Neighbor Removal

[DeletedUser61]

There's a lot of market segment information out there, and while some of it is self-serving the trends are pretty consistent. There's a wealth of readable material at http://www.pwc.com/gx/en/industries/entertainment-media/outlook/segment-definitions.html

An important notion is that you don't have to "steal" customers. The magic of marketing is that you can grow the segment itself, if you have an attractive product. My view is that Elvenar is attempting to broaden the traditional city builder category, which has always been a backwater.

 

[DeletedUser2753]

WINYANDOR UPDATES

7/6/2016: 59944 total, 37344>0, 23844>100
7/7/2016: 58240 total, 35690>0, 23338>100, 19259>250, 15521>1000 & 12802 >2000
7/8/2016: 56891 total, 34299>0, 22895>100, 19085>250, 15466>1000 & 12808 >2000

I don't know if this is the right procedure for the forum, but would like people to know I am editing my previous post and it did not show in the recent post list at the right of the Main forum page. Should I have quoted it instead?

 

[DeletedUser61]

Two points:

1. If you click on the help link at the bottom right of any page, you'll be able to drill down to the BBCodes. We have extensive support for tabular data whereby your data will still look orderly even when the entries aren't the same width anymore. I'm using an Excel spreadsheet that includes columns for the BBCodes, but you may prefer to use http://www.teamopolis.com/tools/bbcode-table-generator.aspx
2. Creating a new forum post triggers the "New Post" page that many of use use as our primary link. The mods want folks to post once per session, in a given thread, but if you're reporting a daily status then sequential posts are just fine, because it IS new information.

 

[DeletedUser2753]

WINYANDOR Neighbor Removal UPDATES

• 7/6/2016: 59944 total, 37344>0, 23844>100
• 7/7/2016: 58240 total, 35690>0, 23338>100, 19259>250, 15521>1000 & 12802 >2000
• 7/8/2016: 56891 total, 34299>0, 22895>100, 19085>250, 15466>1000 & 12808 >2000
• 7/9/2016: 55408 total, 32836>0, 22466>100, 18918>250, 15420>1000 & 12803 >2000

My observations and conclusions:
Disclaimer: Admittedly this is just a start and so far a small sample. I have no idea when the started this purge or what % of the map the purge is on now, it could if all I know, be 95% complete now, but a new purge should be done every 30 days.

1. The Removal/Replacement with Gold Mine sweep continues
   
2. The safe harbor score is above but close to 250
3. There are some removals above that, may be those that have frozen Main Hall Upgrades.
4. The number of players advancing above 250, 1000 or 2000 is not increasing, at least faster then the purge.
5. They are purging about 1400 per day, ergo at this rate they should complete the purge in approx 10 days based on the # of low scoring active players remaining.
6. When the purge is done we will have approx 17,500 players >250 and not purged plus some below 250 that are active.
7. From My personal observation of my fully purged discovered area this percentage is approx 25% of my neighbors. The next problem is of that remaining 25% less then 25% of them are active, the majority of the others played up to scores above 5000 before going AWOL.
8. In conclusion it appears we have a retention rate of about 5% or less than 3000 active players in Winyandor, Hopefully the number of new adds, who will show as players with less then 100 but >0, will start to reverse some of the declining numbers.

I hope this is helpful to all the active players, especially those that have goldmines galore.

 

[DeletedUser61]

They are purging about 1400 per day,

I suspect this is a differential number.

Population Change per day = New Immigrants per day minus New Goldmines per day

Depending on exactly how you're handling the cities that have a score of zero (only 1% of them will be on the map) there may be a few additional uncounted removals.

I'm monitoring a radial on the US1:Arendyll server,
see https://us.forum.elvenar.com/index.php?threads/arendyll-neighborhood-watch.2973/#post-17804
For that small sample of 200 city sectors, I've seen 9 new Immigrants and 16 new Goldmines over the past two weeks.

• Before Goldmines were introduce the removal threshold had crept up above 100 days.
• The develops appear to be examining cities in alternate groups of approximately 300,
  which has the effect of putting the edge of the map in the middle of the cycle.
• The removals (Goldmines) are now occurring after 30+ days, but only half of the cities have been examined, so far as I can tell, and about half of the skipped cities are clearly below the threshold.

If you look at a few of your cities that have a score higher than 250, I suspect that you'll find a lot of level 1 and level 2 Residences and Workshop, rather than a few higher level buildings. I suspect that it's the average upgrade level, rather than the total score, that provides a safe harbor.

 

[DeletedUser2753]

I suspect this is a differential number.

I could not quote your quotes but if the total number of players has been reduced by 1483 and the number of players >0 has deceased by 1463 yes they could have purged more and added more but that is not easily determined. Certainly they could not have purged less that 1400/day with these results, unless they are removing more by other methods and adding more. So Yes. I agree it is a differential but the fact remains the same, a net loss of approx 1400+. Your experience in US1:Arendyll is vastly different then I see here.

Update: Ever since I posted this I have been bothered by agreeing with your quote. What is a differential number anyways? Not a mathematical term i am familiar with. My understanding is in mathematics, differential refers to infinitesimal differences or to the derivatives of functions.

 

[Valtitude]

I was away from Elvenar for a bit recently. Unfortunately, I was released from my Arendyll fellowship during that time - a simple matter of 'no notice' due to an off-game incorrect email address. The email bit's been resolved, but it played a part in how I felt when I returned to to the game.

Imagine the whammy that hit me when I returned to Aren that day! Suddenly, I didn't have a fellowship and found more gold mines than players in my neighborhood - not an exaggeration. I was pretty shocked at first. I'm aware I'm wa-ay out on the fringes of the Aren server - but, 'OUCH!' ... until I realized, clicking a gold mine takes a lot less time than entering a barely active city to 'help' an abandoned city. We call these abandoned cities 'ghost towns' in my area of the World. Usually the gold mines preceed the ghost towns in Real Life, though.

While I was 'drifting' without a fellowship, I really wondered how long I could hang on without slipping even further through the ranks than I already had while away. I'm not primarily driven by points and rank, but hey! I have my standards ... lol

Which brings up a second point ... congratulations, BobbyKitty for playing without diamonds on Aren!

I prefer the challenge of figuring things out to the easy diamonds 'fix'. I'll take this opportunity to tout two of my Elvenar Winyandor mentors/ teachers - azarra and Dhraga. Both of them have spent few, if any diamonds in the game. I think they each purchased an extra Builder, but not much more, if any, diamonds spent. Sorry Inno, I know that doesn't help your bottom line. Still, I almost feel I'm letting azarra and Dhraga down if I spend too many diamonds - ha!

They've taught me the basic structure of the game, strategy - even patience! and more. Thank you both very much, as you already know!

 

[DeletedUser2753]

WINYANDOR Neighbor Removal UPDATES

• 7/6/2016: 59944 total, 37344>0, 23844>100
• 7/7/2016: 58240 total, 35690>0, 23338>100, 19259>250, 15521>1000 & 12802 >2000
• 7/8/2016: 56891 total, 34299>0, 22895>100, 19085>250, 15466>1000 & 12808 >2000
• 7/9/2016: 55408 total, 32836>0, 22466>100, 18918>250, 15420>1000 & 12803 >2000
• 7/10/2016: 54049 total, 31459>0, 22034>100, 18731>250, 15374>1000 & 12805 >2000

My observations and conclusions:

1. See previous update post for conclusions
   
2. My conclusions remain the same because the trends remain the same with one exception pointed out by Katwijk. They are at least purging 1400 players per day but that number may be the difference between new players added and players removed. So The could be removing more, say 2400 and adding 1000
   
3. The result is similar a net loss but if they are adding players then the purge will complete sooner and the declining numbers should reverse themselves sooner. Time will tell.

 

[DeletedUser61]

differential refers to infinitesimal differences or to the derivatives of functions.

A differential is just a change of any sort, and if it's periodic and similar, then so much the better.

Leibniz and Newton developed the techniques that we now refer to as Calculus, but in the case of immigrants we're talking about a random arrival queue and that's an intractable problem as far as calculus is concerned. There are a ton of spline fitting techniques that will allow you to impose an order of sorts on the distribution, but the arrivals are inherently random rather than formulaic, so the more infinitesimal you get, the wronger you will be.

 

[DeletedUser2753]

Deciding to look closer at this I checked the same set of numbers a few minute ago. Previously i would record the numbers at approx 9 AM EST each day. This is 3 PM in Germany. Since as you said the arrivals would be a random arrival queue, it would be a good guess that new arrivals would be as players startup and be a traceable by the hour, with a wave based on demographics and time of day, whereas it would also be my guess that the purge would not be manual, but be some form of script or MySQL DELETE Query run once a day. Now I don't want to track these numbers hourly to graph this out but as i said, i just looked at 4:40 PM, approx 8 hours later, and the number of players is now 54325 and #>0 has increased to 31669, approx 210+ adds. Now we will have to look again at Midnight to see if this is showing a rate off approx 25/hour for new adds. I will go to sleep before midnight and who knows when the script may run. I will look later and report what I find. It is also possible the purge will not run. Still many variables but I think I am now closer to a view of what is happening.


Update 7/10 6:54 EST 54400 Total, 31731>0
.............. 7/10 8:18 -------54446 ------- 31766

Update 7/11 7:20 EST AM 52663 Total, 30089>0

It appears the purge did run last night.
Another quick thought: If the purge is designed to be a continuously run, i would design it to repeat every 30 days so 60,000/30 = 2000 would be my first guess as the best number of purges for the first run through. this would result in a loss of 1400 net if they anticipated 600 new adds per day, after clearing the backlog the number can be adjusted to equal the monthly adds x the 30 day attrition rate. Since the 4 day 100% purge of my own discovered area sample of 115 neighbors only resulted in a loss of about 80% to 90% of my neighbors, a full 2400 daily purge would result in a loss of approx. 2000, so the cycle may end in less than a 30 day first run. say the end of the month.

My current thought is new joins should peek in an hour or two for this US server and the purge should be about 4 AM EST considering a German sys-admin would not want to start the day on a bummer if something failed. so 10 AM German time would allow a cup of coffee and a scan of the system first. We will see.

 

[DeletedUser2753]

WINYANDOR Neighbor Removal UPDATES

• 7/6/2016: 59944 total, 37344>0, 23844>100
• 7/7/2016: 58240 total, 35690>0, 23338>100, 19259>250, 15521>1000 & 12802 >2000
• 7/8/2016: 56891 total, 34299>0, 22895>100, 19085>250, 15466>1000 & 12808 >2000
• 7/9/2016: 55408 total, 32836>0, 22466>100, 18918>250, 15420>1000 & 12803 >2000
• 7/10/2016: 54049 total, 31459>0, 22034>100, 18731>250, 15374>1000 & 12805 >2000
• 7/11/2016: 52698 total, 30116>0, 21663>100, 18575>250, 15324>1000 & 12803 >2000

3 HOUR UPDATES (Noon EST, 3 PM, 6 PM, 9 PM)

• 7/11/2016: 52765 total, 30174>0, 21682>100, 18586>250, 15337>1000 & 12806 >2000
• 7/11/2016: 52854 total, 30258>0, 21722>100, 18601>250, 15342>1000 & 12811 >2000
• 7/11/2016: 52932 total, 30323>0, 21761>100, 18625>250, 15358>1000 & 12816 >2000
• 7/11/2016: 53000 total, 30383>0, 21789>100, 18670>250, 15365>1000 & 12823>2000

My observations and conclusions:

1. The Purge continues at the same time once/day. Somewhere between 10 PM EST and 7 AM EST
2. The Purge less new joins equals approx 1400/day
3. As per previous post, the new joins look like approx 600/day, but closer hourly tracking for 24 hours is needed to determine the number. Update: It looks like 20-30/hour in the US morning/afternoon hours and in the evening showed 30/h (see previous post). During the 10 PM to 7 AM period I would believe we would see the lowest signups and joins. Ergo: 600 may be high for a Monday, looks more like 500 for a day
   
4. The 250 number is low for the average cutoff of the purge, but not much, it may be closer to 300. I wont change the tracking here though.
   
5. We are loosing approx 50 players/day with scores above 1000 but less at 2000/day. This cannot be upward migration unless 50 above 2000 are purged each day, which is possible, but remarkably consistent.
6. My best guess now is the purge is approx 1900/day with 500 joins/day and runs at 9AM German time
7. The a full cycle of the current Winyandor purge, should finish before the end of the month. 53000/2000= 26.5 days max. and in the past 5 days of the 53000 we should have gained maybe 3000+ new players and in all probability it started before it rolled through my area 10 days ago.
   
8. We await a significant change in the numbers to see a change in purges and joins. Only time will tell
9. I will continue of a non periodic basis to look at joins to refine my new guess of 400/day.
10. Everyone can track this by simply clicking on their score and browsing through the full player list

 

[DeletedUser2753]

I will enclose a chart of another game (name undisclosed) just to illustrate the repeated weekly wave of signups, I imagine the demographics are similar to ours, but Winyandor is only a small fraction of Innogames total daily signups.

Why bother with all this? This is a strategy game and neighbors and fellowships are a major part of the game, so understanding the playing field will allow us to make better choices.

 

[DeletedUser43]

How can we use this information to make better choices?

 

[DeletedUser2753]

How can we use this information to make better choices?

I have not finished my research so it is presumptive to give specifics but I will make an analogy:

Imagine you start playing Soccer on grass turf and then they changed the field to crumb-rubber synthetic turf and after playing on it, research showed a possible link to cancer in players, especially goalies. Now what would your choice be? Quit playing the game, play only on older grass fields, protest and require a change in the field or just live with it. Now that is a real world issue now presenting itself, and though soccer is also just a game, cancer is not. I am fairly sure changes here will not be of the same magnitude as me dying, but the choices remain the same.
ref: http://espn.go.com/espnw/news-commentary/article/14206717/how-safe-fields-where-play

I don't mean to sound glib or blow your question off since it is a good one that i myself am trying to answer. I belive we are about 1/3 through with this monthly purge and so 2/3 of the players have not be effected yet, so as time goes on the issues will start to be much clearer.

One thought is if we know the number of goldmines in world, the # of active players, the join rate, and 30 day retention rate we will see how long it will take to get back to a place were neighbors were a great source of coin. At first blush it appears that may never happen without more changes, so Fellowships gain additional importance, you need more Residences, and an Ancient Wonder like Blooming Trader becomes much more desirable. With a retention rate as high as 20% after 30 days for new joins that is about 100/day remaining. To replace the goldmines would take more that 1 year.

Also I have notices a huge difference in trades posted by undiscovered players. i now have 10 pages full of them versus just a few before. If the algorithm was set to the number of players not distance then the purge would increase the number of traders, since players exempt for the purge have a 4 or 5 X active rate then those replaced with goldmines, as new players are added the number of out of discovered area trades will decrease.

 

[DeletedUser2753]

WINYANDOR Neighbor Removal UPDATES

• 7/6/2016: 59944 total, 37344>0, 23844>100
• 7/7/2016: 58240 total, 35690>0, 23338>100, 19259>250, 15521>1000 & 12802 >2000
• 7/8/2016: 56891 total, 34299>0, 22895>100, 19085>250, 15466>1000 & 12808 >2000
• 7/9/2016: 55408 total, 32836>0, 22466>100, 18918>250, 15420>1000 & 12803 >2000
• 7/10/2016: 54049 total, 31459>0, 22034>100, 18731>250, 15374>1000 & 12805 >2000
• 7/11/2016: 52698 total, 30116>0, 21663>100, 18575>250, 15324>1000 & 12803 >2000
• 7/12/2016: 51147 total, 29874>0, 21682>100, 18611>250, 15375>1000 & 12836 >2000

8 hours later at 5 PM EST

• 7/12/2016: 51335 total, 30035>0, 21755>100, 18662>250, 15395>1000 & 12848 >2000

21 Hours later 6 AM EST (after morning purge)

• 7/13/2016: 49534 total, 29741>0, 21716>100, 18662>250, 15418>1000 & 12863 >2000

24 Hours later 9:20 AM EST

• 7/13/2016: 49558 total, 29760>0, 21728>100, 18669>250, 15423>1000 & 12869 >2000

28 Hours later 12:45 PM EST

• 7/13/2016: 49622 total, 29812>0, 21745>100, 18681>250, 17718>400 15429>1000 & 12876 >2000

1. It appears by visiting players will score at 400 they would be at about the cutoff of "Average of Level 3" but some below may have >200 diamonds and still be exempt. I will add >400
2. New Joins have stayed consistent at about 20-25/hour or 500 /day with a low in the morning US of 10/h and a recorded high of 30/h, but only 1/4 of those got on the map, ie: >0. If you visit the vast majority of players with a 0 score you will get a system error

My observations and conclusions:

1. I see a reversal in the trend for most numbers.
2. Though the total purge result remains the same, about 1500, the numbers above >0 have changed trend, it appears something changed with this last purge.
3. I did not witness a large uptick in joins yesterday but did stop looking a 9 PM EST so the purge was less effective on players with score >0 which has been regularly decreasing each day. See the 7/11 Update post for 12 hours of join tracking.
   
4. The number of players with scores>100 has now started to increase for the first time.
   
5. This all may be a blip or the purge is rolling through areas with older players, closer to the center of the map.
6. Only time will tell

 

[DeletedUser1122]

I find your stats and data interesting although I'm not clear if it will be useful in playing the game. Two points.
They have indicated the gold mines are only part of what they are planning to "fix the totally broken map".
The limited, 1/2 purges have not addressed scores above 10,000 and inactive fellowships or inactive fellowship members. [Thank god]
Out of the 220 provinces I discovered, within the space of 3 weeks, became a majority of gold mines and inactive players above 10,000.

 

[DeletedUser2396]

I also am wondering why the cap on Fellowships set at 25 - why not larger Fellowships allowed in game
I mean look just as example i have 14 Culture buildings atm in my City - 5 Builders - Main Hall
Culture click lasts 8 hours so in a day that takes 14x3=42 possible clicks from Fellowship members add 5 more clicks for Builder then there is also the Main Hall that could be quite a few clicks indeed to fill it up depending of course on the lvl of it
So why aren't Fellowship Cap set at like 50 or possibly 60 so your Culture is covered during a 24 hour period

 

[DeletedUser2753]

Culture click lasts 8 hours so in a day that takes 14x3=42 possible clicks from Fellowship members

Hi SilverKNite. Welcome back to your thread.

I am not sure and that is an interesting suggestion, but it may speed up the game to much for more advanced players. I say this because all day I am selling supplies to the wholesaler and coin for PMs. I currently with just 5-8 polishes at a time (13-15/day) run at 185% boost most of the day.
I don't know. Is there any boost cap say 200% or higher? Even so I could delete culture buildings with more polishes

 

[DeletedUser2753]

I find your stats and data interesting although I'm not clear if it will be useful in playing the game. Two points.
They have indicated the gold mines are only part of what they are planning to "fix the totally broken map".
The limited, 1/2 purges have not addressed scores above 10,000 and inactive fellowships or inactive fellowship members. [Thank god]
Out of the 220 provinces I discovered, within the space of 3 weeks, became a majority of gold mines and inactive players above 10,000.

• A few posts back I answered one of the bobbies how it could influence my choices.
• We can only speculate on the future changes that they may be planning, I can only work on the changes now occurring.
• I agree so far it is better for me to have 2o dead players in my discovered area then 20 more goldmines
• I have discovered the same # of provinces approx, and out of 114 players discovered 90 are now goldmines, of the remaining 24 players (all with scores > 5000 except one) only 5 or six are active. This is how we met SilverKNite in the trader for the first time.

 

[DeletedUser]

If there are any Fellowship openings on Arendyll for a daily active new player who has just entered the evolving era, let me know. I'd rather not stumble through making mistakes solo (but I am having fun with the game so far).