Skip to main content

Topic: Discussion: Population spread and server location factors (Read 19758 times) previous topic - next topic

Re: Discussion: Population spread and server location factors
Reply #30
@Ub3rgames

Here is some data for you just so you can see the ping difference for the average NA west coast player. Here is the site and data I used if you want to check for your self.

https://wondernetwork.com/pings

London to Los Angeles - 130.031ms
London to Portland      - 127.15ms
London to Seattle        - 133.455ms

New York to Los Angeles - 65.694ms
New York to Portland      - 62.326ms
New York to Seattle        - 64.603ms

These are not numbers for average players. These are tier 1 network provider, aka backbone latency numbers, optimized for single provider. It will never be that good for your average Joe. I have to add additional 30ms for the New York route and 50 more for London.
  • Last Edit: January 23, 2017, 01:40:41 am by nietoperz

  • MSWG
  • [*]
Re: Discussion: Population spread and server location factors
Reply #31
@Ub3rgames

Here is some data for you just so you can see the ping difference for the average NA west coast player. Here is the site and data I used if you want to check for your self.

https://wondernetwork.com/pings

London to Los Angeles - 130.031ms
London to Portland      - 127.15ms
London to Seattle        - 133.455ms

New York to Los Angeles - 65.694ms
New York to Portland      - 62.326ms
New York to Seattle        - 64.603ms

These are not numbers for average players. These are tier 1 network provider, aka backbone latency numbers, optimized for single provider. It will never be that good for your average Joe. I have to add additional 30ms for the New York route and 50 more for London.

I agree which puts a London server around 180-200 ping for the typical west coast NA player and around 90-100 for the New York server. So moving the server to London does nothing for west coast NA. Again I will say New York is the best option for EVERYONE.

Re: Discussion: Population spread and server location factors
Reply #32
....NY would make the most sense...

Not really, most people don't know that it is more than just distance.

Many people forget that East Coast Silicon Valley is NOT New York but Virginia!  Home of AWS!

In 2013, as much as 70% of the world's Internet traffic traveled through data centers in Loudoun County!

Using above posters method and say server is hosted in https://wondernetwork.com/pings/Wilmington Delaware.

Berlin6591km94.61ms
Buenos Aires8454km149.66ms
Cairo9191km137.32ms
Chicago1045km26.97ms
Istanbul8240km141.64ms
Moscow7675km128.12ms
Paris6006km84.78ms
LA3814km73.78ms
Sydney15854km233.48ms
Toronto540km65.13ms
Vancouver3850km107.97ms

  • SomeBK
  • [*][*][*][*][*]
Re: Discussion: Population spread and server location factors
Reply #33
Looks like the server would be nice in Delaware since it looks like a relatively center mass area

Re: Discussion: Population spread and server location factors
Reply #34
http://www.submarinecablemap.com/#/
This is probably more important than east coast silicon valley and the facebook owned submarine cable to connects Europe with it.

Re: Discussion: Population spread and server location factors
Reply #35
I'm obviously bias, but I really hope for a London server, if it does move to NA, I unfortunately won't be playing, I'd rather not play at all than play with high ping, call me selfish, I don't care it's just not appealing with high ping and I commend any Americans who do play here.

Ideally the game will be so popular two servers won't be a stretch and we can all play in our respective timezones.

What is a high ping? If you're in the UK, you should have around a 90-95ms ping to NYC server on a typical cable/dsl modem, which is about the same ping as someone from Los Angeles or Seattle to NYC.  Ping www.columbia.edu (in NYC) for a benchmark.


Re: Discussion: Population spread and server location factors
Reply #36
@lambdaExpression
A private battleping could be done, but wouldn't be as good as a company dedicated at doing it.
Their tunneling is already proven to improve the experience by a large margin for people having bad routing or packet losses issues.
This helps NA users a lot, and we would recommend west coast users to try it out, but EU connections are good enough that it won't impact much.

@Natris
We have not given up on distributing the server worldwide, however we realized that we need a lot more work on the game itself before doing any such fancy research and development. These projects are postponed to after the creation of features consolidating gameplay and software stability.

@Moritori
Paying customers mean, for now, anyone that ever bought the game.
What we will look at is how ongoing sales are maintained as new features get implemented and marketing campaigns get launched. Activity will be looked at too of course, but it shouldn't diverge too much.
A London server would cost 2 to 3 times more.

The Darkfall: New Dawn Dev Team.

Re: Discussion: Population spread and server location factors
Reply #37
...
- 55% of purchases have been made from North America, with USA representing 48.2% and Canada representing 6.8%.
..
- While we had great hopes for Eastern Europe, Russia only represents 3% of our paying player base, Poland 2% and other countries being negligible.
...


In my view, this is the most definitive information.  What the London vs New York trade off really involves is whether NA player base outside the NE Quadrant or the Eastern Europe/Russia/Turkey player base is given more weight in the ultimate location decision.

If 55% of sales are already from NA, even if its sub-majority is in Eastern states/provinces, it still totally dominates the E. Europe interest in the game.

I would like to add some other factors to consider:
The Latin American market is probably completely avoiding DnD due to ping but had a notable population in DFO NA, particularly from Brazil. 

As DnD begins to in a much more tangible way differentiate itself from RoA and DFO, even more NA interest is likely to appear.  Up to now it's still the case that both games play a lot more like DFO than not and thus NA players with general DF interest have still been biased toward RoA due to ping alone. Ub3rgames will be able to observe if these trends in fact develop in the coming few months as patches begin to make striking differences.
  • Last Edit: January 23, 2017, 02:44:23 am by Crisply Vague

Re: Discussion: Population spread and server location factors
Reply #38
UK seems to be a decent middle ground. It's excessive to move the server across the atlantic just to appease the fairies over in California.
ᛟ Charax Hatehunter ᛟ

Re: Discussion: Population spread and server location factors
Reply #39
@Ub3rgames

Here is some data for you just so you can see the ping difference for the average NA west coast player. Here is the site and data I used if you want to check for your self.

https://wondernetwork.com/pings

London to Los Angeles - 130.031ms
London to Portland      - 127.15ms
London to Seattle        - 133.455ms

New York to Los Angeles - 65.694ms
New York to Portland      - 62.326ms
New York to Seattle        - 64.603ms

These are not numbers for average players. These are tier 1 network provider, aka backbone latency numbers, optimized for single provider. It will never be that good for your average Joe. I have to add additional 30ms for the New York route and 50 more for London.

One needs to add 10ms-20ms for dsl/cable latency to any of these numbers.  I made a matrix from the website that gives a good representation.  When I entered my city, the numbers were very accurate (+/- 2ms) to these numbers as I'm on a fiber optic connection, thus no bonded channel latency like dsl/cable typically have.


Re: Discussion: Population spread and server location factors
Reply #40
UK seems to be a decent middle ground. It's excessive to move the server across the atlantic just to appease the fairies over in California.

Unless they're paying fairies :)

Re: Discussion: Population spread and server location factors
Reply #41
Can we just assume everyone is picking the server location closest to them and skip all the post merely demonstrating this point, we really get everyone wants the server closest to them. really we do.
Mycke Soulslayer - Supreme General: Rastalfarian - An all Alfar clan of all play styles in all time zones, welcoming new and vet players alike.

Re: Discussion: Population spread and server location factors
Reply #42
I wont be buying if it goes closer to NA and EU ping gets screwed.

I'll just wait for Unholy Wars return which no doubt it will be some time soon

WAR BRINGER

Re: Discussion: Population spread and server location factors
Reply #43
I wont be buying if it goes closer to NA and EU ping gets screwed.

I'll just wait for Unholy Wars return which no doubt it will be some time soon

That makes 3 of you playing damn there might even be 10 people on that server by the time its launched, but at least you will all have good ping and prettier graphics to go with the half content.
Mycke Soulslayer - Supreme General: Rastalfarian - An all Alfar clan of all play styles in all time zones, welcoming new and vet players alike.

Re: Discussion: Population spread and server location factors
Reply #44
Can we just assume everyone is picking the server location closest to them and skip all the post merely demonstrating this point, we really get everyone wants the server closest to them. really we do.

No - there actually is data to analyze from a disinterested perspective.  People who make insistent or protesting appeals without an argument should be ignored.

If Ub3rgames original post said that 75%+ of paid players so far are from Europe, I would write that the live server probably should be kept in Europe, even though that'd mean +60ms more latency for me personally.