

Is the 10 days travel time assuming the Prey doesn't set-up camp, or if they do? I mean, since humans need to sleep, that would make the trip longer.

I accounted for the time in my post... sorta. I gave six hours a day to rest, which isn't really all that much. That was assuming that they rest predominantly to give any beasts of burden their time, whereas those in the caravan rested in shifts either on the go or while the BoB rested.

An average would probably be much closer to only about 9-12 hours a day, since many caravans would probably prefer to move in daylight hours (Adventurers would probably prefer to travel in night hours, relying more on stealth and having less people needing magical buffs to function at night than a whole caravan). In such a case, the times I gave would be increased by 150-200%, meaning 7.5-8ish days as the lowest end and practically a month for the high-end. Forced march could very probably decrease the time by a good bit, an example being how far Napoleon could march his forces in a forced march and in how much time. However, forced march would also probably only be an option for adventurers or parties that could slap enchantments on their caravans up the wazoo.

A roughly 60% completed hex version of the full sized Felarya map. There are several mistakes on this one, namely the fact that I likely made Negav, Ur-Sagol, the Giant Tree, and Anko's Bridge far, far too large. I also need to change most of the colors at least a little, to make sure that all the darker ones represent more difficult terrain, etc.

I apologize ahead of time for my rambling and incoherence.



Okay, I finished part 1 of the "lets nail down some distances on the map" project, namely the part that I called "Hexification". If you click on the thumbnail you will be taken to a very large version of the map. Part 2 is figuring out what the travel times across the different hex types should be, and it is here that I could use some help from everyone else.



I have a spreadsheet that I have filled with data, including the tentative travel costs across each hex, and the total cost for a set of trips, with allowable deviation. What I have been doing is adjusting the values for the cost of each hex in order to get as many paths "passing" as possible... but I don't have nearly enough paths to "find" the value for every single hex. That's how everyone else can help me; by makings paths and coming up with a travel time for that path that they think is reasonable for either giant humanoids or regular sized humanoids, or both. It would be quite helpful if you were also to include an image of that path, drawn on top of part* of the hex map, but simply telling me how many hex's of what types it crossed would be helpful too.

The spreadsheet:



The spreadsheet is a little complicated, but the gist of it is this: the invariants are simply little sanity checking rules I wrote to ensure that I don't end up with a situation where it takes less time to travel through heavy jungle than it does through regular jungle. All of the "cost" columns are in minutes, and as you can see some of them have rather strange values in them. Anyone who has learned Linear Algebra probably understands what I am trying to do with my collection of paths; they form a system of equations that, in an ideal world, I could simply apply Gaussian elimination to and get the values I want directly. Since this is not an ideal world however, I have to be a little more indirect.


* - I only need the part of the map with the path on it, no need to post the 1.5 meg map over and over again; just crop it down to size in a image editor.

A few things:

1) How did you get 630 minutes for the Waterfall path for a Predator? Your values for the path would take only approximately 290 minutes to get that far.

2) Many Giant Predator values seem, well, "short". Mountains being one of the chief examples (I just don't see something such as a Naga crossing mountain terrain at a rate anywhere near as high as their rate of travel across "regular" jungle). Unless ascending Giant Predator paths, mountains would not scale as dramatically against human rates as other GP paths. The "Giant Tree to Asc top" is a good example of what's wrong, a distance further from the Great Tree than the Waterfalls, with rougher terrain to boot, at barely a half-hour more.

Malahite wrote:

A few things:
1) How did you get 630 minutes for the Waterfall path for a Predator? Your values for the path would take only approximately 290 minutes to get that far.

Yeah, I figured this would be somewhat problematic; it's mostly the result of me trying to be clever, and the result of having too few paths to test.
But first let me clear something up, since it may be confusing: the column labeled "Expected Time (in minutes)" is the amount of time I think the path should take, and the column labeled "Calculated Time (in minutes)" is the time it actually took given the current hex costs. So the waterfall path currently takes 680 minutes, not 630. I've changed it around some in the spreadsheet so that it will hopefully be less confusing in the future.

The Giant Tree -> Chordoni Waterfall path for predators on my map has the following costs: 1x cost of climbing down out of the Giant Tree from wherever "home base" is located, 5x cost of traversing Regular Jungle, 2x cost of traversing Plains, and 1x cost of River Crossing. With the current values, the cost of 5xRegular Jungle is 250, 2xPlains cost 40, and 1x River Crossing costs 30, for sub-total of 320. The other 360 minutes of travel time come from the cost of climbing down out of the Giant Tree. This value was picked more or less arbitrarily, so it comes as no surprise to me that you think that everything seems to take too little time. Have a suggestion for what value it should be? The trade off of a lower value for it is that the smaller it is, the lower in the Giant Tree "Home Base" is.

Malahite wrote:

2) Many Giant Predator values seem, well, "short". Mountains being one of the chief examples (I just don't see something such as a Naga crossing mountain terrain at a rate anywhere near as high as their rate of travel across "regular" jungle). Unless ascending Giant Predator paths, mountains would not scale as dramatically against human rates as other GP paths. The "Giant Tree to Asc top" is a good example of what's wrong, a distance further from the Great Tree than the Waterfalls, with rougher terrain to boot, at barely a half-hour more.

I just plain screwed up the naming on that particular trip, it's actually not to the tip of the Ascarlin Mountains, but to the nearest edge, Specifically:



But yeah, in general everything feels too "small" to me as well.

I think we need to establish a chart of different terrains and their "values". Like for example swamp-type of terrain being high value, while plains would be low beause it's so much easier to travel ^^

Karbo wrote:

I think we need to establish a chart of different terrains and their "values". Like for example swamp-type of terrain being high value, while plains would be low beause it's so much easier to travel ^^

Yeah, I think I might just have to abandon my attempts at "massaging" the tiles travel costs out of people's estimated travel times, because the results are not very sensible. What I mean by this is that I have written a small program that very quickly tests thousands and thousands of possible tile costs. It takes these tile costs and calculates how long it would take to travel twelve trips specific trips (which are the same as the ones that are on the spreadsheet image that is in this thread a few posts back; I collected the trips from suggestions in this thread.) It then compares the time those trips take with how long the people in this thread [i]think[i] that trip should take.

I let my program run for 10 full minutes, and it calculated over 1,280,000,000 (yes, over 1.2 billion!!!) sets of tile values. The single best set of all those allows 10 of those 12 trips to be completed within their margin of error... but some of the tile values are more than a little funky.

The tile values (time it takes to cross a hexagon full of that particular terrain type):

Regular Jungle: 71.0 minutes
Medium Jungle: 359.0 minutes
Heavy Jungle: 540.0 minutes
Regular Mountains: 67.0 minutes
Plains: 14.0 minutes
Urban Area: 162.0 minutes
River Crossing: 101.0 minutes
Gaint Tree Climb Cost: 129.0 minutes


Yeah.... somehow I don't think that in a realistic setting it would be easier to climb "regular mountain terrain" (at 67 minutes per hex) than it would be to traverse "regular jungle terrain" (at 71 minutes a hex), and the time it takes to cross "plains" is alarmingly low. I think I'll poke at it a few more times, then give up and just make up values for the terrain. ^^;


(Of course on the other hand, using data from earlier in the thread, if a giant pred can cross a "plain" hex in 15 minutes at say 40 miles per hour, that gives the hex's a nice round diameter of 10 miles......which is probably far too big. At 10 miles per hex, almost all of France could fit onto the most recent Felarya map.





Here's a fairly nice one, program just spit it out after I added another constraint, after roughly 750,000,000 tests. It passes 10 of the 12 trips.
Regular Jungle: 72.0
Medium Jungle: 198.0
Heavy Jungle: 298.0
Regular Mountains: 83.0
Plains: 33.0
Urban Area: 274.0
River Crossing: 146.0
Gaint Tree Climb Cost: 80.0

Quote :

I let my program run for 10 full minutes, and it calculated over 1,280,000,000 (yes, over 1.2 billion!!!) sets of tile values.

Technology these days.

Anyway, I think that we shouldn't try to shoehorn the current map into this system.
We need a new map that actually has accurate sizes and distances on it.

I think a good way to do this is determine the number of Hexes that should be between each landmark. then fill in the blank spaces (with hexes) by looking at the current map.

Oldman40k2003 wrote:

(Of course on the other hand, using data from earlier in the thread, if a giant pred can cross a "plain" hex in 15 minutes at say 40 miles per hour, that gives the hex's a nice round diameter of 10 miles......which is probably far too big. At 10 miles per hex, almost all of France could fit onto the most recent Felarya map.

No offense to our denizens of France, but France isn't really that big - roughly twice the size of the state of Colorado. Large for humans perhaps, but for the current 'canon' (which of course doesn't cover all the unnamed/noncanon hybrids) giant hybrid population that's not a lot of roaming territory. I would have pictured the 'known' area of Felarya to be around the same land mass as Russia, with roaming territories of single predators akin to the size of the modern Siberian, perhaps larger. Especially for a predator like a centaur/pantaur - given their beastial natures, they're going to want to have the ability to roam.


A note on scale - the Jewel River in your map is around 40~50 miles at its widest shown points. The largest rivers on Earth are from 7 to 25 miles wide (dry/rainy seasons). The river doesn't really have those types of things I would think, so that generally means its always around the same size/depth. That size fits with how large I would imagine the river to be in its 'larger than life and Earth' mindset, but it calls into question some of the other areas when you hold them side by side, particularly mountain ranges. I think the problem overall you will continue to face with this idea is that as stated previously the map is not to scale, and thus applying a scale to it will create artificial distances that don't 'seem' right.

mhh I agree that 33 min seems really low for crossing a plain hex. Also I don't really get why an hex of jungle would be half the time of an urban area o.o

And to me the map is way way bigger than the size of France ^^;
I'm tarting to think I should indeed redraw a more accurate map

Not sure how such matters would work with Giant Predators, but small inclines can be down-right debilitating when it comes to rate of travel. For just a slight example: Try running on level ground. Okay? Good. Now try running up hill by about, say, 30°. Your rate of running slowed down a bit, didn't it? Now try a hill that's nearly 45° inclined.

Mind that this isn't even rough terrain, this is just a general incline change. While the chart's for a general "walking" pace, the same applies when even for walking (say that the average incline is about 30°. You're trying to cross a 3000m distance (horizontally). Your walk distance has just been increased by 500m, merely by the incline). When including rough terrain, uneven ground, sudden pitfalls / paths that won't do (for Predators, that rickety bridge won't hold their weight across the 120ft. gap at all. Meanwhile, a Prey can't cross the 20' gap without aid of some magical spell, whereas a tall / long-legged Dridder could carefully eek itself across and avoid having to either climb down then up, use magic spells and tools for the non-caster members, go around the gap, or so on), and so on, things get worse.

This link isn't exactly the best, and is applied only to human speeds (duh), but it gives a clear example in that the travel rate's practically doubled for an elevation gain of under a mile. This link corroborates, mentioning a drop upward of 66%+ possible in certain terrain conditions. When considering that many Felaryan obstacles are scaled to be to Giant Predators what the Earthly counterpart is to people, the rough terrain could be a real pickle for them (and practically a valley of plateaus for "Prey").

The map isn't entirely to scale, though, so it's a bit iffy just how much time should be applied.

Karbo wrote:

mhh I agree that 33 min seems really low for crossing a plain hex. Also I don't really get why an hex of jungle would be half the time of an urban area o.o

And to me the map is way way bigger than the size of France ^^;
I'm tarting to think I should indeed redraw a more accurate map

I would definitely be interested in seeing that. Hopefully the locales would remain in their relative locations.

oh yeah the things that would be changing would be basically the space between some elements.

Anime-Junkie (Sun Dec 12, 2010 6:39 am) wrote:

Anyway, I think that we shouldn't try to shoehorn the current map into this system. We need a new map that actually has accurate sizes and distances on it.

Probably. I can think of only one or two more solutions that can salvage the current map, and neither of them are all that useful or easy to implement. Still might post about them, simply to explain why they are impractical.

Anime-Junkie (Sun Dec 12, 2010 6:39 am) wrote:

I think a good way to do this is determine the number of Hexes that should be between each landmark.
then fill in the blank spaces (with hexes) by looking at the current map.

It also might be useful to define how far across a hex is first, then decide how long travel time across various types of terrain should be, and then decide how long you want it to take to get from landmark A to landmark B. By doing it this way you have a wide combination of hexs that all will sum up to the same travel time, meaning that there is a large number of choices for what terrain actually lies between A and B.

aethernavale (Sun Dec 12, 2010 7:32 am) wrote:

No offense to our denizens of France, but France isn't really that big - roughly twice the size of the state of Colorado. Large for humans perhaps, but for the current 'canon' (which of course doesn't cover all the unnamed/noncanon hybrids) giant hybrid population that's not a lot of roaming territory.

All true, though a lot of it depends on just how much food there is in an area; the higher the food density, the less territory needed to supply nutritional needs.

aethernavale (Sun Dec 12, 2010 7:32 am) wrote:

A note on scale - the Jewel River in your map is around 40~50 miles at its widest shown points. The largest rivers on Earth are from 7 to 25 miles wide (dry/rainy seasons). The river doesn't really have those types of things I would think, so that generally means its always around the same size/depth. That size fits with how large I would imagine the river to be in its 'larger than life and Earth' mindset, but it calls into question some of the other areas when you hold them side by side, particularly mountain ranges. I think the problem overall you will continue to face with this idea is that as stated previously the map is not to scale, and thus applying a scale to it will create artificial distances that don't 'seem' right.

Yeah, after reading your comment I was considering marking it as the "Jewel River Esturary", or suggesting a small change in the map so that most of the Jewel River is actually river sized, with it blooming out into an estuary at the very end. The size of the mountain ranges is another good point; at one point I was going to use the size of a local "mountain cluster range" to help figure out the size of the "mountain cluster range" that is Frost Peak, but in the end I didn't think it would work out all that well. (The Olympic Mountains, while a "mountain cluster" like Frost Peak appears to be, are not volcanic in nature, and less than half as tall.)

Karbo (Mon Dec 13, 2010 1:48 am) wrote:

mhh I agree that 33 min seems really low for crossing a plain hex. Also I don't really get why an hex of jungle would be half the time of an urban area o.o

And to me the map is way way bigger than the size of France ^^;
I'm tarting to think I should indeed redraw a more accurate map

Well, since there were only a few trips programmed into the program, some of the terrain types ended up being "leftovers", used to add a small distance while all the real distance was covered by the jungle hexes. I could have mitigated it a little bit with a better means of scoring the solutions, but that would have still left the two major problem unsolved: not enough trips in the program, and inconsistent scaling between people.

It actually wouldn't surprise me terribly much if a predator did move faster in a jungle terrain than in an urban environment. Cities are built for humans, at human scale. They are full of streets in a roughly grid-like pattern that, to the inexperienced, don't point you toward a particular location, or give you direction. The buildings and ruins of buildings are too big to easily climb over, and might be big enough to block long distance sight.

Malahite (Mon Dec 13, 2010 8:18 am) wrote:

Not sure how such matters would work with Giant Predators, but small inclines can be down-right debilitating when it comes to rate of travel. For just a slight example: Try running on level ground. Okay? Good. Now try running up hill by about, say, 30°. Your rate of running slowed down a bit, didn't it? Now try a hill that's nearly 45° inclined.

Mind that this isn't even rough terrain, this is just a general incline change. While the chart's for a general "walking" pace, the same applies when even for walking (say that the average incline is about 30°. You're trying to cross a 3000m distance (horizontally). Your walk distance has just been increased by 500m, merely by the incline). When including rough terrain, uneven ground, sudden pitfalls / paths that won't do (for Predators, that rickety bridge won't hold their weight across the 120ft. gap at all. Meanwhile, a Prey can't cross the 20' gap without aid of some magical spell, whereas a tall / long-legged Dridder could carefully eek itself across and avoid having to either climb down then up, use magic spells and tools for the non-caster members, go around the gap, or so on), and so on, things get worse.

Yep, all very true. You have no idea how hard it was to restrain myself from having the map have two or three "layers": one for the surface material, one for the elevation, one for the roughness, etc. Then I would have had a complicated formula that took all of those factors into consideration (or derived factors, like slope), how fast you were trying to travel, what sort of creature you were, into consideration. I had to keep reminding myself that I was not writing an algorithm that was going to be used only by a computer, and thus the algorithm needed to be very easy to do by hand.

Oldman40k2003 (17/12/2010, 1:47 am) wrote:

I had to keep reminding myself that I was not writing an algorithm that was going to be used only by a computer, and thus the algorithm needed to be very easy to do by hand.

AS I said before...

Anime-Junkie wrote:

It's even be possible to make a flash (or something else) application that allows people to make a route by clicking on a hexagon divided picture of the map. The application would then automatically calculate the travelling time.

So if you want to make that particular algorithm, go right ahead. It would yield more accurate results...

Anime-Junkie (Thu Dec 16, 2010 7:30 am) wrote:

As I said before...

Anime-Junkie wrote:

It's even be possible to make a flash (or something else) application that allows people to make a route by clicking on a hexagon divided picture of the map. The application would then automatically calculate the travelling time.

So if you want to make that particular algorithm, go right ahead. It would yield more accurate results...

Shhh, don't temp me! (I might make it, but that will be after we have a fixed map and other things are taken care of. Otherwise it would probably be putting the cart before the horse and all that.)

Oldman40k2003 (Thu Dec 16, 2010 6:47 am) wrote:

I can think of only one or two more solutions that can salvage the current map, and neither of them are all that useful or easy to implement. Still might post about them, simply to explain why they are impractical.

So, if anyone cares, here are a few ways I thought up that could save the current map; sadly they are impractical for a few different reasons.


The major reason that the current map needs to be "saved" is because it (apparently) has the wrong distances between various important landmarks. Now one could take the easy way out and simply claim that the map was never meant to be topographically accurate, but since this is a map of Felarya, there is another, better explanation: non-Euclidean geometry. Euclidean geometry is "flatness"; graph paper is a good example of a 2d Euclidean surface. No matter how far you travel in a straight line, you will never get closer to where you started from. The surface of a sphere, on the other hand, is non-Euclidean; if you were to start walking in a straight line in any direction, eventually you would end up back where you started! Obviously it makes sense that Felarya, land of dimensional instability, would be non-Euclidean in nature. So the question becomes: how to represent a non-Euclidean space in two dimensions?


No, a spherical Felarya, while non-Euclidean, won't help us here.

Method #1: using a graph/network

If you follow the arrows, you can obviously see that it takes much less time to go from the Giant Tree to Bulvon Woods than it does to return, so a graph/network can display a warped space.
Downsides: Paths and locations must be explicitly provided, there is very limited freedom of movement. For example, from this graph/network it isn't possible to find the time it takes to travel directly from Bulvon Wood to the Deluran Underground Base.


Method #2: Using mathematical formulas to calculate the "warped space".
I'm not certain I even know the math required to do this, and while it would be very precise, it would almost certainly require a computer to calculate any trip at all.


Method #3: Applying the distortions directly to the existing 2d map.
Assuming that there are no discontinuous dimensional warps (like giant region sized portals) then the "warp" caused by non-Euclidean space could be applied to the current map, resulting in a new map that is distortion free. Sections of land would be "expanded", others would be shrunk. Some areas might aquired a twist or two, and then, at the end, a new hex grid would be laid down over the result. The problem is, this would essentially be redrawing the map over from scratch, but with the disadvantage of not having complete control over what gets moved where.

Spoiler:

So in summary, remaking the map is probably the only solution to our problems.

Oldman40k2003 wrote:

It actually wouldn't surprise me terribly much if a predator did move faster in a jungle terrain than in an urban environment. Cities are built for humans, at human scale. They are full of streets in a roughly grid-like pattern that, to the inexperienced, don't point you toward a particular location, or give you direction. The buildings and ruins of buildings are too big to easily climb over, and might be big enough to block long distance sight.

Mhh looking back at it I think you are right

okay, this may be a stupid suggestion (because I don't know much about map-making)
what if instead of squares, or triangles, or hexagons used too make a grid, we had circles, and the interlocking parts of the circles would be areas of felarya, that would feel kind of like the bermuda triangle? it's hard to gauge direction in those parts, and it can be one reason why going from the great tree to two different areas can feel like it took way longer even though they're the same distance apart.