Friday, January 8, 2010

A Treatise on Oculus, Part 4: The Solution



Back to Part 3

So, now we have established that Occulus is different than every other Wrath heroic, and in fact 99.9% of the rest of the game, and we’ve established that, no matter what your personal feelings about the instance, these differences have caused negative consequences, especially when combined with how the Dungeon Finder works. First, I’m going to suggest three simple things to change in Occulus that I believe will “fix” it and for the most part and eliminate the negative impacts that I listed earlier. Then, I will suggest a small set of low-impact changes to the Dungeon Finder itself to put the reality of the player experience more in line with what I stated earlier to be the developers’ primary intent with the system. Finally, I'm going to give some suggestions for the Blizzard developers on how they might avoid similar mistakes in the future.



How to fix Oculus:

1) Change the spawn distribution of the enemy “trash” drakes. Instead of having them spread out in singles throughout the sky, place them in groups of 3 as “packs”. One of these packs will auto-aggro as you approach a robot platform during the second part of the instance, meaning that a normal group will face only 3 of these groups. A few more packs can be around the instance for flavor, as long as they aren’t too close together or likely to be directly in the group’s path.

2) Greatly decrease both the time the final boss takes and his difficulty. Eregos is the only boss in the instance you fight while on drakeback, and as such he should be nerfed, hard. Blizzard already did this to some degree, but I don’t think it was enough. He should only summon drakes twice, and should only banish himself and send sparks after you once. Then just cut about 20% of his health off. Now he’s more on par with the difficulty and time commitment of a boss in any other heroic.

Yes, Eregos can be very easy if you have a great group. But we aren’t talking about fixing the instance for an experienced group of guildies. We are talking about fixing it for random dungeon finder pugs, which are more likely to have ungemmed pvp-geared DKs who have never seen the instance before than they are to have good players.

3) Visual guides. Probably the biggest problem with Oculus compared to other heroics is how easy it is to get lost. There are no hallways, and shockingly few ways for someone who hasn’t been there, or even someone who gets lost, to tell where they are or where they are supposed to go, and this problem is compounded by the vertical nature of the instance. I know that it’s going to sound like I’m rebuilding this instance for a short bus full of blind people (oh wait I am), but when you think about the linear hallways of every single other heroic, you have to admit it makes sense. There are a few options for visual cues that will help people find their way:

  • Lines of light: This would be the best option, but I’m not 100% sure the tech is there on Blizzard’s end. When a boss or pack dies, a beam of light appears leading directly to the next objective. So, you kill the first boss, and a beam of light arcs up from that platform leading to one of the platforms above you with robots on it. Once you beat the last enemy on one of those platforms, another beam of light appears leading to the next one, etc.
  • Dragon Guide: similar to the beams of light. Once you beat the first boss, a friendly drake NPC leashes to your party and helps you fight. It follows a script as each enemy pack dies, leading you to the next one.
  • Platform coloration: Each “level” of the instance (sections between each boss) should be a different color, first of all. Then, ideally, the next platform you need to get to should turn a pulsing bright red until you get there.
Basically any strong, noticeable visual cue will work, but I think these would be the most effective.



Now that Oculus is fixed, let’s move on to how to improve the Dungeon Finder. I do not know if these are technically possible, but they are improvements that directly address the Oculus problem:

1) Weight the instances based on how many people have entered (not completed) the instance that day on that battlegroup. If tons of people are leaving Oculus, it’s less likely to show up for others. This is preferable to the mess of people leaving as soon as the loading screen is done.

2) If someone leaves in first 2 minutes, your group gets teleported out of the dungeon and inserted back into the front of the Q to replace those who left. The next dungeon will be assigned randomly. You can only do this once before you’ll have to leave Q and wait out the debuff if you want to get a different random dungeon.



Future lessons for Blizzard:

Thinking back to the early coverage of Wrath, Oculus was one of the most talked-about and anticipated pieces of content (oh, how things have changed!). It showcased the vehicle tech and promised an era of new and exciting instance design (that was only partially realized). In short, Oculus was actually a good idea. I’m glad they went for it. They just blew the execution spectacularly.

So I definitely think Blizzard should keep trying new things with instances. But how can they avoid a repeat of the failure of Oculus?

For one thing: don’t throw so much new and different stuff at us at once. If you want to add vehicles, then make sure the abilities are simple, the paths are clear and linear, the gear scaling is strong, and the enemies behave in traditional ways. If you want an open-world instance with scattered single enemies, don’t make people fly an unfamiliar drake with unfamiliar abilities in a 3-D plane at the same time. Before I knew what Oculus was actually like, I imagined it would be much like Nexus, only with a hole in the middle that you had to fly your drakes through. That would have worked much better. You clear a linear set of hallways, then do a brief drake flight across a chasm to reach the next clearly-marked linear hallway to dismount into.

Another thing to keep in mind: people like to play their characters much more than they like vehicles. If more than 25% of the instance is vehicle-based, you screwed up. Also, take a lesson from Ulduar and do not force players to start an instance in a vehicle first thing. Nothing was more annoying to me than having to do Flame Leviathan at the beginning of a raid week when I was just itching to play my character and get further into the instance.

Also, keep in mind how the instance will compare to other instances. If you add new mechanics, vehicles, or flying (or swimming! Hint hint!), compensate for it somehow. Make the instance easier or shorter than the more traditional ones. Or add an easy “freebie” boss with good loot. Plus, fully account for scaling from the very beginning. We’ve learned from Oculus that once an instance gets a bad reputation, it can be nearly impossible to improve that image in the eyes of the playerbase. By the time you start hearing complaints about how Oculus 2.0 doesn’t scale so geared people don’t want to go there, it’s already too late. Think of it ahead of time.

Finally, and by far most importantly: hire more fucking people to develop and test instances. There is no goddamn excuse for Oculus to be released the way it was, to be as ill-thought-out as it was, and to remain so shitty for so long. There is not enough content in this game. Blizzard has enabled us to consume content more quickly and accessibly than ever before, but development has not even begun to keep pace despite the literal fortune of cash we send them every month (which gets siphoned directly into Bobby Kotick’s pockets instead, awesome). Get more people on it, and get more testing done by intelligent people who know their shit. I could name 20 people in the community right now who would accept peanuts as pay (literally) if they got to go to Blizzard every day and test instances. Smart people.

People who could have made something better than Oculus with their eyes closed and using nothing but their nipples to type in the lines of code.*

Thanks for reading, and may all your randoms be VH.



*Blizzard devs are smart and skilled, I’m just suggesting they are understaffed. They had too much time pressure and too few resources to get Wrath done, and that’s why Oculus is so bad. Not because they are bad developers or people.

1 comment:

Jong said...

Just got to your "Why I hate Dungeon Finder Now" post.

I hear you. Your dk main can probably make up for a lot of deficiency in technical skill/gear in the party. It feels impossible to make up for lack of people skill/etiquette.