New legendary rank?

[N12]

Well it's lunchtime where I am now and I have just noticed this - so I guess I have become "a legend in my own lunchtime". 

Lol...I did it on call in the hospital, apparently. Cool idea...it'll help know everyone know who has been around for a while.

Too bad I don't get a half green and half yellow final coin :-)

Too bad indeed! I like both.

Also, lol @ "Legendary member". That would remind me of something else.

[1714801685]

1714801685

[1714801685]

1714801685

[1714801685]

1714801685

[dree12]

Well it's lunchtime where I am now and I have just noticed this - so I guess I have become "a legend in my own lunchtime".  

Lol...I did it on call in the hospital, apparently. Cool idea...it'll help know everyone know who has been around for a while.

Too bad I don't get a half green and half yellow final coin :-)

Too bad indeed!

Also, lol @ "Legendary member". That would remind me of something else.

And "Hero member" doesn't?

[oda.krell]

This means that in 2017 people who have registered on here just today have a potential of being "Legendary"? I'm not really a fan of emotes, but I gotta use this one...  

That's what I'm saying. It only takes about 2 years and 4 months for a brand new account to possibly become "Legendary". Now if you think of the meaning of the word, someone can become a legend after ~2 years? Really??

You could just do forum rank according to standard deviations and set legendary rank for accounts with activity above 3 standard deviations.  This would ensure that the highest rank is always highly exclusive.  

Bell curve distribution with no skew has its merits, too.

that is... beautiful...

3 SD above: legendary
2 SD above: hero
1 SD above: senior
average: member
1 SD below: candidate
2 SD below: junior
3 SD below: newbie

I think you'll find that the mean is much lower than you think due to the vast number of new accounts. Anything below the mean would likely be Newbie.

Not sure about that. Not anymore, at least. Looks to me like user registration numbers go through "growth spurts", similar to (and probably highly related with) price. Other than during those growth spurts, registrations seem to stagnate. Lately, I'm under the impression that except for a few newbie troll accounts, I'm only having conversations with seniors and above. Then again, I can only speak for the speculation sub, and I don't have any hard numbers on this, only a feeling. The solution is to look at a site-wide average of activity over time, no? If it keeps falling, you're right.

[dree12]

This means that in 2017 people who have registered on here just today have a potential of being "Legendary"? I'm not really a fan of emotes, but I gotta use this one...  

That's what I'm saying. It only takes about 2 years and 4 months for a brand new account to possibly become "Legendary". Now if you think of the meaning of the word, someone can become a legend after ~2 years? Really??

You could just do forum rank according to standard deviations and set legendary rank for accounts with activity above 3 standard deviations.  This would ensure that the highest rank is always highly exclusive.  

Bell curve distribution with no skew has its merits, too.

that is... beautiful...

3 SD above: legendary
2 SD above: hero
1 SD above: senior
average: member
1 SD below: candidate
2 SD below: junior
3 SD below: newbie

I think you'll find that the mean is much lower than you think due to the vast number of new accounts. Anything below the mean would likely be Newbie.

Not sure about that. Not anymore, at least. Looks to me like user registration numbers go through "growth spurts", similar to (and probably highly related with) price. Other than during those growth spurts, registrations seem to stagnate. Lately, I'm under the impression that except for a few newbie troll accounts, I'm only having conversations with seniors and above. Then again, I can only speak for the speculation sub, and I don't have any hard numbers on this, only a feeling. The solution is to look at a site-wide average of activity over time, no? If it keeps falling, you're right.

According to the stats page, there are 350419 members and 8293716 posts, meaning an average post count of 24. Since activity is always lower than post count, that means the mean user is a Newbie-level.

[oda.krell]

According to the stats page, there are 350419 members and 8293716 posts, meaning an average post count of 24. Since activity is always lower than post count, that means the mean user is a Newbie-level.

Jup, that answers the question. Didn't know we had site stats like that.

[the joint]

This means that in 2017 people who have registered on here just today have a potential of being "Legendary"? I'm not really a fan of emotes, but I gotta use this one...  

That's what I'm saying. It only takes about 2 years and 4 months for a brand new account to possibly become "Legendary". Now if you think of the meaning of the word, someone can become a legend after ~2 years? Really??

You could just do forum rank according to standard deviations and set legendary rank for accounts with activity above 3 standard deviations.  This would ensure that the highest rank is always highly exclusive.  

Bell curve distribution with no skew has its merits, too.

that is... beautiful...

3 SD above: legendary
2 SD above: hero
1 SD above: senior
average: member
1 SD below: candidate
2 SD below: junior
3 SD below: newbie

I think you'll find that the mean is much lower than you think due to the vast number of new accounts. Anything below the mean would likely be Newbie.

Not sure about that. Not anymore, at least. Looks to me like user registration numbers go through "growth spurts", similar to (and probably highly related with) price. Other than during those growth spurts, registrations seem to stagnate. Lately, I'm under the impression that except for a few newbie troll accounts, I'm only having conversations with seniors and above. Then again, I can only speak for the speculation sub, and I don't have any hard numbers on this, only a feeling. The solution is to look at a site-wide average of activity over time, no? If it keeps falling, you're right.

According to the stats page, there are 350419 members and 8293716 posts, meaning an average post count of 24. Since activity is always lower than post count, that means the mean user is a Newbie-level.

There are many reasonable ways you can account for the skewed numerical distribution. 

[Brewins]

This means that in 2017 people who have registered on here just today have a potential of being "Legendary"? I'm not really a fan of emotes, but I gotta use this one...  

That's what I'm saying. It only takes about 2 years and 4 months for a brand new account to possibly become "Legendary". Now if you think of the meaning of the word, someone can become a legend after ~2 years? Really??

You could just do forum rank according to standard deviations and set legendary rank for accounts with activity above 3 standard deviations.  This would ensure that the highest rank is always highly exclusive.  

Bell curve distribution with no skew has its merits, too.

that is... beautiful...

3 SD above: legendary
2 SD above: hero
1 SD above: senior
average: member
1 SD below: candidate
2 SD below: junior
3 SD below: newbie

Do it and see the number of automated registration skyrocket only to lower the requisites to reach member status.


Add other requisites(minimum post or activity to count into the statistics, etc), and you will see bot accounts all around trying o manipulate the system.

It is a nice idea, but I dobt it would work, and I'm sure it would be a pain for the mods.

[Hfleer]

This means that in 2017 people who have registered on here just today have a potential of being "Legendary"? I'm not really a fan of emotes, but I gotta use this one... 

That's what I'm saying. It only takes about 2 years and 4 months for a brand new account to possibly become "Legendary". Now if you think of the meaning of the word, someone can become a legend after ~2 years? Really??

You could just do forum rank according to standard deviations and set legendary rank for accounts with activity above 3 standard deviations.  This would ensure that the highest rank is always highly exclusive. 

Bell curve distribution with no skew has its merits, too.

that is... beautiful...

3 SD above: legendary
2 SD above: hero
1 SD above: senior
average: member
1 SD below: candidate
2 SD below: junior
3 SD below: newbie

Do it and see the number of automated registration skyrocket only to lower the requisites to reach member status.


Add other requisites(minimum post or activity to count into the statistics, etc), and you will see bot accounts all around trying o manipulate the system.

It is a nice idea, but I dobt it would work, and I'm sure it would be a pain for the mods.

Also there are 350000 members total and only 217 legendary. ´3 SD will be wayyyy more

Edit: Well, I was wrong, 3 SD would only be 455 legendaries.

[the joint]

This means that in 2017 people who have registered on here just today have a potential of being "Legendary"? I'm not really a fan of emotes, but I gotta use this one...  

That's what I'm saying. It only takes about 2 years and 4 months for a brand new account to possibly become "Legendary". Now if you think of the meaning of the word, someone can become a legend after ~2 years? Really??

You could just do forum rank according to standard deviations and set legendary rank for accounts with activity above 3 standard deviations.  This would ensure that the highest rank is always highly exclusive.  

Bell curve distribution with no skew has its merits, too.

that is... beautiful...

3 SD above: legendary
2 SD above: hero
1 SD above: senior
average: member
1 SD below: candidate
2 SD below: junior
3 SD below: newbie

Do it and see the number of automated registration skyrocket only to lower the requisites to reach member status.


Add other requisites(minimum post or activity to count into the statistics, etc), and you will see bot accounts all around trying o manipulate the system.

It is a nice idea, but I dobt it would work, and I'm sure it would be a pain for the mods.

The standard deviation idea is incomplete in and of itself, but there are a variety of fairly simple ways that you can mitigate this issue.  For example, a very simple idea would be to set an invariable range for newbie status with an activity threshold set high enough to effectively rule out all irrelevant accounts from the data.  In other words, newbies wouldn't even factor into the mean.  I'd expect a fairly normal distribution to emerge if you were to exclude all accounts with an activity level below, say 50-100.

[tspacepilot]

This means that in 2017 people who have registered on here just today have a potential of being "Legendary"? I'm not really a fan of emotes, but I gotta use this one...  

That's what I'm saying. It only takes about 2 years and 4 months for a brand new account to possibly become "Legendary". Now if you think of the meaning of the word, someone can become a legend after ~2 years? Really??

You could just do forum rank according to standard deviations and set legendary rank for accounts with activity above 3 standard deviations.  This would ensure that the highest rank is always highly exclusive.  

Bell curve distribution with no skew has its merits, too.

that is... beautiful...

3 SD above: legendary
2 SD above: hero
1 SD above: senior
average: member
1 SD below: candidate
2 SD below: junior
3 SD below: newbie

Do it and see the number of automated registration skyrocket only to lower the requisites to reach member status.


Add other requisites(minimum post or activity to count into the statistics, etc), and you will see bot accounts all around trying o manipulate the system.

It is a nice idea, but I dobt it would work, and I'm sure it would be a pain for the mods.

The standard deviation idea is incomplete in and of itself, but there are a variety of fairly simple ways that you can mitigate this issue.  For example, a very simple idea would be to set an invariable range for newbie status with an activity threshold set high enough to effectively rule out all irrelevant accounts from the data.  In other words, newbies wouldn't even factor into the mean.  I'd expect a fairly normal distribution to emerge if you were to exclude all accounts with an activity level below, say 50-100.

Is it possible to download a list of accounts and activity levels in order to play with these stats? 

I also think it'd be kinda cool to do some kind of rank statistic which deals with the fact that we don't want to have to add a new level every 2 years.

[jaberwock]

This means that in 2017 people who have registered on here just today have a potential of being "Legendary"? I'm not really a fan of emotes, but I gotta use this one...  

That's what I'm saying. It only takes about 2 years and 4 months for a brand new account to possibly become "Legendary". Now if you think of the meaning of the word, someone can become a legend after ~2 years? Really??

You could just do forum rank according to standard deviations and set legendary rank for accounts with activity above 3 standard deviations.  This would ensure that the highest rank is always highly exclusive.  

Bell curve distribution with no skew has its merits, too.

that is... beautiful...

3 SD above: legendary
2 SD above: hero
1 SD above: senior
average: member
1 SD below: candidate
2 SD below: junior
3 SD below: newbie

Do it and see the number of automated registration skyrocket only to lower the requisites to reach member status.


Add other requisites(minimum post or activity to count into the statistics, etc), and you will see bot accounts all around trying o manipulate the system.

It is a nice idea, but I dobt it would work, and I'm sure it would be a pain for the mods.

The standard deviation idea is incomplete in and of itself, but there are a variety of fairly simple ways that you can mitigate this issue.  For example, a very simple idea would be to set an invariable range for newbie status with an activity threshold set high enough to effectively rule out all irrelevant accounts from the data.  In other words, newbies wouldn't even factor into the mean.  I'd expect a fairly normal distribution to emerge if you were to exclude all accounts with an activity level below, say 50-100.

Then people will start to mass farm accounts only to drive the requeriments for higher ranks lower. You supposing people will be honest and don't try to manipulate the system, what might not be true

[tspacepilot]

If we have access to the numbers then we could discuss more specifically what such a scheme would look like.  I don't think we have to assume that people are honest (we should probably assume the opposite) that doesn't mean we can't consider other types of systems.  If, for example, we made a statistical system which took skewness into account then we could probably counteract any mass-account-farming action meant to tip the distribution.  I'm just saying that the idea needs more details filled in, without the details it's hard to argue the merits.

[oda.krell]

Then people will start to mass farm accounts only to drive the requeriments for higher ranks lower. You supposing people will be honest and don't try to manipulate the system, what might not be true

Not sure where you get the idea from that this will be so easy. I've seen a few accounts that very obviously are linked to a posting bot. Those posts seem to be reported (and deleted) rather quickly usually.

And if the automated approach should work, then what prevents them from using it to "farm" for high rank accounts directly (in a fixed requirement rank system).

All that said, the deviation defined rank idea looks very nice to me and has a certain nerdy appeal, but in reality, it's not really worth the effort probably. Introducing a new rank each 2 years with about twice the activity requirement is a hack, but an acceptable one imo.

[C.Steven]

Then people will start to mass farm accounts only to drive the requeriments for higher ranks lower. You supposing people will be honest and don't try to manipulate the system, what might not be true

Not sure where you get the idea from that this will be so easy. I've seen a few accounts that very obviously are linked to a posting bot. Those posts seem to be reported (and deleted) rather quickly usually.

And if the automated approach should work, then what prevents them from using it to "farm" for high rank accounts directly (in a fixed requirement rank system).

All that said, the deviation defined rank idea looks very nice to me and has a certain nerdy appeal, but in reality, it's not really worth the effort probably. Introducing a new rank each 2 years with about twice the activity requirement is a hack, but an acceptable one imo.

The current activity required for "level up" are 30, 60, 120, 240, 480, ~900 (middle of 775 and 1030).
So, maybe we will need a new rank 2 years later but then it will be 5 years and 10 years, as the activity required for the new level will be doubled every time.

[freedomno1]

The current activity required for "level up" are 30, 60, 120, 240, 480, ~900 (middle of 775 and 1030).
So, maybe we will need a new rank 2 years later but then it will be 5 years and 10 years, as the activity required for the new level will be doubled every time.

Another way of putting it is that Bitcoin will halve a few times before the next doubling in the not to long future
Kind of a reverse way of looking at bitcoin as the coin circulation halves
The activity to the next rank on the forum doubles.

[Omikifuse]

Then people will start to mass farm accounts only to drive the requeriments for higher ranks lower. You supposing people will be honest and don't try to manipulate the system, what might not be true

Not sure where you get the idea from that this will be so easy. I've seen a few accounts that very obviously are linked to a posting bot. Those posts seem to be reported (and deleted) rather quickly usually.

And if the automated approach should work, then what prevents them from using it to "farm" for high rank accounts directly (in a fixed requirement rank system).

All that said, the deviation defined rank idea looks very nice to me and has a certain nerdy appeal, but in reality, it's not really worth the effort probably. Introducing a new rank each 2 years with about twice the activity requirement is a hack, but an acceptable one imo.

But with the variance stuff, the requeriments for ranks will change every time, so many will be in one position, and next hour in another, then back the the starting one, and so for.

Will generate lots of complains and will be a pain to signature campaigns.

[Chris_Sabian]

Legendary member rank is a good idea.  I think I like it.  ALthough I have a long way to go to make it 

[Koko]

What's going to be the next next rank after Legendary? Godlike?

[Vod]

What's going to be the next next rank after Legendary? Godlike?

Vodlike.

[tspacepilot]

According to the stats page, there are 350419 members and 8293716 posts, meaning an average post count of 24. Since activity is always lower than post count, that means the mean user is a Newbie-level.

Thanks for the link to the stats page,  I've been looking for that!