I dunno, I've used the Census mode since 2009, just because it gives interesting data snapshots. I could look at the end results data of the queries and when I started to see lower players seen on-line(as part of it's scan), it pretty much always matched the following Blizzard quarterly announcements of subscriber drops or gains. It also matched the faction totals. On our server people swore that it was 2 to 1 Horde to alliance, but the data always showed the server being fairly balanced and Alliance normally had 200-300 more people on-line than horde. This was roughly confirmed by sites like WoW progress and such.
Granted I'm on a medium pop server and so the scans normally only takes 6-7 mins to run versus a super server like Illidan-US in which one scan takes over an hour at peak times, So I would accept the data is less accurate for super servers than Med to small pop servers.
Again, I just like the general data about the server. How many people choose Orc, How many made a monk, what was the most common race type for Death Knight etc etc and then maybe go to another server and see how that matches up
Obviously all data should be taken with a grain of salt and understand it gives a good idea and not an exact headcount
Highly unlikely, all the cool kids use Twitch stream viewers as their measuring stick of the state of a game. Which is a funny metric to me, because using a metric of How many people are watching me play a game, versus playing the game just sounds funny to me. It's obvious some folks have a vested interested in seeing Classic fail, so any measuring stick that supports their claims will be used
Last edited by Mad_Murdock; 2019-09-20 at 12:09 PM.
I'm pretty sure that was fixed a while ago.
But, this is Classic, and while they have done some recoding, it does look like the vast majority of the coding remains intact. Which actually makes sense, since Blizz was mostly looking to roll this out on the cheap for a quick cash grab of 1 million or so subs for a few months.
Now, however, it looks like Classic has more staying power than Blizz anticipated, so they are trying to figure out how to deal with older code concepts existing in a world with much faster computer and internet connections. It doesn't help that the conspiracy nuts are hammering this functionality along side those trying to do basic census the same way they've been doing for 8+ years now, not thinking through that they are hitting a 15+ year old game.
If you were to resurrect some old internet sites from 15-20 years ago, you'd see similar flaws. Heck, I used to work on a system 20 years ago where we had to split data across multiple databases because no database at that time could hold it all on a single server.
Probably because it flooded the servers with /who's.
Originally Posted by Blizzard Entertainment
Considering the wildly inaccurate results, and people making server\factions decisions and starting discussions based on them, i'm glad.
Also, having thousands of players with addons spamming /who commands is probably not good for the servers, and i suspect they were the cause of layer lag this past weeks.
People making a big deal out of it is just absurd.
If done right Census was highly accurate for Classic. There was no issue with the results. Do primetime checks for servers and you'll see how accurate they were not only giving you insight on the different server statuses (cap of 15.000 online players until queue, Low -> Middle with 2000 or more online people etc.) but also for faction balance during primetime.
Last edited by Nyel; 2019-09-20 at 12:20 PM.
MAGA - Make Alliance Great Again
Even WoWHead used this addon and made articles about it.
https://pt.wowhead.com/news=295075/c...ough-community
It wasn't a stress to the servers in any way. You see people and mobs around you and your client constantly receives and requests much more data from the server: their guild, what are they casting, what their level is, a whole bunch of data. Just an example: 3 hrs raid combat log is about 1 GB large. Full scan of server using "who" is about 1 MB. Yes, it's not always about the data size, but people been using census addons on retail for ages. Blizzard never complained about it.
The data gathered by the addon was super accurate. That is why Blizzard don't like it. They don't want this data to be available to the public. As an example, after using addon for about a week my scans rarely find new players. For about 3-5k online players each scan added only 20-50 new players.
So, they made "who" call protected. This changes nothing. I think and hope the addon will be updated to include a button which you will have to hit to run each "who" request (this is how protected functions work in addons). The data will become less accurate, because most of players won't do that. There still will be dedicated players who will continue using addon.
So i followed the WoWHead link to the page that has been updated with fans using this addon.
Count me interested in this since Blizzard is trying to prevent it
http://wowpop.appspot.com/
This is the site
The data gathered information on 7,941,842 different characters.
5,513,934 seen in last two weeks.
Cute numbers
I'm not an addon writer, so forgive my ignorance, but: aren't the results that you upload to the site easily modifiable? I edited the resultant .lua in notepad easily enough :/ But again, no idea how the site parses this data.
How do you *actually* know this?
Ah. You don't.
But they have complained about several other addons causing problems with their technology. It's not impossible that the old 1.12 client/server combination does costly db queries, which were never a problem back in the day, because the datasets weren't that large.
How do you *actually* know this?
Ah. You don't. It's just anecdotal bullshit turned into a fact.
Yawn.
Once again blizzard's cowardice in scaring Activision shareholders results in another stupidity...
the fear of seeing the 900k players still playing retail becoming public was enough to make them fart in the flour
Good riddance considering we have roomtemperature IQ scientists making threads how their own census data shows that a realm has like 10 people online.
- - - Updated - - -
What if... they shut it down because it didn't belong in the spirit of classic?
To the ppl saying these census results ate wildly inaccurate, why do you believe so? Do you understand that it uses /who with very specific terms to get player counts over the course of a couple hours? Obvious ppl can log off and in during the scan so it wont be 100%. But with multiple ppl running the scans at once, frequently enough, the data is statistically pretty accurate. Whether or not we should have access to the data is a different discussion, but why do you believe it's no where near accurate? You also realize they log character names so it doesnt double count the same player right?
Do you have any evidence the addon was actually able to track players, rather than characters? Cause tracking characters alone is mostly useless for the purpose of a census; assuming that classic has the 8 characters/server limit you can have as few as 3,750 players for 30,000 characters, and even if you assume most people only have 2-3 characters that still winds up being less than 15k players.
It not double-counting the same character doesn't mean a damn thing, because the number that actually matters is the number of players.
It does not track players, but characters so you are correct. But you can filter by level.
I personally filter by lvl 40 and up to get a good look at active players and the ratio. If someone has 2 lvl 40s in less than a month, then they are a very active player anyway and prob play as much as 2 other players so I have no problems with that player being counted twice.
I usually use the census to compare horde to ally ratios of server by filtering to above 40. These are the characters that are likely to stick around on the server since they've already invested enough play time to be at least half, or near halfway done the levelling process. If a server has a 3:1 faction ratio of players above 40, that server us fucked.