So I think we have at most 999,000 accounts since only ~1/10000 possible ID strings are actually used b/c the first two nd last two digits are scrapped each time.

Knyte, here is a jo*hark* I mean a task for you :) Sort clans based on when its members joined warlight. Not going to yield anything super useful but still a fun factor. Pretty Please.

@TBest: done (http://pastebin.com/nxBVpJuR)

Also, that is a great way to use that information @DanWL. I might just end up tracing WL growth that way from beginning until the end.


@Ben, Well I neither feel that the player base have grown soo much... but then I think about the huge increase in number of trolls on the forum. Now that has changed a lot.

You wouldn't even need to do a lot of scraping to get a decent graph of this, since it would be an XY plot. Just sample a bunch of players (even just from those posting in the forums would be a good start), find each (ID, Join date) pair, and plot it. If your hypothesis is correct, then the result should be a strictly monotonic graph, and it would give a lot of visual information about WL growth, including spurts and trends.

I bet even as low as 50 (relatively random) samples would still give a good rough graph. Certainly knyte's existing data contains enough information already to produce a very good graph, esp. considering there are clans containing very old players and ones containing very new players.

Would underestimate player growth in the recent past and overestimate growth in the far past because people don't join clans immediately.

hey kynte serjestions on how to tell weather they are alts alts check wich ip they use

the only problem is if they use internet cafes

Interesting data there knyte - turns out 13 of REGLs 66 members (20% of us) joined warlight in the month of October 2011.

@muddles: That info isn't exactly publicly accessible.

ok

Edited 12/25/2015 16:30:06

"275549 accounts being made in a total of 602 days seem well to high, in comparison with 55560 being made in a total of 178 days."

(55560/178)*602=187905
275549 =146,6% of 187905.
So in the period of 07-02-14 till 02-10-15 are 46,6% more accouns made the betwean 13-08-13 till 07-02-14

An increase of 46,6% is what I would expect in a period of around 2 years.

estimated 600-700 romanian players
550 known by me
50 active ...

Sneak Peak!



NOTE: This is based on VERY few data points (31), and is inaccurate.

Interested in helping me and DanWL expand it to some detail? You are most certently welcome to help. Just PM me :)

Edited 12/25/2015 21:44:28

my player ID is 711358681, joined 19 VI 2010. How many players joined before or after me? :p

my player ID is 711358681

No, it isn't, but it is if you remove the first and last 2 numbers. So it would be 13586.

I'm prioritizing map creators and the first few people that made an account, ATM.

Edited 12/25/2015 22:12:11

i fucking love warlight. also fizzer doesnt want you to know... just buy a membership and stay docile, which means he has created a matrix and you have all taken the blue pill

MH players plotted into the WL growth chart! Want your clan done? PM me and I will set you up!

Scraped data from the 6,892 players that were visible (i.e., in a clan or having participated in a ladder) as of five days ago. Some quirks and flaws (newest batch of players clearly doesn't have much representation) but this is what I got by taking their true IDs and join dates.

Some weird spikes, too. I took the highest true ID value for each day and set it as the day's corresponding value. Not sure what's going on or whether it's on my end or WarLight's.

Also, looks like the October 2011 climb is visible here as well. Looks like it happened around 10/10/2011.

Edited 12/26/2015 02:00:49

Would underestimate player growth in the recent past and overestimate growth in the far past because people don't join clans immediately.

Wait, is 'join date' the date they joined the *clan*, or the date they joined *Warlight*?

If the latter, which is what I was thinking, then no it would not overestimate. It's an XY plot, and it's monotonic (in theory*), so each point stands on its own. You know their exact Warlight join date and their user ID. You just plot them out.

If the former, then I don't get the point of tracking that.

[* Based on your last comment/chart, looks like either the theory is broke, or somehow the data got messed up.]

Edited 12/26/2015 04:21:19

The join date is WL, join date.

And yes it is a XY plot. However ID's do not always correspond to join date. (FOr example Players that were invited, by email. Once the invite was sent their playerID was assigned to that email, but there Join date does not occur before they actually joined WL.)

Basically, a lot* of the accounts only exist in the system.

*that is an unknow X players that don't exist.

Edited 12/26/2015 04:29:57

The invite thing would explain points where the ID is low and join date is high (downward spikes). Any ideas on what could be causing upward spikes, where people 'join early' but get a 'high ID'?

[Edit: One theoretical possibility just popped into mind. Are there any cases where people had old accounts but for whatever reason were given new accounts and the old 'join date' is imported into the new account? Like if someone lost their password, asked Fizzer to give them a new account, or something along those lines?]

Edited 12/26/2015 04:48:07