Realm Pop: More accurate than WoW Census could ever hope to be

[Elixor]

A very interesting tool. I´ll surely use it if I plan to pay for a month. Start fresh on the highest & balanced realm there is. Unfortunatly I already have on the "perfect" realm, characters .

Anyway, this seems quite nice.

[Kevyne-Shandris]

It doesn't work like that. Try reading; one more time.

The point is to be representative of a server's activity, which is investigated in a social sense. Nobody cares how many lvl 1 unguilded Human Preistesses are rolled on Moon Guard.

Read this: they care if it's level 29s; level 64s; and level 72s.

Why do they care? Because they would like to get into groups, and do content together. That the realm isn't dying, too.

A census that shows 3k 85s, but 200 level 58s (hmmmm, DKs), doesn't show a realm that is friendly for leveling. So many are indeed rerolling toons (if for anything out of sheer boredom), that knowing others are too, helps.

The census is often used to measure what's a "dead" server. If you looked at the census data for Shandris some might think so, but there's more folks there if you've been around to notice...that missing data.

BTW, I'm not "thread crapping", when you simply disagree. Promoting a new product is one thing, but like all products they can and will be reviewed.

[Imbasawrus]

I'm sorry but there is just no way in hell that skullcrusher EU has 12k alliance chars on it.

[Deficineiron]

I'm sorry but there is just no way in hell that skullcrusher EU has 12k alliance chars on it.

edit - looked up warcraftrealms numbers for that server, (3200 TOTAL), and I can at least see reason for skepticism.

so out of curiosity, went to realm pop, narrowed skullcrusher data to level 42's (they say about 280 of them), and clicked on 10 at random

8 existed, 4 had january updates, 2 of those yesterday. 2 had december updates, 1 had a september update, and one some other time.
2 no longer existed 'OOPS' screen. the toon LEVEL was correct, regardless.

another 10-toon sample showed 3 december updates, 1 september, 1 october, 3 current (last week or so), and 2 OOPS.

as I understand armory updates, toons 'update' even when inactive IF the parent account is active, and will show inactive for 6 months until 'oops' pops up. If I am wrong on this, please correct me. [edit - I think I am wrong, it should be last log-out date, but there are cases where it doesn't properly update after logging. the more I read the more confusing it is.] The % of toons with late december update dates is quite high.

I am revising my previous opinion, the realm population data includes characters that have been out of the game for a long time, as I understand it.

Now if they make their data sortable based on armory update date, that would be another matter entirely.

[Relentless]

Read this: they care if it's level 29s; level 64s; and level 72s.

Why do they care? Because they would like to get into groups, and do content together. That the realm isn't dying, too.

A census that shows 3k 85s, but 200 level 58s (hmmmm, DKs), doesn't show a realm that is friendly for leveling. So many are indeed rerolling toons (if for anything out of sheer boredom), that knowing others are too, helps.

The census is often used to measure what's a "dead" server. If you looked at the census data for Shandris some might think so, but there's more folks there if you've been around to notice...that missing data.

BTW, I'm not "thread crapping", when you simply disagree. Promoting a new product is one thing, but like all products they can and will be reviewed.

Yes, and overwhelmingly, by the point those characters reach the levels you've mentioned, they've either used the AH or been in a guild with someone who has, therefore making the population you're attempting to describe statistically irrelevant. You can say that you dislike the layout or that you're interested in seeing the 3 unguilded level 29 people who never use the AH that this missed, but the site is still useful, despite your weak arguments to the contrary.

[Appletini]

It's including characters that have been inactive for a long time in the count. Therefore, I think it should include stratification by activity since particular dates. Should be in the armory API somewhere, right?

[Ghostie]

I don't see how this can be accurate considering the number of alts everyone plays. Am I missing something?

[Deficineiron]

It's including characters that have been inactive for a long time in the count. Therefore, I think it should include stratification by activity since particular dates. Should be in the armory API somewhere, right?

given that it is text on the bottom of the armory page, there has to be a way to pull this data and very simply add another pie graph showing week of last update, or something. THAT would be illuminating, and in fact it is necessary if they aren't going to remove inactive toons if this site is going to have any relevance to anything.

I am equally curious about the OOPS toons still on realmpop lists. My toons went OOPS in late sept and then oct., and none of them are there, so maybe OOPS are being purged when found?

If they were to provide activity-date stratification, it would in fact provide a potentially informative data set to compare to what is known and extrapolated about q3 and eventually q4 western sub data. I have only looked at around 50 armories from that site, but the number of late december 'last active' updates is impressive.

[Kevyne-Shandris]

Yes, and overwhelmingly, by the point those characters reach the levels you've mentioned, they've either used the AH or been in a guild with someone who has, therefore making the population you're attempting to describe statistically irrelevant. You can say that you dislike the layout or that you're interested in seeing the 3 unguilded level 29 people who never use the AH that this missed, but the site is still useful, despite your weak arguments to the contrary.

The problem here is defending the LACK of data.

It's that lack of data that IS the problem. -_-

Building a car, but claiming a tail light isn't needed can get the car out of the factory, but wreck it in a week.

[Deficineiron]

The problem here is defending the LACK of data.

It's that lack of data that IS the problem. -_-

Building a car, but claiming a tail light isn't needed can get the car out of the factory, but wreck it in a week.

I am more than happy to concede that realmpop won't catch people who never used ah nor were in a guild which had ever used ah. I imagine every else is too. You are the only person concerned about the omission of this set of players. These players aren't tail lights, that is a false analogy.

My best recommendation to you is to ignore the site and any claims based on it, that will avoid depending on data which doesn't include toons that don't use ah nor guild themselves, ever.

Further reading on the armory issue suggests that the last update is the last date the character was logged out, with some log locations not causing an armory update (one work-around was to log out in outland or northrend). This gets more confusing than anything, though OOPS is definitely a deleted toon or account 6 months gone.

However, statistics-oriented people, many with a background in mathematics, will see that despite this glaring omission, realmpop has the potential to answer an awful lot of questions about wow western subscriber trends. If anyone knows the group behind the site, the 'update date' in armory would be a tremendously useful data field to include in their sortable/displayable data.

[Nosonia]

last time i did /who org 85 mage (very specific)... i still got 49+ results....

hard to get a census when i cant see EVERYONE due to query limits of 50.

How - To get the list of characters for a realm, first we record all the characters who posted to the auction house. Then we fetch and record their guild rosters. This should cover the majority of characters on a realm. To avoid getting listed, a character must never post to the auction house, and never belong to a guild where a guild member posts to the auction house.

this isnt quite as uncommon as one would think, especially for alts. My alt/main doesnt do any of the selling, it all goes to an AH mule to avoid getting whisper spammed when i undercut or buyout markets.

[Hoticehunter]

Worse than /who.

I and my toons rarely use the AH. If I don't auction, my three accounts worth of alts won't show.

And guess what? I'm on almost every day for hours, too.

"When - We re-fetch guild rosters about once every 10-12 days, and then fetch characters we haven't seen after about 20 days. The goal is to find all characters, not necessarily to have up-to-date guild rosters. The reports that come from this data get updated about once a week."

As long as you've put up an auction, you're in "the list" and if you haven't put up an auction in 20 days, they double check you anyway. If you're in a guild where one person has every put something on the AH, they have that guild in "the list" and double check that guild every 12 days.

Are your alts not in a guild? That's fine. They're probably not at max level and probably aren't really active anyway. This seems like a pretty ingenious way to cover active players, and a ton of inactive ones, though I don't really care if a census misses a couple characters that aren't active players anyway.

[roahn the warlock]

Looks like warlocks are the least played class, WTB class specific legendary that is handed to me on a silver platter. Ktnx

[Thyrial]

@Kevyne - Assuming you are the Holy Paladin Kevyne from Shandris-US which I would assume given your username, a quick look at Shandris on realmpop and narrowing it down to human paladins shows your guildmate Phannos on the selection of 200 players. Which means your character as well is included as you are in the same guild.
Will there be some characters left out? Sure, but considering any character in a guild where ONE SINGLE CHARACTER posts an auction will be counted the characters missed will be insignificant to the overall picture. Especially since the introduction of guild perks, very few people leave their characters guildless. To not be counted in this a person would almost have to actively try to avoid it. As with any collection of data, 100% collection is not necessarily important as a large enough sample size will give an accurate picture of what the realm is like.

[Ghostie]

Still doesn't address the issue of alts skewing the results, one of the realms my family and I play on is a dead realm, that's four people playing but showing maybe x24 active lvl 85's alone if this system is used, not to mention all of the lower level characters since we log in regularly just to check mail. I'm sure my family aren't the only people who are altaholics.

[Zoneseek]

seems accurate where I play

lol @ some of the horde:alliance ratios

[Erorus]

I've never seen such a large group of contrary players outside of the game. If I invented the fork, you'd complain that it doesn't work well when you try to eat your soup.

I wrote Realm Pop. I can answer some questions if you'd like to ask them here.

If you use the word "alt" in your question/statement, you're missing the point. I don't care about how many people are playing. We can't tell, because there's no way to link alts together to one account via any API. I'm counting characters, everyone I can see, regardless of age. The purpose is to compare realm populations, and to look for trends in simple character choices (gender, race, class, etc). To say "Darkspear reports 106k characters" doesn't tell you very much. Instead, saying "35.5% of Darkspear characters are Horde" and "Darkspear reports 106k characters and Dark Iron reports about half of that" means more.

I would like to get activity data eventually. The problem is that the Battle.net API doesn't provide accurate achievement points or last modified dates for characters in guild roster lookups. Without that data in the guild roster API, it would take me 65x as many requests to pull it from the character API, which is a bit ridiculous.

Finally, while the site does count every character it sees, I personally don't care so much about low-level alts, be they bank alts or leveling characters. They don't really have an impact on the game. They don't raid, they don't use max-level LFD, they don't consistently compete in Battlegrounds. If you have a BG twink, it's probably in a guild and/or uses the AH, so it'll get counted. But if you're on some anti-social alt that doesn't guild or use the AH, frankly, you're not really playing the game anyway.

[Deficineiron]

Still doesn't address the issue of alts skewing the results, one of the realms my family and I play on is a dead realm, that's four people playing but showing maybe x24 active lvl 85's alone if this system is used, not to mention all of the lower level characters since we log in regularly just to check mail. I'm sure my family aren't the only people who are altaholics.

Mainly an issue of developing a nominal divisor for avg alts/player/realm. However realmpop isn't filtering canceled accounts at all or until they hit their 6 month OOPS, so I see problems using that data at present until they segregate by update data/oops.

[Kevyne-Shandris]

@Kevyne - Assuming you are the Holy Paladin Kevyne from Shandris-US which I would assume given your username, a quick look at Shandris on realmpop and narrowing it down to human paladins shows your guildmate Phannos on the selection of 200 players. Which means your character as well is included as you are in the same guild.

Why drawing things off the Armory can lead to issues:

http://us.battle.net/wow/en/characte...evyne/advanced

Check the ICC timeline. Check the recent activity. Do the math.

Can't get more blatant about it: lack of data is a lack of data.

[Deleted]

Data is very inaccurate. I've checked few players who stopped playing about a year ago, all of them are listed there even though their pages no longer exist on armory. They are only listed in guild roster.

Looks like it just scans for everything it can find, without checking if characters exist.