WarcraftRealms.com

Re: sample/data cause of weekly faction trend divergence?

1974ER — Wed Oct 12, 2011 1:31 pm

Author: 1974ER

Posted: Wed Oct 12, 2011 1:31 pm

Yes, seems someones have managed to get almost all of US out of 0-state as well... After some more of my patching processes, we might actually have one of the best situations during the entire year, as far as minimum coverage is concerned...

Re: sample/data cause of weekly faction trend divergence?

Balgair — Wed Oct 12, 2011 7:38 am

Author: Balgair

Posted: Wed Oct 12, 2011 7:38 am

It worked, but seeing as I'm asleep during their peak hours and my peak censusing hours are the middle of their night, I got hardly any updates out of it lol I also got an account locked for suspicious activity notice and had to reset my password, so I guess I won't hop over to the US too often

Useful to have the option available though, and it definitely did work, I'm not the only one who's been censusing going on the improved Most Wanted list (heh that's hugely improved actually, wasn't looking that good last night! Someone's been working hard!), but I did remove about 10 factions from there

Re: sample/data cause of weekly faction trend divergence?

1974ER — Wed Oct 12, 2011 7:32 am

Author: 1974ER

Posted: Wed Oct 12, 2011 7:32 am

I have to admire your fighting spirit and innovativeness. I hope your experiment was successful.

I'll stick to the EU realms though... there have already been new holes and there is more to "patch" today as well.

Re: sample/data cause of weekly faction trend divergence?

Balgair — Mon Oct 10, 2011 9:02 pm

Author: Balgair

Posted: Mon Oct 10, 2011 9:02 pm

Ok for US patching, I think I have a stopgap solution, thanks to a PTR thread on how to get premades if you're in Europe (once again, European accounts didn't get any, workaround was to create a US free account ):

Quote:

HOW TO:

under > YOUR GAME ACCOUNTS

1) click Buy Digital Games LINK (don't worry no money is going to be used)
2) in the top right click Game Region and change it to North America.
3) Go back to to YOUR GAME ACCOUNTS
4) Under all your accounts you should now have add WoW starter World of Warcraft�
Americas & Oceania (US)

Combine with the following from Wowpedia regarding realmlists:

Quote:

During patch 3.0.2, the file was moved out of the main World of Warcraft folder into a subfolder (\World of Warcraft\Data\enUS\ or another language).

For US clients:

set realmlist us.logon.worldofwarcraft.com
set patchlist enUS.patch.battle.net
set realmlistbn ""
set portal us

...and US realms access! I've just censused three that were on 1 or 0 on both sides, will see if it processes ok, but fingers crossed it'll work alright. Oh yeah to not screw up access to European realms, I copied my WoW folder and just made a separate install for the US servers, so I don't have to keep switching around. Will see how it works out but it might be some help on those rare occasions when we have our own side of the Atlantic under control

Re: sample/data cause of weekly faction trend divergence?

Balgair — Sun Oct 09, 2011 6:25 am

Author: Balgair

Posted: Sun Oct 09, 2011 6:25 am

I can only agree, as an altaholic with 2 accounts, both full, on my main server, I've had to spillover to another one but would rather all my alts were on the same server. Really can't justify paying for a third account though.

Yeah the English realms used to be better, but I think there were more people covering them then, you and I could fill in the gaps between us but now, there's just too many gaps. US realms seem to be suffering as well, the people who filled in gaps over there I guess have left (there's a few names who used to be in the top 10 a lot that I haven't seen in ages).

Hadn't realised your point about just running a few a day covering so much, good point - I need to apply that to myself, since I usually do loads at a time or none, should just spread it out more and I'll get more done!

Re: sample/data cause of weekly faction trend divergence?

1974ER — Sun Oct 09, 2011 1:00 am

Author: 1974ER

Posted: Sun Oct 09, 2011 1:00 am

Indeed. However, even upping the maximum to just 100 would help tremendously. Though not topic related directly, I also wish they pushed the maximum per realm to 12 or so, since there are persistant rumours of a new class, though people can't agree upon what class that is... :/

Some German realms have enjoyed a high census level for a long time and I think they actually have the best Censususer to number of realms ratio.

Off the cuff I can't recall ever seeing a post by a French Censusplus user, but there obviously have to be some...

I sometimes dip in a minor bit of Spanish data, mostly when multiple realms have had 0s for longer periods.

The English realms... They make me sigh... I can still (very vaguely) remember a time, when most of the ones not being covered by others could be held up at 1 by my efforts alone, but that was somewhere around the time you broke monthly updates and new ones records and my play to just census ratio was different.

And, yes, I am sure your suspicion is entirely correct. Funny thing is... a single person who would wake up early and run a few censuses (about 5 factions per day) on premade "permanent alts" could in theory hold up around 145 factions at 1+ with an time investment of roughly 15 minutes per day. At that sort of rate... 5 people would suffice to give each and every EU realm minimal coverage, if they managed to avoid massive overlapping (there would automatically be some), another 5 would give entire US same minimal coverage.

I already regularly cover a much higher number of factions per day (because I actually play on 6 factions and average 10+ others per day at my current "minimal effort" level), so all I can offer are some occasional activity spikes of "running all over" to "kill off some 0s".

Re: sample/data cause of weekly faction trend divergence?

Balgair — Sat Oct 08, 2011 6:29 am

Author: Balgair

Posted: Sat Oct 08, 2011 6:29 am

Rofl yeah, I can see why they don't allow it but still, would be nice

Been impressed by the German realms lately, whoever's taken them up is doing good things - actually, Eloise seems to have a character on a German realm (although most are on English), perhaps she's responsible? Would make sense if so Some of the Spanish work is me (Tyrande, Los Errantes, Sanguino all both sides, 4-5 others on one side), no idea who's doing the rest but it all keeps up, thankfully!

Those are some nice numbers indeed, if only we could get more realms that well covered! At least my own home realm EU Draenor finally has somebody censusing on Alliance, I used to pop up every time someone on the realm forums moaned about the data (every few weeks there was someone claiming the data said there was 90% Horde pop on the server, when in reality it's about 60-65%) and point out that Alliance data went to the pits after I switched to Horde, evidently someone took the hint and got the addon since it actually keeps up even when I don't play my ally alts now, phew! For one of the highest population servers in Europe, there's only two people censusing though, one on each side (Horde dips to nonexistent if I take a break), that's not too good a ratio.

I suspect a lot of realms are reliant upon one person, or strongly helped anyway, EU-Dun Morogh used to be one I'd notice for really high numbers a year or so back... now, someone I know from another forum plays there, and I know he uses Censusplus. He stopped playing a few months ago and now I see it's down to good-enough-but-not-amazing numbers, no coincidence I suspect! Even one person per realm is something anyway, I always delete my census alts when I see numbers get over about 30, since that indicates somebody's censusing at least enough to keep it covered, just a shame that most English realms in particular are nowhere near that level.

Re: sample/data cause of weekly faction trend divergence?

1974ER — Fri Oct 07, 2011 2:49 pm

Author: 1974ER

Posted: Fri Oct 07, 2011 2:49 pm

I have to agree with you... I just doubt that Blizzard would be very happy with me having approximately 650 characters spread "all over creation".

Yes, the extemely limited number of Spanish realms simplifies that part... as does the fact that someone tries to keep part of them up. Germans tend to have reasonable coverage, apparently by a few people who do not really post here, but do a fair bit of important censusing work. Same goes for a fair portion of French realms.

The English realms are the biggest problem area. There are a lot of them and some of them have had poor coverage for literally years. A fair number of them suffer from generically low population, which amplifies the problem. On a large realm, the likelyhood of at least one person having Censusplus installed goes up quickly.

And some realms are obviously covered by multiple people... It's a shame we can't offer everyone who comes here looking for info just that. A well established database, which isn't reliant on a single person.

US-Whisperwind, 567 Alliance entries
EU-Scarshield Legion, 785 Horde entries (large part of these are mine)
EU-Spinebreaker, 507 Horde entries
EU-Alleria, 321 Horde and 282 Alliance entries (part of both are mine)
EU-Culte de la Rive Noir, 812 Alliance entries
EU-Die Todeskrallen, 622 Horde and 849 Alliance entries
EU-Un'Goro, 402 Horde entries

Couple of things stand out on that list of some of the (currently) best covered factions... US side is massively underpresented and none of the Spanish ones have superior coverage. Having the coverage level of EU-Die Todeskrallen would pretty much be Rollie's dreams come true... Sadly, we are REALLY far off.

Re: sample/data cause of weekly faction trend divergence?

Balgair — Fri Oct 07, 2011 9:31 am

Author: Balgair

Posted: Fri Oct 07, 2011 9:31 am

Yeah the English realms were dreadful. I noticed somebody fixed the French ones a week or so back, which helped a lot, and German always were better covered than the rest, Spanish are badly covered but easy for 1-2 of us to run 1 census a month, given there's so few of them. There's a lot of English realms though and not many that seem to have active censusers playing on them.

I wish there wasn't a character limit on accounts, would be nice to be able to census all needed servers without having to delete and make new characters, I find it far easier to just leave my characters existing!

Re: sample/data cause of weekly faction trend divergence?

1974ER — Fri Oct 07, 2011 9:11 am

Author: 1974ER

Posted: Fri Oct 07, 2011 9:11 am

Well, under current circumstances it's very hard to imagine a day when no realm at all would 0 out. I am hoping the Spanish servers stay that way for the next 24 or so days since there are so few of them. The German and French realms will be more of a challenge... and the English ones were in utter shambles... I ran more "patches" on them than on the French, German and Spanish realms combined. :/

Rollie's long absense is of course regrettable, but real life matters must take first priority, the site is just a hobby after all. Until then, I, you and the other "oldsters" have to try to keep things together at least somewhat and assist the newcomers the best we can...

As for US... One busy person could fix it... in about 1/4th of the time I needed for my share of EU "patching", no less, if (s)he started running censuses very early US time to reduce individual census lengths. Overall it's interesting to note that this might be the first time in almost 2 years when EU realms have better theoretical coverage than the US ones. :/

Re: sample/data cause of weekly faction trend divergence?

Balgair — Fri Oct 07, 2011 8:28 am

Author: Balgair

Posted: Fri Oct 07, 2011 8:28 am

Dammit didn't take long for that 0 to appear - it'll be gone again when my next submission gets processed though, used my only empty character slot on it

And yes, a few people make a big difference here - I feel bad for not doing so much updates as I used to, since that's dragging it down a bit. I try to at least run over all my characters once a month or so, but that's just covering maybe 50-60 servers, and since I prefer to not delete characters from servers that aren't well covered, I don't tend to take on many new ones with 0 submissions.

I suspect things won't improve much until Rollie has the time to look after the site again though, since potential new members will think the site is quite neglected-looking and be put off. Of course some will join anyway, but seems to me most new people on the forum are here to ask why something is out of date, and all we can do is say "it'll be updated when Rollie comes back", which isn't too definite an answer

That being said, we (well, you and Eloise ) have done well to get it back to no 0's for now, it's been absolutely loads of 0's ever since Cataclysm and the issues then, so to get it back to where it is now is a major improvement Now we just need a power-censuser or two on the US servers to get rid of their 0's!

Re: sample/data cause of weekly faction trend divergence?

1974ER — Fri Oct 07, 2011 8:17 am

Author: 1974ER

Posted: Fri Oct 07, 2011 8:17 am

Thank you, though I can't claim all the credit... not even with the fact that Saturday and yesterday combined I updated the status of approximately 120 different factions... I am not sure if Eloise sees this, but if she? does, I want to extend a big thank you that way as well.

As for 0s, they keep slipping back in... there is at least one already again... :/ We really need more coverage... and that seems unlikely. It rather seems that with some rare exceptions, the submission numbers are going down. I don't like saying this, but unless we get at least a few new active members, the day when the "0 war" is lost may come sooner than I would like.

Blizzard can afford to lose 500k customers... We, on the other, bleed horribly if we lose 5... I can just imagine the impact of what would happen if I, you, Gmmmpresser, Postponer and for example Eloise quit entirely. In just 24 hours, multiple realms would start to deteorate and on 31st day, 0s would rule supreme on the most wanted lists and elsewhere. Such a blow would not directly kill the site, but it would render it much less useful.

Re: sample/data cause of weekly faction trend divergence?

Balgair — Fri Oct 07, 2011 5:29 am

Author: Balgair

Posted: Fri Oct 07, 2011 5:29 am

1974ER wrote:

Just out of curiosity, since I have achieved most of what I want to do today in terms of Brewfest, I'll try running censuses on low / no submissions factions during the evening.

Happy censusing, everyone!

Seems you've done some good work - there's no more 0 submission servers in Europe May not last long but it's nice to see the 0's eliminated, been ages since that happened

Re: sample/data cause of weekly faction trend divergence?

1974ER — Tue Oct 04, 2011 3:13 am

Author: 1974ER

Posted: Tue Oct 04, 2011 3:13 am

I am sorry for having to offer just another educated guess, not a fact...

The graph seems to operate with four points per month... BUT... with the exception of February, months aren't exactly 28 days long. Which, I presume, means that actual average weekly activity doesn't get set in stone before the entire month has ended.

And so, the question becomes... is a 31 day long month:

a) 7+7+7+10 days long?
b) 8+8+8+7 days long?
c) some other variant of the options above?
d) something that I didn't even think of?
e) I misread the graph and it's for example 7+7+7+7+3 days long?

Also, we already have a data point for the first "week" of this month, even though it is only the 4th. This seems to support I my suspicion that the data gets recalculated (daily/everytime someone calls for it?) until the month is completely over.

And by "completely over", I mean until no further data is submitted for it. Due to timezone differences, for example, I myself, monthly supply (large amounts of) data that still affects the previous month. Someone who submits very rarely, might cause such alterations much later, possibly even weeks later.

As a stopgap measure, I suggest you might want to monitor some slightly older data point(s)... let's say the weeks of August, for example. They should remain as they are as it's not likely that anyone would add any more "new" data for August.

I apologise for possibly being unclear, I didn't sleep very well last night and am therefore not entirely sure if my "reasoning" works properly... :/

Re: sample/data cause of weekly faction trend divergence?

deficineiron — Mon Oct 03, 2011 8:35 pm

Author: deficineiron

Posted: Mon Oct 03, 2011 8:35 pm

Thanks for your reply. What I mean by lowered is the data points for dates past on the previous graph are lower now than they were last week, e.g. data for 2 weeks ago alliance is lower than the same time/data point was before the update. The entire curve of the last month of alliance has changed from about 4 points back in this last update.

Thanks again for your insight, this is a fascinating process.