Thread by @FoldableHuman

Quick straw poll, if you're planning on checking out WoW Classic, which launches next Monday: do you have an active subscription already, or were you planning to do that just before/shortly after launch?

I actually don't think they have because there's a far bigger bugbear hanging over the whole thing that makes this a tricky situation to solve.

https://twitter.com/Doomcakeplaysg1/status/1163521058726694918?s=20

I estimated months and months ago that WoW Classic was probably going to stabilize around 125k players in the Americas/Oceanic region. Blizzard seems to agree with me because they've got 14 realms for the region which is enough space for a very comfy 8-12k players per realm.

The problem is that launch population is probably going to be in the neighbourhood of 1-2 million players. So they need a system that can handle 1-2 million players at launch without dying once 90% of them get their fill by mid-September and go back to Overwatch.

Since the thought leaders in the Classic community are philosophically opposed to cross-realm tools like Dungeon Finder, Blizzard needs a solution that doesn't just recreate the mistakes they made back in 2004/2005.

When WoW originally launched it exceeded anyone's expectations by hundreds of thousands of players and it was absolutely crushing their hardware. Blizzard's panic solution was to just keep opening more realms. They were launching new realms as fast as they could get machines.

This was kinda the right answer, but they over-shot the sweet spot and once things died down, free transfers were offered, populations stabilized, servers were upgraded to handle 5-6k players per realm instead of 2-3k, a lot of realms just kinda died.

Now, a big thing that needs to be kept in mind is that the death knell for any MMO, no matter how well it was actually doing, was news of "server mergers". If server mergers didn't outright kill your MMO they definitely killed its perceived cultural relevance.

Blizzard subsequently has jumped through a lot of hoops to effectively do server mergers without ever needing to use the term "server merger."

Now, a lot of those tricks, the sharding and layering systems, cross-realm play, are things that the Classic thought leaders are philosophically opposed to and have gotten hilariously upset that Blizzard is using them at all in any capacity.

Now, sharding and layering are just systems for handling large population loads so that, for example, a questing zone has a comfortable number of players around. Enough to feel alive, not so many that every mob is dead the moment it spawns.

In response to the complaints Blizzard has pretty much just said "put on your big boy pants, we're using them to handle launch" but promised to shut them off "permanently" by some date in the next couple months.

Cross-realm, though, is a tool to handle *low populations* and a realm having a critically low population is a much more permanent state than launch attracting a bunch of tourists. You can promise temporary layering, but if you have to turn on cross-realm it's going to stay on.

So in the long run Blizzard would rather deal with "oh no, our game is just *too* popular and making us *too much money* woe is us" than having a buttery smooth launch with a gift of server mergers for Christmas.

Likewise they would rather be adding servers in November if population stays in the 200k range than closing servers in December because they have 120 active players.

To bring this back around to the poll, I assume what Blizzard is counting on is that the less-invested players, the ones who didn't bother to stress test, didn't bother to reserve names, are the players that will follow the path of least resistance and be the most temporary.

These players, while being the biggest chunk of the total launch population, can be used to fill in the cracks. They're mostly not planning on playing much past the end of August, so they're not exactly picky about making sure they end up on a realm with the right people.

WoW decoupled "realm" from "server" a decade ago, so even if people (myself included) use them interchangeably it's useful to know the difference, which I think will illustrate some of the comedy at play here.

https://twitter.com/Xanitross/status/1163824431262765056?s=20

At launch a realm (player collection) and server (physical hardware) were the same thing. One server blade = one realm. Blizzard has actually auctioned off a bunch of the old hardware. I know the guy who owns the blade my realm originally ran on.

How Blizzard avoided needing to use the term "server merger" was playing off the fact that no one actually cares about hardware, they only care about the player groups, the realms, the outward symbol of the server.

So just find a way to put a bunch of realms on the same hardware

This is pretty commonplace cloud server technology these days, but it was innovative stuff at the time and a brilliant sleight of hand.

The cross-server-play drama has nothing to do with hardware and everything to do with a contingent of players who want to be stuck in a bottle. They don't want to interact with players who aren't on the same realm.

There's also a bit of a Purity of Truth thing going on with opposition to load balancing technology like layering because those technologies obscure "the truth" - you're not seeing literally how many players are in a space, you're seeing an idealized, balanced fraction of them.

A lot of the purists idealize the fact that a lot of things in vanilla WoW were luck. You got lucky and chose a good server. You got lucky and found a good guild. You got lucky and met the right people. They want choosing a realm to be a gamble.

If you chose a realm that's got 30,000 players on it they don't want Blizzard obfuscating that by portioning the players into manageable chunks: it should be awful and unplayable. That's the gamble. Ditto for dead realms.

So Blizzard doesn't really need to worry too much about about server population outside edge cases like launch and AQ event because that's hardware and it's all cloud and dynamically scaled

but they need to worry a lot about *realm* population because players have rejected cross-realm play as a solution to population imbalances, and left unchecked early imbalances only tend to get worse as time goes on.

A dead realm becomes a social problem. It gets increasingly difficult to do any group activity. The economy, which has drop rates balanced around the assumption there's a few thousand players active, becomes deeply unstable and vulnerable to manipulation

https://twitter.com/haploidcheek/status/1163870036555157504?s=20

Keep in mind that Classic is deliberately recreating the original leveling curve where getting to level cap (60) takes an average player around 70-80 hours. Dead realms bleed players, because players would rather quit playing than put in another month gambling on a new realm.

This is the gamble that I'm referring to. It's "more interesting" to a lot of the purists if there's a risk that the realm you pick at launch could just die, that when this screen pops up there are, in fact, wrong choices.

So Blizzard is trying really, really, really hard to ensure that realm selection doesn't have wrong choices beyond "oh, maybe I don't want to be on a PvP realm"