On Bottlenecks and the Start of Expansion Rush


World of Warcraft, despite its recent years, is still a pretty big game. One that always gets a surge of interest right at the start of expansions. It’s up to Blizzard to try and make sure that the starts of expansions go as smoothly as possible because as with any game, creating a bottleneck right at the start (or toward the start) of expansions with result in a lot of negative backlash. For the current expansion, Warlords of Draenor, it hit multiple national news sites such as the BBC when the garrison technology either trapped people inside their garrisons, leaving the rest of the world a phased husk with no NPCs outside of the garrison walls, or the exact opposite where trying to get into the garrison left that empty.

However, there was a bottleneck even before then that created the actual bottleneck and caused a lot of rage within the community. As the Alliance, you had to place a flag into the ground, but it could only be done one at a time. Even on a quiet server such as Emerald Dream, you still had to compete with ~100 other players to stick their flag down into the ground. I don’t know what the situation for Horde was like, but from what I hear it had something to do with looking through a spyglass that had the exact same issue.

So it got me thinking, and I’m sure Blizzard have had the same thoughts going into Legion. What can Blizzard do to try and ensure that we don’t have these bottlenecks going into the next expansion? How can they make things go as smoothly as possible, once they know their hardware can handle the mass logging in of thousands of players across multiple realms? With previous expansions, we’ve always been split up to try and even the load of people going to different areas, so that a local spot isn’t overwhelmed with people trying to get one quest objective done (and in their boredom, spam AoE spells and eventually crash the server).


In vanilla, where the hardware was the shakiest, we still had 8 areas where people were split up – every race’s starting zone – that was easily the biggest and best spread we’ve had. But that goes without saying for a fresh new MMO that its starting roots should be the largest, with expansions having less areas for people to flood to. In The Burning Crusade, both the Horde and the Alliance at level 60 flooded into Hellfire Peninsula, with the two factions splitting off into their own questing hubs. Rerollers had the opportunity to also level up the new Draenei or Blood Elves, with particular emphasis on the new Shaman and Paladin classes that were previously unavailable to Alliance and Horde respectively. So here we had 4 bottleneck potential areas for people to get stuck in, yet here again was more simply keeping the servers as a whole up

Going ahead to Wrath of the Lich King we had the first new class introduced to WoW, the Death Knight, where people flocked to en masse. For those at level 70, they had the choice of two entry level zones: Howling Fjord and Borean Tundra, and again split into Horde and Alliance questing areas. This split the playerbase up into 5 local areas. In Cataclysm another two new races were added: the Worgen and Goblins, as well as another two new zones for high level people to go into: Hyjal and Vashj’ir. Although Vashj’ir splits the Alliance and Horde up, in Hyjal (and arguably the more popular zone in hindsight) puts both factions on the same rollercoaster ride through the zone. Again, that’s 5 bottleneck local areas for people to get caught up in. However, new race/class combos were introduced and with the revamp to the 1-60 questing experience, I’m sure the smaller minority spread themselves over the 8 starter zones for other races (for those that didn’t readily race change anyway).


Mists of Pandaria brought us Pandaren that had a notorious quest to set a banner on fire, again with only one player being able to do it at a time, with the banner respawning a while later for the rest of the people in the “queue”. Level 90s were pigeon-holed into the Jade Forest, with the two factions having their own hubs once again. Similar to Cataclysm, there were a few that levelled in the rest of the starting zones, this time to level non-Pandaren Monks. Though with how small the Monk population already was anyway, I doubt that this had much of an impact to the bottlenecks, especially in the Wandering Isle as well as taking several rounds on a helicopter vehicle before even getting an action bar to complete the quest… At least instance servers were stable for at least half an hour to filter people around. Realistically, people were bottle-necked in MoP into 3 areas, and the game definitely felt it as questing technology increased faster than the capabilities of the servers handling it large-scale.

Cue Warlords of Draenor, with its launch bringing sub numbers back up to 10 million again, and Blizzard used technology extremely well with the Assault on the Dark Portal scenario, allowing small groups of people to go into the entry experience in Draenor. Even though I was on a quiet realm anyway, I never saw more than 10 people in one area, though I believe that the scenario split people up into packs of ~50 to ease up on lag. The only problem was that once you left this scenario, you were put into Shadowmoon Valley or Frostfire Ridge, and as I outlined earlier, a similar problem to the Pandaren fire banner quest occurred: only one person at a time could click on the spyglass or flag in order to set up your garrison – this was outside of the problem of having to keep on returning to this solo phased area to update missions, pick up new breadcrumb quests to lead and fly you to new zones and update buildings and work orders… all for it to become an extremely unstable place to be for several days after launch. With no new races (thanks to the art team being busy updating the pre-Cata races) and no new classes, people had very little reason to start at a lower level, so the vast majority of people levelled from level 90 in the two new zones – the smallest range of bottlenecks to date! Because of the problems of filtering people outside of the garrisons, it’s hard to tell where else there were problems, but once people started catching up to the lucky few that got out, you certainly noticed the increase in server lag.


In Legion, Blizzard are trying out a new style of levelling – allowing the zones to level around you while you play through the story of different zones. With current pacing, it’ll take around 2.5 zones to get you to level 100, and it takes around the same amount of time to get through each zone’s full story. Without access to beta, I’m not entirely certain how factions will be split up in each zone, but I know for sure that Stormheim splits Horde and Alliance up, while Val’sharah has a similar Hyjal feel of keeping everyone together in one place (though with less rollercoaster questing as you get to choose which hubs you want to go to). The biggest bottleneck that I fear for however will be that as soon as the expansion launches everyone will start in the new Dalaran. I don’t know if you remember how laggy Wrath of the Lich King’s Dalaran was at the best of times, but if everyone starts off here as soon as the expansion launches in order to get their artifact weapons… Let’s hope that live servers will be more stable as even in videos I’ve seen in alpha, it sometimes took a while for the emissary to lead you to your artifact took a while to even turn up (despite them spamming your chat log that everything is urgent).

Blizzard have done fantastic with everything leading away from it, as long as instance servers can be found in order to do your artifact scenario, as it will lead people away, initially, into 36 different areas, to group up into class order halls (note, not your own personal garrison) of 12 different areas dotted around the world, to then filter out into 4 levelling zones of at least 5 concentrated areas of action. Demon Hunters will be introduced as well, though pre-orders can play their starting experience and play around in HFC between 7.0 and Legion’s launch. It all depends on if Lagaran has become more stable in its new incarnation for Legion’s biggest bottleneck. One of the bigger problems will be for those that are classically fast levellers: they will continue seeing plenty of players levelling with them even if they’re one of the 5 people that are 108+. No more hiding away in Nagrand, Dread Wastes, Twilight Highlands, Icecrown or Netherstorm lag-free and continuing your dominance to server first level X.

Time will tell if Blizzard have finally hit the nail on the head on how to keep a game’s launch stable.

Scaling Up Content


“The technology for scaling up old content is mostly in place now, but the devs haven’t come up with a good rewards scheme for scaled up old dungeons and raids yet. Today, players can go back and solo old content for transmog gear, but if they were going to go back and do it in an appropriately sized group, they would expect something useful other than transmog gear.”

So this quote on MMO-C as a result from the Vanion interview with Chilton and Stockton at Gamescom intrigued me last week. Blizzard have openly admitted that they have technology for scaling up old content, but don’t see the point in implementing it just yet, as they can just blast through the old content without the scaling now for transmog gear. Firstly, it would need to be a toggle-able thing where people that still want to go back and smash the back doors of Molten Core in can still do so.

The vocal minority for sure have certainly asked to be able to raid places like MC or BWL without facerolling through it, but when the reward of nostalgia has taken its toll, what could continue to bring people back to these places? Would pulling C’thun only to wipe 5secs later because people were standing too close together really bring back the fun memories of raiding back in the day? Especially with the LONG corpse run back.


For one, Blizzard could turn the old raids into a pseudo-challenge mode, where not only could the mobs and bosses be brought up to level, but our gear gets taken down to a certain level, like in Challenge Modes or Proving Grounds. You could begin to compete on a world-basis for different raids. Though how long that would keep people interested other than the select few trying to overtake eachother for their own prestige might dwindle quickly.

Achievements are also a good way to get people to go along – but again this is really only a short-term method for keeping people interested, with the only way to increase longevity would be to turn the achievements into something similar to Ulduar in creating multiple achievements for the same bosses. However, Blizz have already mentioned that there is an achievement bloat, so this may not end up being the case.

Another reward might be useful in this difficulty of raiding, is to make ‘timeless’ rewards such as mounts that only drop in this difficulty – though mount collectors tend to prefer content to farm in places that they can solo (at least I know I’d prefer to farm for the ugly stone drake from Stone Core, rather than the awesome Astral Serpent from Elegon, despite the latter being my favourite flying mount model in the game).

Other ‘timeless’ rewards could include heirloom armour and weapons, that you can send to alts and use them as alternatives to the JP/Honor heirlooms, similar to the Garrosh weapons in 5.4, we could see something like the Bulwark of Azzinoth being an alternative tank shield for scaling level 70-90, for example.

Heck, we could see the scaled up raids act as stepping stones for gear, for those that care more about the purples than why they’re there. They could have similar iLvl gear as Flex raids, in the models of the gear that dropped there back when the raid was current content.

Why it shouldn’t happen

As much as it might sound like I’m being a hipster, I actually don’t believe these raids should be scaled up to allow for us to raid these old places again. I’ve had amazing memories of some of them, but they were back in a time when raiding mechanics were very simple. For example, just a single phase of a fight in MoP was pretty much the complexity of entire boss-fights in TBC and vanilla. If I went back to progressing on these fights, the only thing stopping wiping the place out in a few hours on launch would be gear checks. And Metzen help us if we have to force tanks into cloth gear because it has the best Nature resistance/stamina combo for Huhuran.

Which brings me to another point – the main throttle for progression back in those days were stuck behind requiring things like a special cloak for Nefarian, the materials of which you could only gain a limited amount of per week, or more importantly, many bosses had high magical damage that could only be beaten by having high amounts of resistance gear to that school of magic. You may complain about lack of space in your inventory these days, but that’s nothing in comparison to having to hold 4 different sets of resistance gear, on top of things like DPS gear if you were fortunate enough to be allowed non-tanking gear if you wanted to farm Tubers and the like.

As such, something between the two HAS to give, while I would enjoy seeing some of my guildies somehow forget to REMOVE THEIR HANDS FROM THE KEYBOARD on the Shade of Aran fight, either mechanics will have to become more complex, or we add these resistance blocks back into the game again, to ensure that we don’t burn ourselves out too quickly on these encounters.

I’d much prefer to look back on those encounters with rose-tinted glasses and have fond memories of them, rather than have to butcher those fights to suit today’s standard of raiding. What do you think? Should scaled up content be implemented in the game? What rewards would you offer people to make it worthwhile, yet not make it more important than the current tier’s raiding and other PvE content?