darkSpyro

No More.

My second post will deal with the actual amount of users over time.

First up, the total amount of users by year.

http://goo.gl/slSwWX

Then, the total amount of users by month.


The amount of new users who joined by year.

http://goo.gl/ZA46rz

The amount of new users who joined by month

For my third post, I'll deal with the topic and post count by forum section.

This is a bar chart showing the total post count by forum section. Sections with subforums have been split up and included as their own section. i.e. the S&N bar accounts for only the posts in S&N itself, posts in forum games and roleplay have separate bars.


And here we have topic count by section.


Here's the change in post and topic count by section between yesterday and today at the same time both days.(A 24 hour period)

Here's my data:

Banned/Suspended/Unactivated accounts(incomplete, pls help)
Current gem counts for all active accounts
Monthly user totals.
Posts/Topics per section
Potential gem range by join date
Active users and their avatar status

Previous 2 topics
//forum.darkspyro.net/spyro/viewposts.php?topic=117671
//forum.darkspyro.net/spyro/viewposts.php?topic=118192

[btw think of these as histograms(which they are, they show frequencies of a discrete variable(gems)), I just couldn't get rid of the space between bars]

Alright, so I thought it would be interesting to see the gem count distribution at a finer scale that by Sparx Rank.

So, initially, I decided to break it down by 100's as seen here:
As you can tell, this simply won't work. The 0-99 column overshadows the next closest column by a factor of more than 4. The fine details of the rest of the graph is near impossible to see, and the bars on the right hand side don't show up at all.

To deal with this I made another histogram with the 0-99 and the 100-199 bars removed.

The first few bars are still pretty big but you can see all the bars a lot better here. Of course we can't leave the 0-99 and 100-199 bars out entirely so I broke them down by 10's and made a separate graph.



Something else I made that might be of interest. A histogram showing gem counts 0-30 individually.

good work my man, i can finally look back at all the banned users and reminisce

You know things are bad for Idle Chatter when the Non-Spyro section somehow seems to have more posts. <.<

This is quite impressive

Quote: spyro and sonic

Banned/Suspended/Unactivated accounts(incomplete, pls help)

I can do scf's column if you want~

Quote: AvatariDragon

This is quite impressive

Quote: spyro and sonic

Banned/Suspended/Unactivated accounts(incomplete, pls help)

I can do scf's column if you want~

Anything you can do will be appreciated.

Done~

I'm working on something else for this topic atm.

In the meantime, a fact, if you have at least 200 gems you already have more gems then ~74% of the users here. Not counting banned/suspended/inactivated accounts of course. This is due to the massive amount of accounts that are never used, or only ever used for a week or two.

This is cumulative from 2005 to now.
It seems more people join in January then any other month by quite a bit.

All the multiple posts!! For the glory of science

Needs more scientist


Meanwhile

Will this be on the test?

Any clue on which banned user has the most gems?

Quote: spyro and sonic

This is cumulative from 2005 to now.
It seems more people join in January then any other month by quite a bit.

I joined in September. Yay, I'm in the minority!

Quote: Mrmorrises

Any clue on which banned user has the most gems?

I want to say this one. Highest gem count of any banned user I've found so far.

Quote: spyro and sonic

Quote: Mrmorrises

Any clue on which banned user has the most gems?

I want to say this one. Highest gem count of any banned user I've found so far.

I thought it was her. These guys are up there too:

//forum.darkspyro.net/spyro/profile.php?user=1775
//forum.darkspyro.net/spyro/profile.php?user=1433

Quote: spyro and sonic

Quote: Mrmorrises

Any clue on which banned user has the most gems?

I want to say this one. Highest gem count of any banned user I've found so far.

Whatever does a user with that many gems do to get banned?

Quote: theuone

Quote: spyro and sonic

Quote: Mrmorrises

Any clue on which banned user has the most gems?

I want to say this one. Highest gem count of any banned user I've found so far.

Whatever does a user with that many gems do to get banned?

Gem count doesn't matter when it comes to users being banned. This user asked for it.

Quote: Mrmorrises

Quote: theuone

Quote: spyro and sonic

I want to say this one. Highest gem count of any banned user I've found so far.

Whatever does a user with that many gems do to get banned?

Gem count doesn't matter when it comes to users being banned. This user asked for it.

Yeah, that's probably the only way. My logic is that if you've been around for that long, then you should know how not to get banned.

Here, we have the maximum and minimum gem counts a user could have as of today ordered by the week they joined, assuming they logged in and posted every week.

The next thing i'm gonna do is a scatterplot of account age vs. gem count. I'm expecting a noticeable positive correlation but with the massive amount of never used or barely used accounts it might not be so. This could take me a while to do though so bear with me.

Is it possible to do one on Birthdays?

Quote: somePerson

Is it possible to do one on Birthdays?

Sure, luckily enough I have 18 pages of people volunteering that info.

By request, some stuff on birthdays. That topic only provided me with a sample size of 492 but it should be enough.

Here is the distribution of birthdays by month:


Also:

it would be cool if this was sticky'd

Quote: Pixilism

What was your method of getting your data? Like, specifically for the gems, did you go down all the users on the userlist and add them all up, along with counting which ones belong in their specific tiers? Or was there an easier way of doing it?

I went and copied the entire userlist table into excel, while having it ordered from highest to lowest gem count. Then, I copied the gem count column into its own sheet. Since they're already ordered, I can just select say, all the numbers in the column until I see a number less than 5000. Excel will automatically tell you how many cells have been selected and I can write that number down somewhere else. Then do the same with values between 3000 and 4999 and then gold sparx and so on.

Afterwards I also split it into separate columns for each rank to simplify things. I think that can be seen in one of the Sheets in my data post on this topic. Yeah, on the Math here tab here

This topic has made me feel stupid and inconvenient in a good way.

These statistics are supremely interesting. You have made incomparable job doing these and your authentic interest in doing this is a personality feature everybody should look up to. Thank you for doing these and feel free to use my dS stats in numbers as a part of your project (which I hope, someday, will be named) to give most authentic descriptions of this site.

- - -

Quote: spyro and sonic

- - -

Sprink on Skylanders, taking over everything.

Also, some, err, basic things that I just realized I haven't put anywhere yet.

• The mean gem count is 252.9
  
• The median gem count is 39.
  
• The mode gem count is 0.
  
• The standard deviation is 549.3
  
• The first quartile is 12.0
  
• The third quartile is 214.0
  
• The interquartile range is Q3-Q1=214-12=202
  

Going by the usual definition of outliers there would be about 1800 outliers in this data set so I'll just ignore those for this next part, the box plot. (Expand the image because it's ****ing huge)


What this means is simply, 25% of users have between 0-12 gems, 25% of users have between 13-39 gems(the first box), 25% of users have between 40-214 gems(the second box), and 25% of users have between 215-7946 gems.

As promised, a scatter plot of account age vs. gem count. (with bonus Spyro) Huge image, expand pls.

You probably notice that there are no data points above a sort of invisible line, running from bottom left to top right. This happens simply because for any given age there is a maximum number of gems an account can have.(ex. someone who joined last month cannot have a gem count of greater than say, 1000.) This invisible line represents that maximum value of gems as a function of account age.

Quote: Windumup

All the multiple posts!! For the glory of science

Needs more scientist

What? Are you tryyying to pull me out of the crowd?

What... the heck...?

But seriously now. I'm highly suprised somebody takes the time to figure all this data out.
I'm curious, what is the percentage of active users with an avatar? I know it sounds stupid but I'm curious.

Quote: BlackWing116

But seriously now. I'm highly suprised somebody takes the time to figure all this data out.
I'm curious, what is the percentage of active users with an avatar? I know it sounds stupid but I'm curious.

nah, that's not stupid. I think I'll just look through like the past week's worth of topics to get my info, in all the sections.

Quote: spyro and sonic

Quote: BlackWing116

But seriously now. I'm highly suprised somebody takes the time to figure all this data out.
I'm curious, what is the percentage of active users with an avatar? I know it sounds stupid but I'm curious.

nah, that's not stupid. I think I'll just look through like the past week's worth of topics to get my info, in all the sections.

OH wow, realy? YOu can seriously do that? That's amazing.

Alright, here we go!

Data is here.

I took a selection of 300 users who logged in and posted within the past week. This obviously does not include every user who's been active recently but I figure 300 is a good enough sample size to provide an accurate representation. Based on past predictions only about twice this amount of users is active any given week anyway.

Of these 300, 277 have an avatar and 23 do not. This means 92.33% of these active users have an avatar. 7.66% do not.


Out of those 277 with avatars, 33 are using preset avatars.

This was the last post in the first topic I made and so it probably went unseen so:

Some things I've discovered:

The last id number to say "no user exists with this id #" is
//forum.darkspyro.net/spyro/profile.php?user=1295
This id number is sandwiched between 2 accounts made on May 24, 2008, meaning it also was.
This is around the time the ability to post as a guest was discontinued(which all evidence points to as happening sometime in the first half of 2008). This leads me to believe guest accounts(Fodder) are somehow related to these "missing" users.

The first Crystallized account didn't show up until July 22, 2008.
//forum.darkspyro.net/spyro/profile.php?user=1607
The first Rhynoc account didn't show up until July 24, 2008.
//forum.darkspyro.net/spyro/profile.php?user=1631

This is simply due to the fact that the Rhynoc and Crystallized ranks did not exist in the prehistoric darkSpyro, as can be seen by their absence from the old help files.

I took a sample of 103 users that have been fairly active in S&N since the start of this year and came up with this.

Interesting. After the first one, Dark must have decided that users needed to be at least recently active in order to receive gem day gems.

Or does this imply something else? I can't be entirely certain...

Quote: Philomena

Interesting. After the first one, Dark must have decided that users needed to be at least recently active in order to receive gem day gems.

Or does this imply something else? I can't be entirely certain...

I'm not sure what it means tbh. And also some of the earlier accounts like this one have a good amount of gems without ever logging on or posting.


EDIT: I found an explanation.

According to dark's post in this topic, gems were not always given out how they are now.

Quote: dark52

...the current gems system is just based off how long you've been registered as well as how many posts you've made, moreso to do with posts.

So I assume early on your gems would accumulate just by having an account, even if you never did anything with it.

a lot of these crystallized users seem to be a person making three different accounts each time because they have no idea how to use an e-mail to validate their account

HOly mother SAS litterly took my Avatar thing serious


Okay, YOLO on it. what's the general Character Length of Signatures by lets say around 200 users?

Quote: BlackWing116

HOly mother SAS litterly took my Avatar thing serious


Okay, YOLO on it. what's the general Character Length of Signatures by lets say around 200 users?

How many signatures contain links and how many are Quote:

site quotes

?

Quote: Philomena

Quote: BlackWing116

HOly mother SAS litterly took my Avatar thing serious


Okay, YOLO on it. what's the general Character Length of Signatures by lets say around 200 users?

How many signatures contain links and how many are Quote:

site quotes

?

Alright, took a little while to get 200 names but I did it. Data here.

To answer the first question, the average signature length is 82 characters. For reference, eighty-two characters is about the length of this current sentence.

As for the second question, 34 signatures(17% of the total) contained links, and 10(5%) contained quotes.

I'll probably stick a chart of some sort and a bit more details in here when it's not half past midnight.