On Bottlenecks and the Start of Expansion Rush

417673-for-the-alliance

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).

Classic_Eastern_Kingdoms_loading_screen

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).

WoWScrnShot_092512_004200

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.

LegionLogo_lessglow

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.

Do we need another round of Connected Realms?

Connected_Realms

Before I begin, let me make a disclaimer that numbers for populations are according to Realmpop, which retrieves data from Blizzard’s Armory. While it is fairly accurate, it is not Blizzard’s internal data. In addition, the numbers that I bring up for the populations themselves are for characters, not players. I’m sure plenty of us have more than one character per realm, so numbers may be inflated somewhat. I’m also talking in context of EU-English realms for the most part.

One of the more welcomed features of the MoP/WoD transition was the introduction of new hardware from Blizzard’s part – the merging of realms in order to aim to boost realm populations to be at a stage where servers didn’t feel like ghost towns. Most realms that got this treatment were paired up with a single other realm, but some were connected to make a conglomerate of up to 10 realms.

How Blizzard decided who should be mixed with who is still a slight mystery as that large merger brought the meta-realm up to the 7th biggest realm in the EU region, according to Realmpop, with the only realms larger than that one are the behemoths that are single realms anyway: Outland (~537k), Stormscale (~455k), Silvermoon (~414k), Draenor (~380k), Kazzak (~379k) and Sylvanas (~370k). The 10 meta-realm sits closely behind Sylvanas at ~369k characters. Thanks to faction imbalance, Outland’s Alliance population exceeds Stormscale’s entire realm pop.

So if they can merge 10 realms together to rival the lower end of the behemoth realms, why are the rest of us floating in the 100-250k character range? That’s a big difference to see the majority of realms at that range, then to have scattered outliers between 250-350k, and then the select few mentioned above. My personal realm is in the vicinity of ~174k characters, with a ~2.2:1 Alliance:Horde bias. Our Horde population still outmatches Silvermoon’s Hordeside thanks to their hysterical faction imbalance, but for recruitment for mythic it becomes incredibly difficult to find new blood for Blizzard’s waning patch 6.2. At peak times the server itself barely breaches medium population, but for most of the day we’re low pop – the Hordeside being very reminiscent of pre-connections.

WoWScrnShot_091413_195259

The main problem is that it still feels quietWhether or not those 174k characters are active or not, the biggest group of people I see are in front of Hellfire Citadel before a typical raid start time – but I doubt I could even fill a 40man group with those that hang out there. The above screenshot was when I hopped onto Kazzak to get a Huolon kill back in MoP, and I was told that was an average amount of people to see at places like the Shrines, in front of SoO before raid start, and of course farming rares on Timeless Isle. In comparison, Doomhammer-Turalyon makes you feel the entire world can be your garrison sometimes with how few people you see walking around.

It’s not a good feeling, and I would heavily suggest for Blizzard to continue the trend of merging realms again in the future. The main choke-points of larger realms are when new content is released, especially expansions, and people are forced to either wait in queue to get online or once they are online, are confronted with tons of lag and competition of mobs and doodads.

Four Realms to Rule Them All

One thing Blizzard could do is to simply merge all realms of the same type to one server. All PvE realms would be together, all PvP realms in one group, RP realms and RP-PvP realms – though I believe the latter is already the case.

With regards to faction balance, Blizzard have already mentioned that according to their internal statistics, Horde tend to venture toward PvP servers while Alliance populate PvE. With mega servers, you’d be looking at PvE realms having nearly 2.8m characters on Alliance side while Horde sits just under 1.7m characters. For PvP realms, Alliance become the underdogs with nearly 2.7m characters against the Horde’s 3.6m. What would happen to the PvP realms if this were the case? Many people prefer to be on the dominant faction when world PvP is concerned, so would we see mass exodus from the Alliance to go to the PvE realms (especially where places like Outland are concerned when they’ve been so used to outnumbering the Horde by so much).

princessofazeroth

One interesting statistic I saw was when I looked at the RP realms however: Currently there are 3 RP realms: Argent Dawn on its own, Darkmoon Faire/Earthen Ring and Moonglade/Steamwheedle Cartel/The Sha’tar. If they merged them all together, Argent Dawn would still make up just over half of the population on its own.

To look at a similar game that’s already gone through this, Wildstar originally started with separate realms. As the initial hype left along with more and more players, Carbine eventually merged the realms into a mega server and the Exiles far outweighed the Dominion. When they further created two servers – one for PvE and one for PvP – most people ended up moving to the PvE one and left the PvP one as a wasteland. RP-PVP realms in WoW seem to be unique to this however, as they are already all connected and their balance is ~162k Horde to 171k Alliance. Could the mega server work for WoW then?

As I mentioned earlier, the biggest downfall of merging everyone together would be in major content patches or even just at prime time when you’re confronted with a queue screen. When you do get online, you probably won’t be able to get anywhere as either the lag will be insane or the things you need will be missing or dead. Blizzard has made improvements to this with areas like the Tanaan starting zone at level 90, where packets of people were phased at a time, but once choke-points like setting up your garrison occurred, that’s where things broke down. Majorly.

Aim for High Population per Realm

This is where Blizzard would probably know more on the matter than what you or I could. Because of the data that we know, we can’t tell for sure how many of those characters belong to a single player, and we don’t even know how active or what activities they partake in. I largely take part in the crowd most influenced by the Massively Multiplayer side of an MMO, the raiding side, but constantly each week we’re struggling to fill out our mythic raid team to its full 20, and the 15-18 of us are getting tired of heroic HFC but feel like it’s an uphill battle to find people capable of mythic.

13524_world_of_warcraft_the_horde

There’s only 3 Horde guilds on our realm that’s progressing in mythic HFC, and the only two have plenty of people happy in their current guild, and we get on well with them enough to not want to poach from them, and the rest of the Horde’s guilds have yet to kill Mannoroth on heroic, let alone Archimonde. 6.2.3 will help a ton with us being able to sell the Grove Warden mount, and to be able to finally regularly raid mythic with crossrealm raiding from the patch’s launch, but in that space of time where we can’t field enough players for mythic from our own realm, it feels utterly crippling.

So instead of the mega realms suggested previously considering even the realms with 350k+ characters are struggling with prime time play, perhaps merge realms together to aim for around 300k characters. This would be easy for the sub 150k realms that are already merged, but the challenge would be for those between the 150k-300k mark. As it stands you could safely merge the twosome and threesome RP realms together from above, and keep Argent Dawn as its own realm and call it a day, but what about for all the PvP and PvE realms?

Faction balance would always be an issue: for PvE realms at least out of 21 realm groups only 3 have more Horde than Alliance, but on a PvE realm that’s not much of an issue anyway – I’d just like to have more people to play with, not necessarily caring much about if the other faction has a monstrous size in comparison. I’m not going to be activating PvP much for it to matter! Ignoring realms with over 250k characters, Doomhammer/Turalyon is the 3rd quietest Hordeside PvE realm according to Realmpop, so connecting us with the Horde 3rd busiest will merge us with the Kilrogg/Nagrand/Runetotem group of realms, bringing us to a huge 413k mix of realms – around the business of Silvermoon but hopefully with a ratio of ~2:1 Alliance/Horde (instead of 22.7:1). I don’t know if that will help at all, especially considering even with the quiet/busy merge brings the total population to those of problem prime time realms, but during the quiet drought between patches it will definitely help a lot.

For PvP realms, there are only 5 realm groups that have less than 250k character, but none under 200k. I feel that Blizzard probably did as best as they could with those realms, but their problem is that all of them are Horde dominated. The Alliance are going to feel very quiet and farmed on those realms, and any mergers are going to bring them to Stormscale’s level of business. I’d probably suggest to merge Al’Akir/Skullcrusher/Xavius with Bladefist/Frostwhisper/Zenedar to make a 428k size realm group, and let the smallest realm be a size of 214k until Blizzard’s engineers can update the servers to handle more people on a single realm. Of course, staggering these mergers is always going to be rocky until we can turn into a mega realm described above, as you can’t split these realms up later on down the line – it’d be like saying the Undead are removed from the Horde and have to fend for themselves!

league-of-legends-logo-images-photos-0322211237

The Bottom Line

The MMO genre is a “dying breed”, going the way of RTS’s, construction games and alternative sports games. With the way that people generally want quick games that they don’t have spend hours pouring into (unless they can safely leave and re-enter again later), it’s easy to see why MOBAs, FPSs, CCGs and casual games are rising to the top of Twitch consistently. It’s also why MMOs that people have said will be “WoW Killers” end up not. WoW is still a great game and is one of the best at what it does in the genre, it’s just that genre isn’t where gamers are going any more.

With these gamers leaving the genre, it’s imperative for those of us left behind to still feel that MM part of MMORPG. Blizzard may not want to spend the big bucks to improve the servers on a game they know is in the twilight of its life, and they may want to invest more money into Hearthstone, Heroes or Overwatch, or even into their new esports division. One thing that could bring WoW back into the limelight will be the changes that Legion bring to WoW PvP, but until then we need to have those empty spots on our realms, on our raid teams filled.

Patch 6.2.3 Hype!

623 header

Blizzard suckerpunched us not too long ago with news of future updates! Hoping for some teasers of Legion stuff as soon as I saw that wall of text, it ended up being content between now and 7.0. Well, content in a loose sense – we’re not getting a Ruby Sanctum on our hands here.

While it seems that Blizzard have learned their lessons from Warlords about letting up too much information too quickly on the next expansion and ending up going back on their word (e.g. Farahlon, Bladespire Fortress/Temple of Karabor), it’s promising to know that what we saw at gamescom is likely going to stick around with only a few minor tweaks. We’ll know more about Legion in a little over 3 weeks as of this post, but until then we have changes in store for Warlords that should keep people logging in for the next 6 months or more.

Pit of Saron

Following on from the success of 6.2’s Timewalking dungeons, Blizzard are now expanding the events to include a Cataclysm event, in addition to expanding the TBC with Magister’s Terrace and WLK with Pit of Saron. I’d almost be erring on the side of caution for CC with MgT coming to the tables if it’s anything like the live version back in TBC. Then again, there are Hunters and Warriors out there doing 15-20k AoE DPS thanks to a mixture of legendaries, set bonuses, gems and so on.

The list for the Cataclysm set of dungeons are Grim Batol, Stonecore, Lost City of Tol’vir, The Vortex Pinnacle, Throne of Tides, and End Time. No ZG or ZA just yet, nor the revitalising (again) of Deadmines or SFK, though I have to say the one I’m most surprised about not seeing on the list is Well of Eternity, considering the links into Legion with it. They may extend it out to that during the next expansion when they introduce the Mists dungeons, we’ll see.

In addition, any of the timewalking bosses will have a chance to drop the mount in the header image – the Infinite Timereaver! Dragons have always been awesome mounts, and the Infinite have always been a really interesting faction of them.

Hellfire Citadel

This is where I find it interesting they’re calling this patch 6.2.3 instead of 6.3 instead. Although there’s no new groundbreaking content like new raids or new systems or even a selfie stick/social media integration, Blizzard are treating this as a “next content patch” by allowing mythic Hellfire Citadel to be raided cross-realm! Personally for me and my guild, this is an insane change that will help us to progress and see the end of mythic raiding. Although we’re stepping our toes into it currently, on a low-pop merged realm we’re struggling to get a healthy bench of people to cover for emergencies/real life issues/etc.

With people still thinking about leaving the game as well as being cynical about any carrots on a stick that Blizzard are aiming in front of us, this could open up recruitment much better than a simple advertisement on Wowprogress, Reddit, MMO-C or even trade chat ever could. People come and people go – for servers such as Outland-EU and Illidan-US it isn’t too much of an issue as long as you’re the right faction – though for those that are thinking of leaving the busy servers in preparation for the queue times when Legion hits, this could be a blessing for us.

nerzhuldraenor

Still within the realm of mythic content, though this time for 5mans, Blizzard have incentivised the small group demographic by allowing 725 loot to drop. All warforged drops from mythic dungeons are no longer simple +20 iLvls, but have a chance to be 690, 695, etc. All the way up to 725. In addition, item upgrade levels are being re-introduced from their concept back in Mists, alongside valor points! Up to 10 iLvl increases at 2 upgrade levels, players can once again earn valor points and spend them on whichever pieces they want to upgrade! This’ll be easier to get new people up to the right gear level to fill in those last few spots on your 20man team, or to get alts up to scratch if you need a different role. Finally, there’s also a mysterious new heirloom trinket that has a chance to drop that goes up to level 110. I’ll laugh if it ends up being a fel void tentacle trinket, but at the same time cackle maniacally if it’s a 1-110 trinket instead of just from 100. Double hentai trinkets? Hell yeah!

Mythic 5mans too hardcore for you too? Blizzard still has prizes in store for you with the baleful items that drop in Tanaan Jungle. They now scale between 655-695 to get people to raid heroic HFC that much easier, and raid finder, heroic dungeons and bonus events will also give you those valor points to upgrade them another 10 iLvls. PvP players will also see Warlords Arena Season 3 come along with this patch – if it coincides with the patch launch, we’ve got a date for it being between Nov 24-Jan 19, going by Blizzard’s 22-30 week cycle for seasons.

79WMBVYOFME71444790679062

Finally, Ghostcrawler’s final promise to the players is finally entering the World of Warcraft! The Grove Warden, previously thought to be a store mount, is now going to be available through killing heroic/mythic Archimonde and doing a quest chain following. It is such an amazing match with my Druid’s challenge mode set, I’m not sure I’ll be switching from it for a while, even if there are going to be hundreds of people also riding it. We finally get our moose!

TL;DR: This is no more than a carrot on a stick to get people to stay subbed between now and Legion, but for those that are calling it quits… Well we’ve now got ways to still raid mythic, get our alts to mythic level and maybe get a shot at that Felsteel Annihilator before 7.0 comes out and makes it super rare.

For the full post by Blizzard, check out the link here.


EDIT: The trinkets have been datamined, as have the valor costs/rewards! There’s 5 trinkets in total, levelling from 100-110, and have demon-focused procs to help with our levelling experience: different trinkets for tanking, healing, agi/str/int DPS. It says they’re unique-equipped, but I wonder if I can use both the tanking and the agility one for my bear… Will be interesting to find out!

As for the valor costs and rewards, MMO-C have found out:

  • Each 5 item level upgrade costs 250 Valor. Only HFC/Baleful gear can be upgraded right now.
  • Random dungeons reward 100 Valor.
  • The weekly Bonus Event quest rewards 500 Valor.
  • The weekly Raid Finder run rewards 75/150 Valor.