I just started a project that is dedicated to gathering census data via an ingame addon and renderring it into some filterable charts on a website. So far I've collected some decent samples and also some other people from the beta joined and submitted data.
If you wanna have a look on it, you can find it here:
Really intresting, well done
It seems to confirm what I suspected: mages are more popular than warriors. It seems a pretty clear result of increasing game knowledge, but I wonder if it will stay through to release. Hunter numbers are obviously slightly inflated for this reason of leveling. I’m not sure of the reason for priests. Maybe they fact that they’re known as the best healer by far except for a few niche situations, and shadow melts faces?
Well at least I won't be hurting for food and water with all those mages running around.
Ups, the screenshot is a snapshot from the beginning of the website from like 3 days ago. Now that other people started submitting data, the numbers look a bit different, at least for level 40, see here: https://wowclassicpopulation.com/charac ... Beta%20PvP
Edit: Also updated the picture in the first post
Really great job! I love seeing census data from WoW. I'll have to download the addon to contribute. Gotta get as much data as possible!
Ah, damn, guess I was wrong. That's the strength of being the only class that can tank I suppose.
Right? I always like to look at these stats and then imagine how that impacts the quality of life of individual classes and specs. A warrior on alliance side will have a TOTALLY different PvE (raiding/gearing) experience than a horde warrior. A warlock on alliance side will have a totally different PvP experience than a horde warlock.
~10% of alliance are warlocks, even fewer of those will be actively participating in PvP and going for rank. Yet a whopping 38% of players (40% on the Beta PvP server) they will be competing against will be undead with will of the forsaken. That would be a brutal reality. Say what you will, but that is a massive difference in quality of life. Im not advocating that alliance warlocks are trash or that they arent relevant, simply that they will have a much more challenging time in PvP. This is especially true when meaningful buffs they use like soul link are just getting purged by shamans.
Lets look at the alliance warrior. 15% are warriors, 12% are rogues and 14% are paladins. Now look at the human population.... .... .... 42%!!!!! The alliance human warrior (BiS PvE racial) is competing for 2-hander swords with the rets, plate against the rets, 1-hander swords against the rogues and other warriors! Soooo much loot competition for swords to capitalize on the human racial! The horde warrior? 16% of horde are warriors. Only 27% are orc, actively seeking axes. No ret paladins to compete for plate. Rogues cant use the 1-hander axes. SO much easier to gear up for the min/max orc warrior than it is for the min/max human warrior. How about for the Orc warrior who is seeking to temporarily tank during his gearing phase? While paladin tanks are few and far between, they are still more common than shaman tanks. The Orc warrior has even more demand as a tank in 5 mans allowing him another opportunity to gear up a bit faster than the human warrior. All of these little advantages and differences add up.
There are tons and tons of instances of this, but it is interesting to consider the different adversity each of these class race combos will face depending on their faction. This makes the game super dynamic and I would not want it any other way. I don't think many people even consider the nuances in these early choices and how it will impact them throughout their entire experience. This is one of the great design pillars of Classic. You have choices and the game does not hold your hand... The game will let you make a choice that will hinder you. Perhaps you want the extra challenge, or perhaps you value some niche thing that you are willing to sacrifice for. Super interesting to think about though.
Great work. I'm really eager for this to get more data.
It's interesting to see how the current data lines up with the reddit class survey. The main discrepancy seems to be fewer Warlocks and more Hunters at level 40. But given the level cap it makes perfect sense — Death Coil (Rank 1) is trained at 42. If the cap is raised to 50 we might see that ratio change a lot.
Good, good... the fewer warlocks I run into the better.
You still didnt explained why it's "not that detailed by far". I'm very open to feedback and want to improve the project so I just want to understand what makes you feel that the stats you've linked are by far more detailled.
I AM TALKING ABOUT THE LIGHTSHOPE STATS - Because I mentioned in THIS THREAD for people who wanna see some ACTUAL STATS from the Private Servers (They are bigger than the beta, also the new Silverhand realm which got released on Saturday is already bigger than the beta)
It has nothing to do with your project.
Ahhh now I think I understand, you just talked shit for the sake of an attention grabbing introduction to advertise a pserver that will be dead in 2 month. Grossaeh wrote: ↑7 months agoIt has nothing to do with your project.
Not really, I just sent the people interested in stats a link where they can see how it is currently on the private servers. I'm sorry, but the beta stats aren't really meaningful at the moment.
On the private servers you currently have a much larger range of players. Best example is Silver Hand, which was opened on Saturday to give people without beta access the chance to play one more time before the release.
And it's not advertising either. Apart from that, private servers and everything that has to do with them is allowed in this forum, it's not shit ;)
I don't want to attack your project or present it badly, in the long run your project will probably become a very great stats site.
But here you have already lost a user...
Test wanna see if some time or something change on the post.
What do you mean in with "not in PvP?"
What the hell. All he did was to link pserver stats in case someone was interested. Not once did he "talk shit" or "attention grabbed". Even complimenting you! The first thing he said was that yours is more detailed.christophrus wrote: ↑7 months agoAhhh now I think I understand, you just talked shit for the sake of an attention grabbing introduction to advertise a pserver that will be dead in 2 month. Grossaeh wrote: ↑7 months agoIt has nothing to do with your project.
Why are you this hostile?
What? Tanks don't really exist in pvp unless we are talking about FC which is the one thing where druids shine more than anyone in vanilla. You are making no sense.
I think people in this thread are.... broken.
Well, there is a huge difference between the detailedness of a statistic and the sample set it uses. It's obviously not my fault, when you talk about the detailedness of the projects statistics, but actually mean it's sample set.
I absolutely agree that the sample set from the beta is maybe highly biased, but we dont really know that.
But do you really think that the sample set of this private server is not biased? First of all only people that don't care about piracy join them and thats a huge bias from a statistical point of view, especially when you aim for representative statistic about the whole playerbase. The second huge thing that prevents avarage people from being represented in the set is accessibility. Initially you have to be aware that pservers are a thing, then you have to find one, register on their site, search the net for a client that is compatible with their server and edit a config file. I know thats no that difficult at all for people like us, but its definitly too much effort for an avarage person, only to join a server that becomes meaningless 2 month later. So what you probably see in a sample set from this specific pserver is highly invested hardcore players. Just to be a bit hyperbolic on this point for fun, thats like trying to make predictions on a real market stock by analyzing a random dark net market place inside the tor net :-P
The next big thing you have to think about is, that sample size is not everything that matters in statistic. Sure in general more samples are always better, but when you can select from a pool of people that you have a lot of information about, then a hand-selected set of 1k samples is far superior to a set with a size of like 10k samples that has a lot of unknown bias in it.
Yes it's true. we don't know how Blizzard actually selected the player base for the beta, but I would expect, that they've statistical professionals on their team that finely selected the samples to ensure a high level of diversety, so that they can collect meaningful, representative statistics for themselfs, to be able to predict different things that are important for them for launch. But yeah the data in the census project isn't very complete, at least not for pvp realm horde side, because we are getting data from that side only since thursday. But that only really matters for the overall statistic of that realm, and not so much for the horde specific statistic at level 40. So after all, if I had to make an assertion I would say, that the data from the beta is more representative compared to the pservers one.
But finally I dont think anybody of us is really capable of finding out which sample set is "worse" or "better" and so no one can really make a valid statement about it.
For the past 5 months, I've been working on a 'overlay' style project for Classic. It's called Over(p)lay! - Would meanFor program updates, go to my twitter found here!
Ever since I've been playing original vanilla WoW, I've been wondering if there...Deaththinius • 4 months agoCaperfin
4 months ago
5 months ago