It seems that differing animation cycles in the game have resulted in an unforseen side effect: female avatars do less damage then male ones [previously at WCFTWD]. At the time, the Funcom developers said they’d let us know as soon as they could about a fix. Today they let us know, but the news is mixed. The good news is that ’straight damage’ (what they call white damage) has been fixed. Simple autoattack routines use few animations and they’ve all been fixed and pushed to the live servers.
The problem comes in with the much-vaunted realtime combat system, which requires heavy use of combos and unique animations. There are almost a thousand of those animations, and every single one will apparently have to be tweaked by animators and then retuned by a designer.
Wow.
Truth: live teams can’t depend on anything but design time, because artists get moved to other projects and coders are busy fixing shit. Even if you’re so sure that your project will be better!, just play it safe. Design data should always be the final arbiter of anything that may change during live.
I don’t care that your animation was painstakingly handcrafted to look great at a four second cast time — I will change that cast time someday, and I do NOT want to ask for art time to do it.
One of the commenters asks why female characters’ DPS couldn’t be fixed with damage multipliers. That would be a nice quick fix, but I would guess they can’t do it — the ability system would need to be able to check gender, which would require code and data format changes. One would assume that they don’t have a lot of code time to spare! And then it’d be a huge amount of crap for designers to maintain in the future.
I return from ION 2008 and a whirlwind tour of northwest Oregon and southeast Washington. We drove from Seattle to Portland to Salem to the coast and up 101 until somewhere in Washington before returning to Seattle via Olympia … in two days.
The conference itself was pretty good. Executive heavy.
Rather than talks, I mostly went to panels on topics that I already knew well, which made them into interesting exercises of seeing who knew what they were talking about. Sadly, I missed the couple of real talks that I was interested in.
My talk, Design for Longevity, was a little under the radar — I think I should have gone with the original title, “Your Tools Suck.” Because that’s what it’s about: you can get more done with fewer people and less money if you think about workflow up front. But most designers and tools programmers and middleware providers never thought about it too hard, so it takes us forever to make content, and then we lose subscribers the longer we go without updates. Obvious stuff, but people keep making the same mistakes. Here are the slides. I cleaned up the notes so that it could feel more like you were there.
There may be better ways to approach solutions to these problems, but I pointed out the ones that have worked for me — tools that I can build on my own with (comparatively) limited technical skills. And remember that it’s from the perspective of a small-team balance designer, which is what I was doing Back In The Day.
(Aside: I would like to point out that, while presenter view is a great idea, it really sucks that you can’t practice it until you’re hooked up to two monitors. When you create the bulk of your slides on a plane and in a hotel room, you only have one monitor. And hey, would you know it, but clicking on different parts of the screen in presentation mode appears to do different things, but of course you couldn’t know it, because you couldn’t try it! They are apparently taking cues from World of Warcraft raid encounter design … once again, a failure in workflow foresight.)
Two weeks ago, one of the top World of Warcraft raiding guilds announced that they were leaving the game because they were unhappy with the current risk and reward ratio. Despite the fact that the next patch is probably next week, they’ve followed up with a tangent on the trouble with handcrafted content: (copy-pasting quite a bit because the link doesn’t look stable)
Do you want the truth? Blizzard is giving nobody that pays for a subscription to WoW what they deserve. With the amount of money they are making, they should be putting out at least 4 5-mans and 1 10-man per month and 1 25-man every-other month. Not to mention they should probably be adding 1 new battleground each month, and 1 special tournament every other month. […] To be honest, I’m more irritated with this badge loot bullshit […] No, I really really don’t want to do more Kara, more of the same heroics, or anything else that gives badges. That is OLD CONTENT, regardless of the rewards.
Do people that enjoy 5-mans and heroics not enjoy seeing different encounters in different instances? […] . As was said earlier, -everyone- should be getting content. […] It is all being put out in this game slower than it is in AC1, and AC1 has been out for what, 9-10 years, and has less than 1% of the player-base of WoW.
Asheron’s Call 1 has nine years of monthly content updates. They had what, a live team of ten? But they had a famously awesome toolset, and people said they had an obscene retention rate.
Following up on Joe and Damion’s posts, Raph says “a designer who only works inside of data-driven systems will not have a career path and training path to learn systems design unless they are given access to system modifications.”
Well, shit. Somebody tell my boss.
I design “infrastructure systems,” like items and abilities, and I do so all the better because I have years of experience in data management. Five years from now, the live team will be three guys in a cardboard box. If I’ve done my job right, they’ll do more than tweak numbers — they have room for creativity because the systems are well-thought-out and allow for combinations to last for the service’s lifetime. And if they need something the system doesn’t support, they can get an hour’s worth of code and be on their way.
If the player ends up feeling like they’re “wandering through a database,” then the systems were poorly designed.
There’s nothing wrong with design within constraints. That design will be easier to polish, balance, and maintain in the long run, because it’s easier to run a query than to comb through scripts. And we’re failing if we’re not in it for the long run.
Buried away in an article on Zul’Aman, an aside to resurrect an old topic:
We were also told about a philosophical change at Blizzard, namely logical loot. Previously, the loot tables were somewhat random. They’re not going back and fixing years of content, but now, players can expect creatures to drop things that one would expect. If that bad guy is holding a big, shiny axe, then odds are his loot is a big shiny axe. This is a general goal and a monumental exercise in inter-departmental organization* (they need to make sure the artists and loot guys play nice with each other), but one they seem rather committed to. It’s a small touch, but it sure beats wolves coughing up fishing poles!
I’m crying on the inside. I bet the content guys are crying on the outside. I would.
* hahahahaha.
Joe Ludwig’s 10 ways to improve developer efficiency are a great read, especially for all of you folks out there trying to build your first MMO. The problems Joe is avoiding are ones I’ve seen before, and I suspect they’re very common on newbie teams. While he’s talking about making your game easy to build, my god, it also goes a very long way towards making your game maintainable during live service. I can’t overstate the importance of it.
On data and text files, points 2 and 4: it’s also worth noting that it’s an even bigger win if your designers are at all technically savvy. When I was a full-time data manager dealing with text files, I got shit done fast because I took the time to learn Perl. I took the time to learn Perl because I was tired of doing repetitive tasks in text files, and Perl did the job for me. If the files weren’t easy to process, fellow data mangers would be stuck doing those repetitive tasks manually until the end of time. (On a technical note: I understand that Perl isn’t cool anymore, but Python’s regular expression support sucks ass. At least, it did the last time I checked. See also this comic.)
On fast reboots, point 9: this is also valuable in live service. Your servers will crash, you may have to do what we delicately call “scheduled maintenance;” the less time the servers are down, the less time people will have to gunk up your message boards and hate you.
Juice Analytics’ Excel training worksheet is available for download. It covers their previously identified core Excel skills.
I thought I was pretty good with Excel, but a lot of this stuff is new to me. I’m a bad Excel guru. (Nobody tell Damion; he still sends me a lot of Excel questions.) Excel skills are very important for design. That, and data entry. (Tip for would-be designers: go work some shitty temp data entry jobs, and I’ll like you better. It not only improves your implementation skills — faster work, fewer errors — but it exposes you to a lot of bad UIs. Temp data entry jobs are a great way to learn usability.)
The Excel books on my shelf that I’ve read and love:
- Statistical Analysis with Excel for Dummies. No, really! This is a great book. You can read statistics books, which are full of hard math and don’t tell you how to actually implement anything, or you can read Excel books, which tell you how to implement everything but not why you’d want to. This book does both, and it skips the hard math parts.
The Excel books on my shelf that I haven’t read, but think I’ll love:
- Excel Best Practices for Business. It’s about building spreadsheets for usability and efficiency.
- Pivot Table Data Crunching. I just started using pivot tables recently. They’re stupidly powerful and stupidly hard to use. I’m hoping this book will tell me how to build pivot tables that the rest of my team can use.
Subscribe