Wow Realm Population

[Twdft]

I have active chars in 4 different guilds, only 2 of those guilds are even listed.

[Jargathnan]

This site pulls data in, arguably, the best way available. But, it’s still not entirely accurate.

My guild, which has somewhere in the ballpark of 30-40 active players (unique individuals, and this is a light count of people I can think of, that log in more days throughout the week than not); even more accounts (some, likely myself, multibox), and certainly many more active characters, given the number of active alts we have. Yet our guild only shows 37 active characters. I can count over 50 characters off the top of my head that are active per their definition. And that’s just among the players making up our raid team, never mind the dozen or more casual/social members and the alts our raiders have that I am not thinking about off the top of my head.

If my guild is that light in count, how many more are?

[Lochton]

I love how you enter every negative thread and instantly dismiss it instead of contributing to the actual discussion.

Did you even read the whole OP? He didn't try to state these stats are 100% irrefutable stats.


They are however real.


This is the best way to make an approximate estimate, which is what he has done. So why instantly dismiss this? Honestly why are you so defensive?

Didn't really dismiss them but you should know where this topic tends to go. As there is nothing even close to proof, it tends to conspiracy and guess work, which tends to be shut down.

I entered with my original comment because, that is a true statement.

[InTheEnd]

Anyone who thinks WOW has more than 2 million sub numbers is dreaming.

[VanishingAct]

All I know is that classic is calling

[Puupi]

Anyone who thinks WOW has more than 2 million sub numbers is dreaming.

I'd be surprised if it even had a million subs. I mean, who really plays WoW anymore? I know a handful of people. I've literally known thousands of people who have played the game and they have nearly all quit. If 99% of the people who used to play have quit. there can't be millions of subscribers left.

[rda]

No clue how they come up with their data (I understand you said they have to have guilds/players seeded), but the realm information for my server couldn't be more wrong.... Some of the biggest, most active, most progressed guilds aren't listed as active at all and the top HK guy on the server isn't listed... And I know he's active because he heads up the RBG group I play with..

My guess is that if people have to forcibly search for their guild/character to initiate the process then it's likely a huge number of guilds and players are absent.

People don't have to forcibly search for their guild / character to enter them into the system even as a seed - afaik, the seeding is done traditionally via /who scans (plus there're AH scans, so I'd personally add them as well).

I hear you -- and several others in the thread (thanks, guys) -- on the info for your server missing important details. It is expected for the data to have holes, what's important is to determine how big they are. I cannot come up with a useful test that would also be surmountable (eg, that I could do myself in a couple of evenings) at the moment, but I might in the future. For now, let's assume that the data has holes, that we don't know how big they are, so the data is mostly useful for dynamics - eg, for changes. Maybe later we can make the data useful for more than changes. Although there's always this big unknown of how to translate characters to players which will prevent us from going to players even if we solve everything else.

[MrLachyG]

oh look another unique thread about sub numbers. why the fuck are people so obsessed with finding out how many people play the game? and discussing it ad infinitum despite having no solid evidence of how many subs WoW has

[Drusin]

As I say every time population is brought up: My instance Q times are the same and my auctions sell the same, I have no issues with population.

- - - Updated - - -

oh look another unique thread about sub numbers. why the fuck are people so obsessed with finding out how many people play the game? and discussing it ad infinitum despite having no solid evidence of how many subs WoW has

Back in the day people used to use sub numbers as an indicator of how the game is doing, Blizz stopped releasing sub numbers so now players are literally dying inside without having a metric to point at lol

[rda]

As I say every time population is brought up: My instance Q times are the same and my auctions sell the same, I have no issues with population.

By the way, you would appear to be a little unique in the above, or maybe your auctions are just bog-standard "latest materials and consumables". Because auctions outside of "latest materials and consumables" (transmog, pets, etc) sell way slower than in previous expansions, and that's global.

[quite an expert in wow]

even if only thousand people played wow: as long as i can keep playing content the way i want to play it, i dont care at all

it feels like half of the people that 'care' about this are not even playing the game and only desperately try to convince others that the game is shit, to force developers to change the game to something they like

[rda]

I love how we are using a sample and we are entirely focused on how we can divide it to a smaller number (to make up for characters per account) instead of how we can first increase it to the actual user base, which is bound to be higher.

Who's talking about a sample? The site (attempts to) capture all active characters. It doesn't succeed, to an unknown extent, yes, but the idea is to try and capture everything or close to everything.

[faithbane]

I have active chars in 4 different guilds, only 2 of those guilds are even listed.

lol /endthread?

[rda]

And we know it cannot capture everything since we have ready anecdotal proof of it. Which means that it IS sampling.

No. It is trying to capture everything. There are holes, yes. The relative size of holes is unknown. But this is not sampling, this is trying to capture everything. There is no extrapolation from what's captured to "everything" as there would be with sampling. Because the system is trying to - you guessed - capture everything from the beginning.

Why am I saying the above instead of just saying "if you want to call that sampling, feel free"? Because of this phrase of yours - "I love how we are using a sample and we are entirely focused on how we can divide it to a smaller number (to make up for characters per account) instead of how we can first increase it to the actual user base, which is bound to be higher." What you call "how we can first increase it to the actual user base" (character base) is important, absolutely, but since the site is trying to capture everything, this might simply involve determining how big the holes are (perhaps using a sample for *that*) and then extrapolating the ratio to the whole. If the site was trying to sample in the first place, we'd be talking about measuring something else. The distinction is important.

But sure, it's important to know how many characters were really active in, say, EU+US. No disagreement.

[Qnubi]

Didn't really dismiss them but you should know where this topic tends to go. As there is nothing even close to proof, it tends to conspiracy and guess work, which tends to be shut down.

I entered with my original comment because, that is a true statement.

The OPs numbers are also true, it's true data. The minimum amount of players is exactly that. And, just like best before dates, the actual subscription numbers could be exactly the amount given by the site, or a little more or a lot more. Now, the question is, is WoW a milk product that quickly goes bad after the best before date or is it candy that can be good to be eaten way after it expired.

[rda]

BTW, the success of the Warcraft movie in China likely has way more to do with the popularity of Warcraft (ROC/TFT) there than WoW. They do like WoW, too, but they Warcraft much more. At least that's what I remember from analysis pieces after the success of the movie.

[Sughs]

3-4M sounds about right. No way they have more than 4M or less than 3M right now.

[Dormie]

Even if we knew how many subs, it's such a pointless statistic.

[rda]

For it to not be sampling then the methodology must have the potential to capture everything. If it had a crawler to read the entirety of the Armory instead of calling based on name lists then you can say that it is not sampling. But by default, characters who are not in a guild or using the AH actively would not end up in the name lists. This means that it starts from a sample (the initial name lists) and tries to expand to a greater whole. It is not canvassing.
I understand the distinction you are trying to make and I agree it is meaningful, I simply disagree that the methodology can be considered complete. The more correct way to do this would be to fully canvass a small number of realms via the Armory, compare it with the numbers they get with their current methodology, get weighted ratios and then extrapolate.

And I honestly feel that we have no decent way to make a call on an average number for active characters per account. We cannot even try to extrapolate from the past when subs were reported since alting behaviour may have significantly changed. I will readily assume e.g. that during Legion the number of alts likely ballooned compared to any other point in time in WoW as it does during the last patch of each expansion. Which means that even if we can make a solid estimate on how many active characters exist, it is very difficult to make any estimate on active accounts.

Why do you think the methodology cannot be considered reasonably complete? Here's what happens: we seed the search, doesn't matter how now (but it was /who afaik). Then we have character names and guild names. We take a character name, ask for its data, and as part of that we might get a new guild name - the (new) guild of that character. We take a guild name, ask for its data, and we get character names - guild members - some of which might be new. We then add a periodic AH scan on top, which keeps adding character names and guild names as well, and run this all periodically. This all grows to cover new and new characters and guilds. Sometimes we lose a name when a character or guild gets deleted, too. This process needs time to stabilize - initially the numbers (of all, not active, characters and guilds) grow fast, then they saturate and grow or decrease about as much as they do in reality. If our scans are regular enough, that is. The numbers on the site passed the fast growing phase a long time ago (6 months for most of EU+US), that's long enough. I don't know how regular their scans are though, that's true.

Who do we not capture in the above process? We don't capture levels below 10 because of API limitations (we could have been capturing some of them, but we'd have to handle them differently from 10+ characters, so the site didn't bother - I would do the same, I don't think levels below 10 are important). We don't capture people who (a) aren't in guilds and (b) don't post anything on the AH. That might be sizeable, true, although perhaps only potentially. I think it's good enough to have an approximation on the number of active characters and then say that WoW has that plus an unknown number of people who also play, but outside of guilds and without using the AH.

Regarding no decent way to make a call on an average number of active characters per account - yes, this is a tough nut. This is not the first time this question stands and there was never a good answer of how to even estimate it suggested. Agree here.

- - - Updated - - -

Anyway, the point of the thread was that this is one more data source, here's how it works, seems to be reasonably good, etc.

Such things are useful even without having a good way to estimate all the juicy numeric factors that in the end produce absolute sub numbers. Eg, we can look at changes of numbers that are measured and talk about getting X% more players in June than in July or vice versa, and across the year / patch / expansion, etc.

[Dystemper]

Even if we knew how many subs, it's such a pointless statistic.

It's a sub based mmo , not pointless at all