• Dear forum reader,

    To actively participate on the forum by joining discussions or starting your own threads or topics, you need a game account and to REGISTER HERE!

World Map Climbing the Penrose Stairs

DeletedUser

Guest
For sure there is the urgent need to make possible proper city movement, but this has to be done in a way, that there are no changes on important existing game features.

Edit: With your city moved you should NOT get new chances to complete very cheap provinces for example....
 

DeletedUser867

Guest
I very much agree with you that average costs should not be effected, and folks who are roaming farther and farther from their home base should certainly have higher costs.

What we DON'T KNOW is the underlying assumptions that InnoGames made when they selected the current cost structure.
  • It may be that they expected to have a huge level variance in static neighborhoods, with a noob being tossed in whenever an inactive player is removed. If that's the case then we'd need to adjust, perhaps flatten, the cost structure.
  • It may be that they expected to optimize the map all along, and thereby minimize the level variance in each neighborhood, in which case we're merely talking about the movement algorithm, not the underlying cost structure.
I'd prefer active neighbors who are near my level, to facilitate trading and polishing. Guilds, when they are implemented, will probably take some of the pressure off of the neighbors, but I suspect that there will eventually be SOME SORT of neighborhood optimization, and I hope that it will be
  • Elegant
  • Fair
  • Strategically interesting
 
Last edited by a moderator:

DeletedUser

Guest
With all respect for idea and effort that went into this proposal I cannot but oppose it (for comparatively non-mathematical reasons):
This is game is a city builder providing an overland map, where you build your settlement at a certain location, providing you with certain (boosted) goods and certain neighbours you all get used to, in short: Static surroundings in which you make a home (something I like in Elvenar, and much more than a mere neighbourhood list in FoE). I would therefore not like constant shifting or changes of these surroundings I have become accustomed to, even if it would to some extent further my material needs.
Neither does something as large as a city perpetually teleport around the world, nor are my people wandering nomads. I wish them to stay where they are.
I acknowledge the necessity to remove and replace inactive Players/cities for gameplay reasons, but that's as far as it goes. If I seek new lands and neighbours, I explore outwards.
 

DeletedUser867

Guest
Ah, but "people" do move all the time. Every few years. More kids, different job, better schools, a swimming pool, you name it.
Static neighborhoods are a fiction.

Regarding cities, when they not getting it done we throw the bums out and elect new officials or (but hopefully not in Elvenar) a city is conquered. So while I'd agree that the physical city isn't a trailer park that you can just pick up and move, the residents of a particular location DOES change all of the time.

Equally to the point, as you progress, your like-minded neighbors will move WITH you, and you will, in fact, find more like-minded friends.
Toss your coin again, and hopefully it will come up Heads this time. :rolleyes:
 

DeletedUser

Guest
You don't have to prove that people/residents do move all the time.
To be clearer - I dislike the thought of Paris suddenly appearing north of London so to say, as a mere calculatory consequence, and some time later somewhere else again, even if London moved with Paris, and even if Paris gets more like-minded friends that way, as you put it. If you introduce such shifting and arbitrary localities, you could as well scrap the overland map (which in my opinion is a nice feature adding a territorial feeling to the game) alltogether.
 

DeletedUser867

Guest
Perhaps you'd be more comfortable thinking of the World Map as a College Dormitory, or as an Office Cubicle farm?
I think those are actually better analogies, because
  • Our neighborhoods are very homogenous, and are neither unique nor funky
  • We're all doing pretty much the same things
  • We have every intention of moving to a better location as soon as we can.
You can personalize your "city" to a very limited extent, but it's a village at best and it's NEVER going to be Paris nor London.
 
Last edited by a moderator:

DeletedUser

Guest
I wouldn't say that the top players have a village. A city is first a place of trade. Active players are constantly trading and doing whatever a city makes a city besides having a paper declaring it a city.
 

DeletedUser43

Guest
I'd like to know if I understand the idea correctly. Please correct me if I'm wrong. ;)

The idea is to have the nr. 1 player at the center of the World Map and to let the other players spiral out, sorted by ranking. The placing of the cities would be adjusted on a specific date, e.g. one per week. You could predict where your city will be places when you keep an eye on your own ranking (and the rankings of your neighbours, if you want to keep your city close to theirs). Once you've discovered a city or province, this city or province will stay discovered when your city is moved. Because your city moves, you could be placed next to undiscovered provinces, which will be cheaper to scout and which will have easier encounters, as they're closer to your city. So, with all your actions, e.g. gathering relics, scouting provinces or negotiating/fighting encouters, you should take into consideration where your city might be moved to.
 

DeletedUser867

Guest
>>> Very close <<<
You could predict where your city will be places when you keep an eye on your own ranking (and the rankings of your neighbours, if you want to keep your city close to theirs).
Click to expand...
I've separated Rank and the underlying City Score.
  • Rank is where you are on the map. It's a street address. The rank of each particular city sector could be hard-coded. For example, the highest ranked city sector in a particular ring will ALWAYS be at the 12 o'clock position, and it will always be the square of an odd number.
  • As the program walks the hard-coded spiral, once a week seems about right, the system would consider the 3x3 grid that's centered on YOUR city. Those 9 cities would be sorted by the UNDERLYING CITY SCORE, with the highest scoring city ending up in the highest ranked sector in the 3x3 grid, ... , and with the lowest scoring city ending up in the lowest ranked sector in the 3x3 grid.
So there's a very high correlation between your underlying city score and your rank, but they are NOT the same thing.
Let's look at Rank #15 in the following example.
  • The three players with the highest scores will end up in ring 3, which contains ranks #2 through #9 (3 squared), in Rank #3, #4, or #5.
  • The three players who score in the middle will end up in ring 5, which contains ranks #10 through #25 (5 squared), in Rank #14, #15, or #16.
  • The three players with the lowest scores will end up in ring 7, which contains ranks #26 through #49 (7 squared), in Rank #33, #34, or #35.
Spiral.jpg
Spiral.jpg
compass.png
compass.png

Let's look at the consequences of walking the spiral starting with the #1 ranked sector in the center of the map.
  • Dedicated players will have a chance to climb into a higher ring, once per week, without spiraling completely around the mountain. If they were in ring 9, for example, then when ring 7 is calculated they'll have a chance to capture one of the top 6 ranks, and be pulled up a ring, or even two rings. But they'd only get three tries per week to climb, while the three adjacent sectors in the next higher ring were being calculated.
  • Most players will remain in the middle ring, or perhaps wobble up and down a bit, and will spiral around and around the mountain as their city grows, happily exploring additional relic sectors and discovering new neighbors, or renewing old acquaintances from several weeks ago.
  • If a player deleted all of the buildings in their city, which would dramatically drop their underlying city score, they could fall a LONG WAYS. They'd fall a ring when 7 was calculated, fall another ring when 9 was calculated, etc, most the way to the fringes of the map.
The dedicated players will have the tactically interesting challenge of trying to make sure that their underlying city score, just BEFORE the recalc, scores them as one of the top three cities in their local 3x3 grid. In particular, they'll need a HIGHER score than at least one the three folks who were placed immediately above them the previous week. But those guys know that too, and they might be sandbagging.

OR, as you indicated, friends might want to to keep their city scores similar to AVOID getting separated. Folks who are growing at a similar rate will tend to stay clustered, although the interior rings will rotate a bit faster than the exterior rings, which have a few more cities.

Archiving inactive players is inherent in the algorithm.
  • Dealing with low activity players is simple, they'll drift to the fringes of the map where nobody will be depending on them, as average scores climb, and if they quit playing altogether you'd simply vacate their city and move the lower ranked cities up one rank, as you go through the optimization.
  • Reactivating an archived player is very nearly the same operation. When they return they can temporarily "rent a room" in the city that has the next higher underlying score and can use that city as their temporary base of operations. During the next recalc you simply insert their city in the next lower rank, which will bump the lower ranked players down one rank.
  • Note that a deletion or an insertion rotates EACH of the lower rings a bit. The neighborhoods will skew a bit, because the lower rings have a few more cities, but near neighbors will still be near neighbors even if they are in adjacent rings, because BOTH rings will rotate by nearly the same amount.
The IMPORTANT concepts are that the optimization will be:
  • Elegant: You'll move smoothly toward the center of the map, based on your efforts
  • Fair: You'll gain rank because you gained more city points than your nearest neighbors. Leaving the doorstops behind is probably MORE important than your own progression, psychologically, because you'll know that you're not going to be stuck with them forever.
  • Strategically Interesting: Depending on where you think you'll end up as a result of the recalc, you'll want to clean up some of the sectors "behind" you, as they will become more expensive, while you might want to wait a week before you clear out the sectors that are "ahead" of you, as they'll be cheaper next week. If you guessed right. :cool:
 
Last edited by a moderator:

DeletedUser43

Guest
Alright, I get the difference between Rank and City Score now. :) The only thing I don't understand yet, is this:
Katwijk said:
As the program walks the hard-coded spiral, once a week seems about right, the system would consider the 3x3 grid that's centered on YOUR city. Those 9 cities would be sorted by the UNDERLYING CITY SCORE, with the highest scoring city ending up in the highest ranked sector in the 3x3 grid, ... , and with the lowest scoring city ending up in the lowest ranked sector in the 3x3 grid.
Click to expand...

Could this really be possible? If I understand you correctly, Rank #15 would share his grid with Ranks #3-5, 14, 16, 33-34, while Rank #3 would share his grid with Ranks #1-2, 4, 11-15, as he has another grid centered on his city.

I'll try to explain my concern with an example. Let's assume that Rank #3 turns out to have the highest City Score when the movement takes place. This would mean that he gets to be at the centre of the mountain, as Rank #1, as he has beaten the score of the previous Rank #1. However, in the grid of Rank #15, Rank #3 also played a role. As he obviously has the highest score in that grid, shouldn't he stay at the same place as he was?
Another example. Let's assume that Rank #3 removes a lot of buildings and turns out to have the lowest City Score when the movement takes place. He would then take the place of Rank #15, the lowest Rank in his grid. But what if the previous Rank #16 would also have a higher score? Wouldn't this affect the placing of the lowered Rank #3? Rank #3 and Rank #16 didn't share a grid before the movement, so in theory they wouldn't affect each other, but this would mean that the placement of the other Ranks in your grid don't have to make sense in terms of City Scores. A player with a low City Score could be placed more at the center of the World Map than a player with a higher City Score, just because he did have a high score earlier (and it takes a while before your city would be placed at the edge of the World Map).

Somehow I think it could work to let each Rank have his own grid, but I still need some more convincing. I know it would be difficult, but could you explain how the map would change after one movement? Perhaps you could start with the easy *unrealistic) scenario that the Ranks are the same as the rankings of City Scores and then change the City Scores a bit to show how the cities would move. :)
 

DeletedUser867

Guest
During the optimization, each Rank sector would be looked at 9 different times. The algorithm is a very local, one-pass process, and while there are some programming tricks that you could use to skip steps, it's likely that the optimization will run faster if the program simply goes through ALL of the motions each time, so that the code can be kept as clean as possible. For example, while there are several pretty algorithms for determining the adjacencies, the code would run much faster if you simply compiled a lookup table for the Rank Spiral and the adjacent ranks.
  • Start at Rank #1 and walk down the spiral one rank at at time
  • Decide if the current city needs to be archived or if a guest needs to be reinserted
  • Sort the 9 cites in the local 3x3 grid
  • Rinse and repeat
As Rank will be a physical location on the World Map, we've :D addressed :D most of the navigational problems.
Clicking on Rank would jump directly to that player's location on the World Map.

Rankings.jpg

Rankings.jpg

Let's look at a couple of scenarios.

#35 melted her credit card.
  • When 3,4,5 are calculated, #35 is still out of reach
  • When 14 is calculated, #35 is still out of reach
  • When 15 is calculated, #35 will jump to #3
  • When 16 is calculated, she's out of reach again, and will remain so for the rest of the optimization
  • Climbing UP the mountain is NOT easy, but scores don't climb very fast anyway
CreditCard-Trap480.png
CreditCard-Trap480.png

The next week (she's #3 now) she got her credit card bill, and she was HORRIFIED. So she wiped her city.
  • When 1 is calculated, #3 will drop to #9
  • When 2 is calculated, #9 will drop to #10
  • 3 through 8 she's out of reach
  • When 9 is calculated #10 will drop to #25
  • When 10 is calculated #25 will drop to #28
  • When 11 is calculated #28 will drop to #29
  • Rinse and Repeat
  • She'll be chased down the hill until she reaches the Rank that's appropriate for her dramatically reduced score,
    which now includes points only from solved encounters
  • Falling DOWN the mountain IS easy, because scores can drop really fast
But normally you won't see either of these extremes.
  • Half of the people in your neighborhood will improve a bit each week, so they'll move along the spiral pretty much as a group
  • The other half, who are just barely playing, will drift toward the fringes of the map,
    because everyone else will shuffle past them.
 
Last edited by a moderator:

DeletedUser43

Guest
I've forwarded this idea as an option to solve the issue of inactive neighbours. I've also made sure to name the consequences of implementing this idea, e.g. that players would leave a trail of discovered cities and provinces behind and that players would have to make strategic choices as regards to discovering and completing provinces, because it could be seen as both strategically interesting and plain irritating.

I'd like to underline the fact that, while this idea has been forwarded, it's no guarantee that it will be implemented. It's forwarded so developers would consider the idea as a possible solution for the inactive neighbourhoods. :)
 

DeletedUser

Guest
As I posted here as well:

This thread is rather uncannily complicated and technical for me; however, trying to understand the suggestion there, it seemed to me that that it would result in all the higher levels flocking together, with the lower levels at the far end of the map.

Although I do see the benefits for the higher level players, personally I think that as a consequence it could easily result in the higher end players speeding up even more (as they have all the necessary goods available close by), and thus rapidly enlarging the gap with the lower level (including newer) players, who will have no such access. I'm not sure if that's any good for the balance of the game; these two ends will never meet again then.

To be honest, I'm happy with release notes 0.16/the solution that were released on May 12th! ;)
 

DeletedUser

Guest
It isn't particularly hard to understand, Tintagel. What it is all about is this.
You have a certain group. The best player stands above and then clockwise down the lower ranked players follow. The player in the middle is the worst player.
 

DeletedUser867

Guest
Tintagel said:
It seemed to me that that it would result in all the higher levels flocking together, with the lower levels at the far end of the map.
Click to expand...
Folks would indeed work their way toward the center of the map, but the lowest ranked players would drift to the fringes of the map all the way around. So it's like a bullseye. You get the most points in the middle.

The reason WHY you want to cluster similar cities, is that they have common interests. I doesn't do either of us much good if I have too much Steel and my Neighbor has too much Magic Dust. I don't want any more steel, and he couldn't use Magic Dust if he had some. We all need trading, and polishing partner who have reciprocal needs.
 

DeletedUser

Guest
Alhough that in itself is true (having similar cities next to each other) - the current map is made in such a way that you're not depending on a single neighbour, but on a whole score of neighbours. As some will progress faster than others, there'll always be some around your own level - and at this moment we have but three goods levels, so the chance someone is at your level is pretty large! :)
Especially, now that inactive players are being replaced.

I think, that having players in your neighbourhood that are further, will show you what your own village may look like, thus encouraging you to keep playing. And that is what we need, players who stay and will progress along us. New ones will certainly catch up :)
 

DeletedUser

Guest
This proposal is well over my head with regard to dev tools and programming, lol. Your visual aids give me some idea of wut you're suggesting, though, and I think I like it :)
 
You must log in or register to reply here.
Top