Sunday, March 16, 2014

[WoW] Should Monolithic Expansions be Obsoleted?

Be careful what you wish for, because sometimes you might just get it.

Blizzard has promised faster content delivery for years, and the player base has been demanding it be rolled out faster as well. Four months for a small tier of five to seven bosses and a quest area is probably a little bit too long (i.e.: Firelands, Trial of the Crusader), and a year for any dearth of content is much too long. To meet that demand, Blizzard actually managed to release patches on a pretty quick cadence for Mists of Pandaria. In fact, they went almost too quickly...

When you look at how long it was between patches, it was 2 months from the release of Mists to Landfall, which contained no raids, but did contain a plethora of new Scenarios, PvP sub-quests, dailies, storyline, gear, etc. From there, it was another 3 months until Throne of the Thunder King. 5 months for 14 bosses in 3 “wings”, which was probably a tad too fast.

Arguably Blizzard could have taken a bit more time to stagger the wings, as I know of many Heroic raiders who felt fatigued after doing 3 races nearly back-to-back, with only a couple weeks between each one, whereas the rest of us were just making headway into the first wing, let alone ready for Heart of Fear or Terrace of Endless Spring.

Two and a half months after that, Escalation came out, offering some new grinds for folks to participate in, and then 4 months after that was Siege of Orgrimmar. Again, about 6 months for 13 bosses (14 if you count Ra-Den). I think our raid, which did get started on ToT right on the button, could have gone an extra month in ToT. We actually downed Lei Shen with a week to spare because we just started locking the raid once we were at boss 8, and did so for a month. Granted, we only raid 3 hours a week, so I don’t expect everyone else to move at our speed, but I think an extra month wouldn’t have hurt too many people.

And now we’re looking at a year of more for Siege of Orgrimmar. That’s definitely way too long. Took us 6 months to get Lei Shen on Normal, and we’re looking at Garrosh in about 6 months as well--we’re locking until we kill him. We’re at about 45 wipes across three weeks so far. I think one more week and we’ll have him. So by that measure, another 6 - 9 months is appalling.

Take a look at the patch timeline below (requires Flash sadly, but mouseover for magic!). Those last two dots on the end? That’s the release window. “Fall 2014”. A total of 12 to 15 months after Siege of Orgrimmar dropped. If they’re hedging their bets, that hedge overwhelmingly needs trimming.


Guilds and raids fall apart in that empty space between expansions. With nothing new to do, people move on to other things, and inevitably some don’t come back for the next expansion when it drops. It’s pretty rough. So what’s a company to do to reduce that time?

If I recall correctly, the developer team at Blizzard is really split into two, and for MoP, staggered releases. Team 1 was working on Patch 5.1 while Team 2 was working on Patch 5.2. Once 5.1 was done, Team 1 started on 5.3, and so on. Ghostcrawler mentioned that at some point, I just don’t have the reference handy. And that’s a pretty damn smart way to go about things, up until you have the monolithic Expansion, where everything changes.

So the easy way they could have reduced the end-game time was let patches sit fallow for an extra month. Finish up 5.1, sit on it for an extra month while you move on, and then release it. Repeat for each patch, and we’d just be starting Siege of Orgrimmar in November or December rather than September of last year. Granted, by the end of the expansion they’d have SoO done months in advance, but is that a bad thing to get ahead of the curve for content rather than releasing everything the second it’s ready?

The other half of the story is the monolithic expansion. That’s a lot of work. When you look at the staggered approach they took during Mists, it starts to look a fair bit like Agile development, whereas monolithic expansions are clearly ye olde Waterfall development. Plan a bunch of stuff in advance, build it all, release it in one go. And for something like the Squish, that’s probably about the only way you could go about doing it rationally.

There’s also the bonus of getting an extra influx of cash thanks to selling the expansion, which any business would be crazy to let go of as long as it’s the expectation. When you’re talking about that much content in a single go, it really is a new game, too. I mean, look at how much goes into a single WoW expansion, and look at how much goes into FIFA 2014, or Gears of War 3. They’re not building a new engine, but they are upgrading it, and providing an immense amount of new content, so why shouldn’t they get paid for it?

But when you look at having to release more content and be faster at it, monolithic expansion development flies directly in the face of that, as evidenced by the fact that it takes a year and a half of development time to build said expansion. So should the monolithic expansion be obsoleted, at the risk of not being able to hit the reset button for things like the Squish, or at the risk of not releasing an entire continent at a time, but instead parceling out zones every 3 months? Is that really a better model?

Monolithic expansions are too lucrative for companies to give them up, but I wonder how long it will be before MMOs bleed out their populations for a game that updates faster, or just get bored with the current pace of delivery. Or are things just peachy and while people grumble, the status quo is fine and continue indefinitely?
 

 
Confidential to Mei Francis: STAHP.
#BadDesign, #Expansion, #Waiting, #WarlordsOfDraenor, #WoW

14 comments:

  1. And before anybody says, "Just hire more people!" that doesn't necessarily work for two reasons:
    1) They've already HIRED more people
    2) The mythical man-month. People need time to come up to speed, the more people you have, the more relationships you need to manage and the more management needs to organize around all of those people checking all of those things into the same product. Production does not scale linearly.

    ReplyDelete
    Replies
    1. I still don't see why that extra expansion money can't be put into a slow build up of additional employees, at least to start covering some art assets and other things.

      No magic bullet, but it makes perfect sense to me to have more people ready to go, even if just temporarily!

      Delete
    2. You still get the problem of too many chefs in the kitchen, though. There's only so many more people you can add to a project before the management aspect becomes, well, unmanageable. That's not to say Blizzard's hit that yet, but given they've mentioned that they've doubled the staff, I'd be surprised if their efficiency hasn't taken a hit for it.

      Delete
  2. I agree that the content could be better-paced. We had another one back during ICC, and it was terrible -- an overtuned fight (on Heroic, anyway) with mediocre drops and very little storyline justification. Blizzard has admitted that they don't want to do another Ruby Sanctum, however, and that means these large gaps become even more problematic.

    Because, as you say: what else is there to do? People can point to Heroic mode and tell your raid team to get cracking, but not every raid team is capable of that level of reflexes, coordination, and -- let's admit it -- masochism. It's also largely the same content, with rewards that are dialed up a few item levels but otherwise identical.

    There are alts, but anyone dedicated enough to do Normal raiding probably has a stable of those by now. Pet battles can be a nice diversion, but again, anyone playing this long likely has collected plenty of companions and mounts already. And daily hubs are usually something one endures as a means to an end (reputation grinds or specific rewards) and not something people generally do because they enjoy the gating itself.

    On the other hand, WoW has a vast amount of older content that lies largely unused. Plenty of people like going back to solo older raids, but there could be incentives to group up for it: make Dungeon Finder satchels available as rewards for everyone, as one example. Make it a weekly 5-man quest for a random raid, similar to the random dungeon/raid boss quests available during BC/Wrath, and you've got a new activity that requires very little developer time to create and manage.

    There are other possibilities if they want to spend a little bit more developer time. Imagine Challenge Mode scaling applied to those older raids, with a few recolored raid sets available for transmogrification. The Caverns of Time could actually have a storyline built around sending you to these older areas, since blah blah timeline corruption blah blah alternate universe blah blah blah. And they could always steal ideas from other games: Instant Adventure from Rift (level scaling + Kill X of Y quests + regular teleports) is something I've been waiting for another MMO to copy. I'd love an Instant Adventure in Elwynn Forest with 20 people scaled to level 10, raiding Hogger. :P

    But all of these ideas (and more) have been on their Suggestion Forums for years. The real question is, why do they avoid reusing old content? Apparently, it only suits them when they run out of ideas for heroic dungeons.

    --Taaniel

    ReplyDelete
    Replies
    1. To be fair, the heroic dungeon remakes don't save them THAT much time or resources. They still redo all the textures, and the polygonal geometry needs upscaling (downscaling is easy, upscaling not always); they redo all the bosses, and add new voice acting, etc. Really the only thing they get to keep is the general theme and layout, which does save some time. Just less than what you're joking about.

      As much as I disliked Ruby Sanctum, it was better than nothing. Even if I never raided it myself, it kept more of the playerbase engaged, and therefore more people doing other activities. There's a lot to be said for at least the illusion of a large population, even if most people are checked out besides raiding (because the secondary activities, like the economy, continue to support those raiders).

      Agree with the other ideas, though. Blizzard *has* the scaling tech. Heck, there was someone who accidentally triggered it in Stratholme once a while back, so it's not like they're THAT far from that solution. Who knows?

      Delete
  3. I think it would help if Blizzard had a better grasp on what happens between the last content patch and the next expansion.

    Assuming your Group 1 and Group 2 alternating content theory is correct (and I think I saw likewise at some point), the various intermediate content patches seem to indicate that it takes a single group 3-6 months to release an intermediate content patch. I'm not sure how Blizzard expects that same model to turn out a full expansion in 6 months. Even if an expansion is only 4x the content of a content patch (it's probably closer to 10x), the math doesn't work.

    More realistically, they'd need to have 3 content patches basically banked and ready to go while releasing both groups to work on the expansion 12-18 months in advance and only going back to tweak the patches as required. Maybe keep small groups separate to handle that with the bulk moving on to the next expansion.

    Either way, something like that needs to happen. I found it a bit disturbing that Blizzard were having "What would YOU like to see the next expansion?" polls as of a few months ago... that type of stuff should have been baked 6+ months ago. They seem to think they're a lot more agile than they ever turn out to be.

    Blizzard has long held out expansions as the time for major overhauls... "We'd like to see that happen, too, but that's something best handled in a new expansion, not mid-expansion", etc. That type of thinking would have to fundamentally change. I actually LIKE the current expansion paradigm, I like that it gives players the ability to step away and come back at obvious times. A constant flow of mid-size content would be nice for many but likely troublesome for others... not everyone plays the same amount or constantly from week to week or month to month... there is something to be said for a release schedule that has some obvious best times to take a breather, all the outrage notwithstanding.

    ReplyDelete
    Replies
    1. A couple of reasons I could see as to why Blizzard wouldn't want to make patches that far ahead of time are 1) leaks, and 2) not being able to respond to feedback as quickly. People railed against the dailies at MoP launch. They didn't get to apply as much of that learning as they'd have liked for Landfall. Not until Throne of Thunder did they get to. Building full patches well ahead of time would make that even longer.

      As Taaniel above suggests, I think a Ruby Sanctum is better than nothing. At least something to give the playerbase a bit more to do, even if it's not that big.

      I agree with you about the expansion model though. While I question the model given Blizzard's stated goals of a faster release cadence, I think I prefer the Expansion model if only because of all the benefits. A meaty chunk of coherent content; some downtime (though, not too much!); the media blitz and excitement that goes along with it (the announcement is half the fun!); and the chance for the reset, like with the Squish.

      Delete
    2. Yeah, I'm not saying that's a good idea or even possible, just that from a logistical standpoint I think that's how they'd have to do it in order to flip an expansion that quickly after the last content patch. Or they'd have to have an entirely different team working on the bulk of the expansion and only bring in the other teams in the last 3-6 months to polishing.

      I dunno, I can't speak for anyone else but my raid got a kill on normal our first night in for Ruby Sanctum and tried heroic the next week but felt the main mechanic was too gimicky to be worth fighting against... and, frankly, we knew it was just meant to be transition content so there wasn't much motivation. It was basically a complete waste for us and I don't know anyone personally who ever got more than one kill in there while it was current, the other raids I knew at the time either couldn't kill it or killed it once and stopped.

      Yeah, I have no real issue with what's coming and I'm not sure if a faster expansion release would change much. Raids are collapsing all around me NOW, have been since early Feb... only the absolute most optimistic guesses had an early year expansion target. I was June, myself, but even that would leave 3-5 months of not raiding.

      I think an issue that doesn't get enough attention is gear. One of the reasons why players tend to stick with mid-tier raids through their full life is because there's usually an upgrade or two still to get no matter how many times you've done it and that gear will help you in the next tier when it's released. There's a fundamental benefit to continuing to run it, both for yourself (while you need gear) and for your raid (when you don't but others still do, them getting gear helps you).

      That is basically nullified when it comes to the last tier of an expansion... once you've "completed" it, whether it means a normal completion, a heroic completion or just hitting your raid's particular skill/execution wall, there isn't any real incentive to keep running. Adding a Ruby Sanctum doesn't really change that, nobody's gearing up for that type of content.

      Ultimately, in the last tier of an expansion, players will only raid until the will to not raid exceeds the will to raid and it only takes 1 or 2 players to bail before things go south... there there are usually 1 or 2 players who pull the ripcord pretty early.

      A faster expansion schedule will reduce the time between collapse an expansion but it won't eliminate it... and functionally does it much matter whether the break is 5 months or 9 months? I don't think it does, not really, either players will come back or they won't.

      It's a bit different for non-raiders or folks who'll stay engaged with the game regardless (including me, I don't expect to take a break although I do expect my playtime will shorten as time goes on), shorter would in most cases probably be better... but those players aren't living the patch-to-patch life that raiders do so I don't think it's all that critical there either.

      Plus, there's more time to get Challenge Modes done before 6.0 hits than if it was a June release. :)

      Delete
  4. "Again, about 6 months for 13 bosses (14 if you count Ra-Den)"

    *cough* 12 and 13 *cough*

    Overall though, yeah, I agree. And my guild literally killed Heroic Sha of Fear about 4 hours before 5.2 hit -- that was an insanely short tier. Throne of Thunder could definitely have lasted a bit longer too.

    ReplyDelete
    Replies
    1. Derp. I'm a Pure Mathematics Minor. I can't count or calculate. Unless the math involves letters, I'm hosed :P

      But good to hear the opinion I've heard before echoed here. I suppose it's interesting to see just how fast Blizzard can go when they put their engineering efforts directly to it.

      Delete
  5. Frankly, I think expansions are too important, especially to subscription model games. Beyond revenue, they generate a ton of buzz, give you something concrete to market, give you a chance to truly change your game around, and do a great job of getting older players interested again.

    I'm vaguely considering Warlords just because new expansion launch is honestly the best time to play World of Warcraft!

    ReplyDelete
    Replies
    1. Furthermore, I miss expansions across gaming. DLC doesn't deliver quite the same punch. Every time a new Civilization expansion drops, I treat it like a brand new game. It's awesome!

      Delete
    2. The media blitz is a huge benefit. Existing players get the excitement too! And yeah, agreed that DLC isn't quite the same, even if it is a mini-expansion of sorts. Also, shush about Civ, or I'll have to go play more, and I can't afford that with all the OTHER games out :P

      Delete