• Good day, Stranger! — Are you new to our forums?

    Have I seen you here before? To participate in or to create forum discussions, you will need your own forum account. Register your account here!

Answered Trader range and number of trades

DeletedUser

Guest
NOTE: This thread has become quite pic intensive due to the many informative visuals posted by members, and can take some time to load.
Please keep that in mind. :)

So... my first post (meant 'posted thread') in this section. :p

Now here are two questions I've been wondering about for some time:

1.What exactly is the maximum range of the trader for neighbours?

2. Is there a maximum limit to the amount of trades you can see from other players (apart from the limit of posting a max. of 60 trades per player)?​

The furthest trade I've come across (that I can remember) was 21 rings out from my city.

As for the limit of how many trades one can see... I've been wondering whether it's just pure bad luck or I'm actually not able to see many of the trades posted in my neighbourhood. The reason I come to this conclusion is simple: I check the trader hourly about 12-16 times on some days, though I'd say 6-10 times a day is more common, and until recently (about a week ago) I haven't really seen many neighbourhood trades posted. That appears odd for two reasons:

1. If my trader's range is at least 21 rings out from my city, then that would make a potential 462 neighbours to trade with if my calculation is correct. Taking into consideration the enormous amount of inactive players, I'd argue about 20% of those 462 are active (at best) - leaving about 92 potential trade partners.​

2. Now taking into consideration that there is a good number of (active?) players in fellowships, one could think they've got trades posted every now and again right?​

Well, if they ever do... Then I have hardly ever seen them (as from October until about two weeks ago maybe two to three neighbours had trades up every now and again). I mean, am I the only player out of an estimated 92 active players (and theoretically 462+ players) that posts trades regularly, never mind up to the full 60 trade limit (which I often reach)? This just seems odd. And before you ask: Until very recently, my fellowship members were the only one's to accept any of my trades.
 
Last edited by a moderator:

DeletedUser

Guest
  • There's no known limit on the number of trades that you can see
  • In addition to the neighbors that you have discovered, you'll also see radio button traders who are on the fringes of your discovered territory, up to a limit of 200 traders.
  • If you visit a radio button trader, click on their World Icon, and drag the map just a bit, you can follow the blue arrow back to your discovered territory. "Fringe" seems to be a pretty generous concept, as I have a few radio button trading partners who are at least a half dozen rings outside of my discovered area.
  • I haven't seen any reports regarding what happens when you've already discovered 200 neighbors, but at that point you're going to be swamped with manufacturing in any case, because you'll have to negotiate rather than fight.
 

DeletedUser

Guest
Thank you Katwijk,

I was hoping you might respond here - as you are one of the most knowledgeable players in our community for this kind of subject in my opinion. :)

  • There's no known limit on the number of trades that you can see
I would certainly hope so. Wondering whether there is a limit if one were to consider the extreme of (using your 200 neighbour limit you mention below) 200 players x 60 offers each = 12000 trades posted. It's still far too early in the game for us to come across such a scenario though...

And by the way: where did you get your info about the limit being 200 neighbours?

  • In addition to the neighbors that you have discovered, you'll also see radio button traders who are on the fringes of your discovered territory, up to a limit of 200 traders.
Now this I'm a bit confused about. What exactly do you mean by "radio button traders"? I can't quite get my head around the "radio button" part (only know that to be the graphical control element for selections, which makes little sense to me here). Or are you saying the undiscovered "radio button" traders are predefined and only a select few per ring visible in your trader for example?

  • If you visit a radio button trader, click on their World Icon, and drag the map just a bit, you can follow the blue arrow back to your discovered territory. "Fringe" seems to be a pretty generous concept, as I have a few radio button trading partners who are at least a half dozen rings outside of my discovered area.
I did this in a slightly more complex manner to be honest (though much easier to understand IMO): Zoomed the world map out to 25%, took a screen shot, then painted in the rings (through which I learned I used to count the rings wrong, lol), and then determined which ring a specific player belongs to. That's how I discovered the furthest trade partner I've had so far was 21 rings out from my city, and 13 rings out from the furthest ring I've discovered (which is the 8th ring).

  • I haven't seen any reports regarding what happens when you've already discovered 200 neighbors, but at that point you're going to be swamped with manufacturing in any case, because you'll have to negotiate rather than fight.
Hmm... Guess I could go asking around. NecroDee might be able to answer this one perhaps, or Goldie...?
 

DeletedUser

Guest
We're happy to announce an additional change to the release notes, which is in effect as of today:
  • The amount of possible trading partners has been increased to a maximum of 200 (was: 100), which should give you more available offers in the Trader
See https://beta.forum.elvenar.com/index.php?threads/release-notes-version-0-15.2332/#post-13507

Radio Buttons are a term of the art for "pick one of multiple choices" buttons that are typically presented as a group of round buttons. In the case of the trader they're actually just a passive info icon, but they sure look like a puma.
RadioButton.jpg


In an hexagonal array, each even numbered ring will have 6 cities more than the previous even numbered ring, so the progression is

Cities Total
1 1
6 7
12 19
18 37
24 61
30 91
36 127
42 169
48 217
54 271
60 331
66 397
72 469

The least confusing way to count rings is to count on a diagonal radial as you'll thereby only be seeing even number rings, and then you can likewise follow the the outer edge of the equilateral wedge to the next diagonal radial. It's convenient to think of the radials as 1 o'clock, 3 o'clock, ... ll o'clock, which is the convention that I'm using in https://us.forum.elvenar.com/index.php?threads/felyndral-neighborhood-watch.1247/page-2

There's an ok visual at https://en.wiki.elvenar.com/index.php?title=World_Map#Province_Distance
  • Scouting costs involve two factors. A base cost for the ring, plus an incremental cost for the number of scouted sectors.
  • The Global Rank value of a relic is weighted by Ring distance
  • The difficulty of combat increases with distance, but similar sectors will always have similar defenses.
    .
  • Your own boosts are determined by the sectors to your immediate southwest, southeast, and double southeast (which is always identical to north)
  • Every 2nd sector on a ll o'clock to 5 o'clock diagonal will have identical relics, every 3rd sector if you want to count the city sectors as well
  • Your most valuable trading partners are therefore in the diagonal to your double southwest and your triple either southwest or southeast
  • Adjacent relic rows are offset by 4.5 of 9 relics, so you're never more than 5 rings away from a trading partner with any one of the nine unique city+relic combinations, and more trading partners of that type will be diagonally adjacent to them, in every second city.
 
Last edited by a moderator:

DeletedUser

Guest
First off, thanks for the very detailed information Katwijk. :)

I was made aware of most of the points you mention (in that last bit you added to your post later on), through my search of trying to find answers to my two questions. Certainly useful information regardless. It does not quite satisfy my curiosity though I'm afraid. :D

In an hexagonal array, each even numbered ring will have 6 cities more than the previous even numbered ring, so the progression is

Cities Total1 16 712 1918 3724 6130 9136 12742 16948 21754 27160 33166 39772 469
Not sure I can agree with that statement. According to the graphic I created based on my initial findings and the visual you've linked to from the wiki (which was how I had first learned that I counted rings wrong), I'm getting an increase of 6 neighbours after every 3rd ring after the 1st ring.

Here my graphic:​

neighbourhood_grid.png


Ring #City # / RingTotal # of Cities

[TR1][TD1]1[/TD1][TD1]0[/TD1][TD1]0[/TD1][/TR1]
[TR2][TD1]2[/TD1][TD1]6[/TD1][TD1]6[/TD1][/TR2]
[TR1][TD1]3[/TD1][TD1]6[/TD1][TD1]12[/TD1][/TR1]
[TR2][TD1]4[/TD1][TD1]6[/TD1][TD1]18[/TD1][/TR2]
[TR1][TD1]5[/TD1][TD1]12[/TD1][TD1]30[/TD1][/TR1]
[TR2][TD1]6[/TD1][TD1]12[/TD1][TD1]42[/TD1][/TR2]
[TR1][TD1]7[/TD1][TD1]12[/TD1][TD1]54[/TD1][/TR1]
[TR2][TD1]8[/TD1][TD1]18[/TD1][TD1]72[/TD1][/TR2]
[TR1][TD1]9[/TD1][TD1]18[/TD1][TD1]90[/TD1][/TR1]
[TR2][TD1]10[/TD1][TD1]18[/TD1][TD1]108[/TD1][/TR2]
[TR1][TD1]11[/TD1][TD1]24[/TD1][TD1]132[/TD1][/TR1]
[TR2][TD1]12[/TD1][TD1]24[/TD1][TD1]156[/TD1][/TR2]
[TR1][TD1]13[/TD1][TD1]24[/TD1][TD1]180[/TD1][/TR1]
[TR2][TD1]14[/TD1][TD1]30[/TD1][TD1]210[/TD1][/TR2]
This would also mean that once reaching ring 14, a player will have "already" hit the trading cap of 200. The post you quoted from the beta forums doesn't specify if the cap is referring to just undiscovered players or any potential trade partners in general. My understanding therefore is that the total of any combination of undiscovered and discovered neighbours as trade partners is capped at 200. Guessing once you've discovered more than 200 players only the first 200 you've discovered would remain as trade partners then?

EDIT: Just realised now that your table adds up for the number of tiles per ring and tiles in total, beginning with the first tile (player's own city) as ring #1 (which I'm excluding by the way). Still doesn't add up for me in regards to the number of cities though...
 
Last edited by a moderator:

DeletedUser

Guest
If you plow through the information at http://www.redblobgames.com/grids/hexagons/, you'll find that we're using a "flat topped odd q" arrangement for the hexagons on our world map, so you'll need to rotate your model by 30 degrees.

If you go to the outermost ring of any Elvenar world map, you'll be able to see that the vertices are on the odd-numbered radials. I happen to have the data for us3, so take a look at https://us.forum.elvenar.com/index.php?threads/felyndral-neighborhood-watch.1247/page-2. Our current vertices are:

Since Score Name (us3) Radial Ring Status
10/16/15 3286 Nanabarb 0 o'clock Ring 000
12/28/15 110 sherri30721 1 o'clock Ring 114 was agoodone49
11/10/15 29 sundin 1 o'clock Ring 214
1/2/16 26 krionos 1 o'clock Ring 244
1/2/16 22 Doridan 3 o'clock Ring 244
1/7/16 0 yoal3 5 o'clock Ring 244
1/7/16 161 RedDeadCabbage 7 o'clock Ring 244
1/8/16 0 Lingman 9 o'clock Ring 244
1/2/16 183 Terrhkklh999 II o'clock Ring 242
1/8/16 134 tbard324 SpiralE Ring 244 10:59
 
Last edited by a moderator:

DeletedUser

Guest
Er... Why rotate the model by 30°? I am using a "flat topped" arrangement in my graphic just like this depiction bellow from the website you linked:

upload_2016-1-10_1-34-55.png
 

DeletedUser

Guest
To state the issue more clearly, your map is fine with a flat-topped sector at 12 o'clock, but your RINGS need to be rotated 30 degrees to match the filling sequence that's being used on the Elvenar maps. We're flat-topped both locally and globally.

Take a look at your world map. The top of the populated area is a horizontal row from 11 o'clock to 1 o'clock. Once you rotate the rings to match what we have on the world map, the patterns become a lot more obvious. The top and bottom of the world map are both rows of cities, while the other 4 sides have cities on a clean diagonal.
 
Last edited by a moderator:

DeletedUser

Guest
Here the graphic rotated at 30°:

path7338.png


This seems completely wrong to me though... At least for the world map that is. If what you say is true about the ring formation on the outer edges of the map, then the rings might be counted differently? Something's certainly out of place though.
 

DeletedUser

Guest
This seems completely wrong to me though
Oh I agree! :p Unfortunately it's not that easy, BOTH the local hex AND the Ring itself need to be flat-topped.

If you click on your on city, and then scroll north on the map, you'll find one of three conditions.
  • A 9 o'clock to 11 o'clock outermost diagonal ring if you're near the left edge of the map
  • An 11 o'clock to 1 o'clock outermost ROW ring if you're toward the middle of the map.
    At 12 o'clock there will be a jog, due to the start of a new ring.
  • A 1 o'clock to 3 o'clock outermost diagonal ring if you're near the right edge of the map
If you go to the us3 map and visit genevadugger, then click on the world map icon, you'll get a full-sized view of ring 244, which STARTS at noon and then progresses clockwise in a hexagonal spiral that's formed by jogging to a new ring, at 12 o'clock, each time a ring is completed.

StartOfRing.png
 
Last edited by a moderator:

Deleted User - 13667

Guest
This seems most helpful. Thanks a bunch for the effort to share this, guys/gals

:)
 

DeletedUser

Guest
I've just gone full circle around the entirety of the world map on Winyandor here on the EN servers and I can confirm that you are right, Katwijk. New players are placed in a clockwise spiral and this is a flat-topped pattern with diagonal sides. I first noticed the clock-wise rotation myself when I first started playing the game on the still relatively new world of Winyandor as it was very evident at the time. Never thought much of the pattern though. Apart from viewing the outskirts, one can also see the clockwise rotation when looking at the 'Swiss-cheese-like-belt' (caused through the deletion of inactive players) that is gradually shifting to the right and being filled from the left (while at the top of the map).

There is one small issue outstanding however that I had come across while taking a closer look at the world map. The rings for provinces don't match the rings for the placement of new cities. Based on the costs of scouting new provinces, the rings are pointed-topped and not flat-topped. Therefore my initial graphic is correct for provinces and determining their distance in relation to costs.

So does this mean there are separate rings in place for cities and provinces? Or am I still missing something here?

P.S.: Will update the graphic for city rings later on.
 

DeletedUser1316

Guest
You both are so excellent! :)

Thank you for posting all this useful information! I love seeing players get together and posting things like this, it makes me feel all happy. :p

Btw - I loved the tables and graphics the most. *giggle*
 

DeletedUser

Guest
So does this mean there are separate rings in place for cities and provinces? Or am I still missing something here?
Now you want consistent too?? :rolleyes:

I'm an electrical engineer, so I'm completely comfortable with transforming a problem into a space where the calculations are easy. In our particular instance, the most likely approach is that the programmers are transforming the map into an orthogonally viewed 3 space so that distances are simple addition, and then transforming it back again for presentation. If you try to think about it in 2-space, you'll just get a headache.

Without access to the Elvenar database schema it's not possible to determine the most efficient approach, and even then the programmers probably didn't write anything from scratch. Curiously, it's not likely that they used the Forge of Empire GvG map model, wherein the rows, rather than the columns, are staggered.

I prefer using the flat-top local plus flat-topped global model for analysing the Elvenar world map, because the patterns are more obvious which helps with finding similar cities and relics, which is important to me, while scouting costs and relic values average out anyway, regardless of how you manipulate the rings.
 
Last edited by a moderator:

DeletedUser

Guest
I did some more noodling, and I suspect that the distance/cost calculations do actually use the 3-space model that you illustrated in https://en.forum.elvenar.com/index.php?threads/trader-range-and-number-of-trades.1827/#post-11770 from http://www.redblobgames.com/grids/hexagons/.

If that's the case, and assuming that cities are regarded as null relic sectors, the distance calculations are likely to be 3-space local, based on the coordinate offset distances. To accurately represent the equidistant sectors in a transform 2-space, you'd need to use a polygon of order =INT(2*PI()*r), which wouldn't be a hexagon at all except for the very first ring, because the higher order polygons would bulge out a bit between the odd-numbered global map vertices.

Regardless of how we rotate the rings, or which spaces we use, in the worst case the distances will vary smoothly between r and SQRT(3)/2*r, and the variances will cancel out every 60 degrees.
 

DeletedUser

Guest
Okay, so I've 'refined' the initial graphic I made based on the relic cost distribution using a pointed-topped hexagonal pattern. (It's effectively the same graphic as before, just with more numbers, :D)

neighbourhoodringrefined.png

To my current knowledge, this graphic above is accurate in regards to how the developers presumably want us to count the rings. I've looked at rings 6-8 specifically and the costs to scout these provinces are definitely in a pointed-topped pattern and include player cities as "null relic sectors". (The visual found in the official Wiki that you've linked in a previous post, Katwijk, follows the same principal.)

As for the rings of how player cities are added to the game, this spiral depicts the respective principal the game appears to follow:​

nhringspiral.png

I suspect the reason behind the spiral following a flat-topped pattern is to distribute new players along the edges more effectively in relation to the relic boosts they receive. If the pattern were pointed-topped, then every player added along the south-western edge or north-eastern edge would have the exact same boosts, until the spiral reaches the next corner.

EDIT: changed the 'spiral' graphic according to Katwijk's studies. Nevermind :p
 
Last edited by a moderator:

DeletedUser

Guest
Yeah, the 9-bar on the relics is pretty convenient.

9Bar.jpg

Calculating distances in 3-space is just SQRT(SUMSQ(deltaX, deltaY, deltaZ), using the conventional orientation with z at 12 o'clock, x at 4 o'clock, and y at 8 o'clock. If you were to model the results in 2-space they would converge to a circle that has the same area as the current polygon, but a pointy topped hexagon is a close approximation for small regions.

If you're near the 2 left or 2 right edges of the global map you're chasing a diagonal boundary in any case, and a distance calculation that nicely approximates the outermost edge of the global map 4 times out of 6 is actually a very elegant solution, and even on the top and bottom edges you're only losing the tip of the hexagon.
 
Last edited by a moderator:

DeletedUser

Guest
Knowing the relic distribution pattern is very helpful in many ways. I made a similar graphic using a spreadsheet a couple months back, however it looks a bit different (I used alternating rows). Might add an improved version to this post/thread later that matches the style of the other graphics I've posted here. As for the initial questions I raised, there is a limit to both the distance and number of trades a player can see in the trader:
  • The maximum number of trades visible must be 12000 as the number of trade partners is capped at 200. (200 trade partners x 60 trades each = 12000 trades)
More interesting however:
  • The cap of 200 trade partners would mean the furthest trade partner you could currently have would be in ring 198 + the maximum distance an undiscovered player can be as a trade partner. Generally speaking the maximum distance would be a player in ring 199 though. This would be the case if you were to only scout in one direction from the very start of the game.
  • If a player scouts ring by ring (effectively making the most efficient use of province costs), then after discovering their 200th neighbour, the furthest trade partner they could have would be in ring 14.
These two facts are rather interesting when taking into consideration the amount of inactive players within the game. It is thus extremely inefficient from a trading perspective to scout ring by ring, as this reduces the maximum distance to any potential trade partners considerably over time. Far more efficient would be to scout in a "star-like" formation from the corners of a ring in one consistent direction. If you match your scouting pattern to your respective relic boosts, this also increases your efficiency in gaining useful relics. (Don't worry, you'll still get plenty of non-boosted relics for AWs in the process.)
 

DeletedUser

Guest
"Bump" :p

As mentioned in my previous post above, here the improved graphics explaining the distribution of relics/provinces and concluding with a brief explanation of what I meant with scouting in a "star-like" formation. Enjoy. :)

To begin with, the graphic bellow displays an entire "sequence" of the relic distribution pattern in a rectangle form (as you would likely notice it on your screen). In the next image I will prove that this pattern repeats itself.
pattern1.png


patterproof.png


These sequences can be very useful when navigating the world map as they can be divided into two simple repeating patterns:

horizontal-7.png


vertical.png


Alternatively you can argue the entire relic pattern is made up of just a repetition of this image:

base.png


This also allows for a third diagonal pattern in direction of North-West to South-East and South-West to North-East respectively. The following example should highlight this for the North-West to South-East diagonal:

diagonal1.png


With this information, planning the efficient scouting routes should be much easier. As an example: We'll assume our boosts are Steel, Crystal and Elixir. So here's how scouting in a "star-like" formation could be beneficial in this case:
scouting2.png


Admittedly, it doesn't look quite like a "star", but rather more like parallel lines if you want to be specific. The scouting process is a bit "star-like" as you will scout the closest relic provinces first, then gradually move farther and farther away however not scouting each ring fully.​

If you liked this, let me know and I might just turn it into a guide. :)
 
Last edited by a moderator:

DeletedUser

Guest
There are two considerations that you need to think about, as we might as well be entirely accurate if we're going to put this much effort into it. I can supply graphics if words don't suffice.
  • While the "star" will pick up your boosts, it neglects your ideal trading partners, and you NEED somewhere cheap to exchange all of those boosted goods. I'd even argue that chasing active radio button traders who have your missing boosts is the "best" strategy, and that your own boosts will fall into place by happenstance.
  • On the spiral, it's more elegant to create a new ring by overrunning the 1 o'clock corner, but that's not what actually happens. Each ring starts at 12 o'clock, and when the ring closes at 11:59, there's a jog and a new ring is started, again, at 12 o'clock. You can verify that pattern with a close look at us3: genevadugger, or by looking at the top edge on any Elvenar world map, at exactly 12 o'clock where you'll find the start of the most recent ring.
 
Top