Game: World of Warcraft: The Burning Crusade
Game Element: Netherspite, a boss in Karazhan
Discipline: Content Design
I’ve had more people ask me to do this post-mortem than any other in the history of every project I worked on at Blizzard.
What is it about this encounter that struck people so poignantly that 7 years later, I still hear about it?
Much anticipated, let’s talk about my first novelty boss fight – aka one that breaks the normal rules of the game at a deep level.
Let’s explore what made this boss so intriguing to people.
Well, it all started with a drawing.
Nether Dragons
Scott: Good morning gentlemen.
Joe/Me: *mumbles/grunts*
Scott: Chipper, today.
Me: *sleepy 9 am glare*
Scott: I have just the thing to wake you up. TADA
Scott showed us a drawing of a very badass serpentine creature with energy flowing out in every direction.
Me: Oooh, cool!
Scott: Yessss. And this is the creature Metzen and I think we should use in Medivh’s Observatory.
We had left the room empty for a little while now, as one of the optional encounters, it was less important to be filled out early on in production.
Joe: What do you want it to do?
Scott: I’m not sure yet. Let’s discuss this at the next meeting.
At the meeting the next day…
Geoff: Nether dragons look pretty cool. If the story goes that they existed in Outland though, how on earth did they get here.
Scott: Well, the current story is that the empty space outside of Karazhan is open to the Twisting Nether – which means these creatures just wander the empty space and fly in there.
Geoff: That’s still a little weird. What if it could open portals and it came in through one of those tears instead.
Me: Oh that’s interesting, maybe we could use it in the actual fight abilities.
Geoff: Ooh, yeah, like portal open up and blast him with energy. Then spin around the room.
Me: Uh sure…. who’s working on this one again?
Scott: Hmmm… Geoff, you seem pretty excited. Do you want to take it on?
Geoff: Sure, but that would mean giving up the demon prince at the end.
Scott: Erm, no, nevermind then. I really want you to finish that one first. Joe?
Joe: I’m still working on the sacrifice mechanic for the Satyr.
Scott: Well, Alex, how’s Shade?
Me: Done.
Scott: *skeptical* Done?
Me: Well… I have to have more QA playtests… but I mean yeah, he works.
Scott: Alright, I guess the design for that fight was pretty tight. Go ahead and work with Geoff to come up with the kit for this one.
Me: Okay!
Nether Planning
Geoff: Yeah, so I am pretty busy. What do you think about those portals?
Me: It’s a cool idea. Maybe the beams could come out and infuse you with power.
Geoff: Oh yeah!
What if when the beams hit him he powers up and you step into the beams you steal the power for yourself.
Me: That sound awesome!
Geoff: Go ahead and be crazy with this one.
Make the players like veritable gods when they have the beam on, hahaha.
Me: Alright!!
I hurriedly rushed off to create the kit.
Netherspite Version 1:
I spent the majority of the next week crafting the scripts that made the portals “work”.
At this point in WoW’s development, the one gameplay programmer, Sam Lantinga, was in high demand… and it was nearly impossible to get new game features.
Thus all of the new features were generally forcibly scripted into the game with gratuitous amounts of Lua.
This one was no exception.
After a week:
|
| Creature Model |
Version 1 abilities:
- Melee – 50% harder than anything else in the game
- Shadow Damage – unavoidable shadow damage to everyone in the room.
 |
| Blue Beam – Target deals 100% more damage. |
 |
| Red Beam – Target takes 50% damage. |
 |
| Green Beam – Target gains 50% healing and restores 10% of max hp. |
This boss was surely intense! Just look at those numbers! Such power, so amaze!
First testing attempt:
Jeff Kaplan: Okay! here I go!
*dies instantly*
Jeff: Ha! Okay, I just need to shield wall on the pull. Revive, I got this.
*dies 10 seconds later*
Jeff: Uhm, okay. Maybe this is overtuned. I’m going to double my health.
*30 seconds into the fight, someone misses the green beam and netherspite instantly reheals*
Jeff: Uh… that sucks.
*1 min into the fight, someone with the blue beam pulls aggro and instantly dies*
Jeff: Hahah, Joe! Learn to hit evade.
*2 min into the fight, someone misses the red beam and Netherspite instantly one shots him*
Jeff: OKAY, WHAT THE FUCK?!
Scott: I thought I was standing in the beam.
Joe: I thought the same thing too… but I couldn’t be sure, the dragon was right in my face.
Alex: It was really hard to tell…. because it shot over your head.
Scott: Why does it do that?
Alex: Well, what are my other options?
The bloody wyrms are floating 10 feet above you at that scale.
Scott: Ugh. I’ll go talk to the animators. An hour later…
Scott: Uh, guys. I have some bad news….
Alex: They couldn’t fix the animation?
Scott: Uh… worse. All of our animators quit…
Joe: …
Alex: So …
Scott: Just make the beams snap to the ground and replace the model with that transparent green dragon model Roman Kenney made for now.
We’ll replace it with something better before ship.
Alex: Got it.
Life Lesson:
“Unpredictable shit happens. When it does, you need to make the best of it.”
Joe: Oh yeah, and Alex, these beams should be fun for the people who use them.
Can you do something to make them a little bit more
Jeff: Also… if I’m in this beam, I want to be like the god of tanks.
Alex: Okay…
Joe: But Jeff, if you’re the god of tanks… why would anyone else ever want to use it?
Geoff: Oh yeah… that’s a good point.
It really seems like you should want to switch around who is using each beam.
Alex: How would I go about doing that?
Geoff: I don’t know… force them to switch who is on the beam… just kill them or something.
Alex: Alright….
Netherspite Version 2:
 |
| “Temp” Dragon Model |
Version 2 abilities:
- Empowerment – Gains 100% additional damage 15 seconds after being aggroed.
- Shadow Damage – unavoidable shadow damage to everyone in the room.
|
| Blue Beam – Target deals 100% more damage. Dies if in the beam for 20 seconds. |
|
| Red Beam – Target takes 50% damage Gains bonus health. Dies if in the beam for 20 seconds. |
|
| Green Beam – Target gains 50% healing and restores 10% of max hp. Dies if in the beam for 20 seconds. |
Mindset: I’ll do exactly what I’m told. That’s totally going to work, right?
Life Lesson:
“Listening is good. Blindly obeying is bad.”
Playtest Round 2:
Jeff: Okay! here I go!
*10 seconds later*
Jeff: Sweet! I’m alive.*15 seconds*
Me: Jeff. Give someone else the beam.
Jeff: What? No way, this is awesome. *20 seconds*
Jeff: Shit! What killed me? Did I get crit?
Me: You didn’t give anyone the beam…
Jeff: Why would that kill me?
Me: Because if you keep it for 20 seconds… you die.
Jeff: *facepalms then starts laughing* Oh god.
Next attempt: Everything goes smoothly. Joe and Jeff swap the beams. Repeat for 5 minutes.
Joe: Hm…
Geoff: Yeah…
Jeff: Well, it worked…
Me: Great!
Geoff: Not exactly… well… its kinda…
Joe: Boring. It’s boring.
Me: But I worked so hard to get those beams working.
Geoff: Yeah… that impresses us… but the players don’t really care.
Ya know?
Life Lesson:
“It doesn’t matter how hard you worked on something if the end experience isn’t great.”
Joe: There’s a couple things here… one… there needs to be more going on.
More tension… and well… anything else to break up the fight.
Jeff: Go ask some other designers for their thoughts.
Life Lesson:
“Farming other people for opinions is a great way to get new, raw ideas.”
Wyatt: Movement? You know, I got a lot of movement by using those void zones in Naxxramas.
Kevin: Huh… I noticed everything is percent max health, damage, whatever… wouldn’t it be better if some of this stuff scaled with gear?
Stephen: Its really hard to SEE changes like percent reduction… like buff the crap out of them in a way they can notice.
Eric: Instantly dying is bullshit. You need to give them some sort of warning.
UI can communicate this perhaps if you do it right.
Jeff: Tedious fights that do the same thing over and over again get boring.
Please mix it up somehow.
Scott: Keep in mind there might not be two tanks in this group.
Make this fight work with like one tank and a melee shaman or rogue please.
Joe: I keep running out of energy on my rogue.
This means I’m way worse in this fight than everyone else.
That’s not cool.
A. A.: These names are really placeholder.
Can you spice them up a bit?
Life Lesson:
“Raw ideas are great. However, getting overwhelmed by all of those ideas… not so great.”
Me: Augh! This is way too much. Fuck it.
I’ll just do all of this stuff at once.
Also, I’m going to make it way easier to tank this.
Me: In fact, why should I even make the tank worry about aggro?
I’ll just give it to him.
What makes tanking work? Defense? Fine. MORE DEFENSE FOR ALL.
Me: I also don’t think its cool that the healers get to have all of the fun. Know what?
No energy costs in the green beam.
Me: Hell, fuck those DPS guys who just stand in the blue beam. They should want to get out of that beam.
Me: Names? I can DO names…. oh yeah! Get Thesaurus’d, baby!
Me: Crazy? A break? Fine. Uhm… crap… all of that work just took a very long time.
I have less than a week to finish this instance up.
Ugh, I’ll slap on a mode where he does knockback on random people for a while.
Me: Shit. This is really hard to tune.
I’m going to make separate versions of the beam effects for Netherspite and players.
Me: Predictable? PAH. I’ll show them unpredictable.
What if you can’t predict where the portals will be.
SITUATIONAL AWARENESS. Oh yeahhhh!
Life Lesson:
“You really, really need to stop when you get on a roll.”
Netherspite Version 3:
|
| “Temp” Dragon Model |
Version 3 abilities:
- Empowerment – Gains 100% additional damage 15 seconds after being aggroed.
- Void Zones – Opens a massive void portal that lasts 25 seconds.
- Shadow Damage – unavoidable shadow damage to everyone in the room.
- Every 90 sec, Netherspite banishes the portals and randomly casts Netherbreath on all targets.
- Netherbreath: Knocks back an enemy, dealing damage.
- Portal Resummoning: Portals randomize their positions after each Netherbreath cycle.
|
| Dominance – Gain 1 stack per second. Netherspite: +1% magic damage per stack. Player: +5% damage dealt per stack. Player: -1% healing received per stack. Player: damage taken by spells increased by 8% per stack. |
|
| Perseverance – Gain 1 stack per second. Netherspite: -1% damage taken per stack. Player: Netherspite will attack you. Player: -1% damage taken per stack. Player: +5 defense per stack. Player: +32,000 health. Player: -1,000 health per stack. |
|
| Serenity – Gain 1 stack per second. Netherspite: +4000 health per second per stack. Player: Healing done increased by 5% per stack. Player: Spell and ability cost reduced by 1% per stack. Player: Maximum mana reduced by 200 per stack. |
Me: Huff… huff… WHEW. Good deal. I got all of that shit in. Boom! Play this game my way. Bwahaahaha.
Scott: What the hell… Brazie, those tooltips are a paragraph long each.
Me: So…?
Joe: Uh… does this even work?
Me: Well… QA managed to beat it… Once.
Geoff: …
Joe: …
Scott: …
Ship it.
At a certain point, you run out of time.
When we playtested Netherspite, he worked and with some iteration, we found he was consistently beatable, took a unique strategy to beat and was highly memorable.
Players adapted.
They shared their strategies, in the form of webpages, wiki entries and images.
The encounter seemed fine to many… but I could never shake the feeling that something was wrong. Even minor errors were punished. Few situations were recoverable.
As Steve Burke would later say: “It was a dance. You had to take exactly the proper steps or you failed.”
Looking back many years later, with more design experience under my belt, I understood. It was such a highly “plotted” experience… you were completely forced to play it exactly the way I built it.
That was what was wrong. It wasn’t really a game. It was a physical execution and coordination check.
There were few windows for unique success, personal glory or even sacrifice for the team.
You just did your job, collected your loot and wandered on.
It showed in the long-term play results.
Players skipped Netherspite, regularly heading to the Chess event.
Later, when better geared, some groups would beat it for their resto shaman.
But few felt the effort was worth the price.
Retrospective
Looking back, I see a million places to simplify and improve this encounter.
Telegraph abilities that require moments of protection from the beams.
Simplify the beam mechanics. Simplify the spell layering.
But most of all… create situations that allow the players to use their own tools, not just yours.
What did you think of Netherspite?
Share in the comments! I’ll post the best ones here.
Brett Hundley – Machinama: “looking forward to some good stories in the comments, whatever guild I was with in BC was certainly not attempting him and his scary magic, haha”
Lessons
I wanted to create a novel and memorable fight that forced people to perform roles they wouldn’t normally have.
It was so jarringly different from most WoW fights and brutally punishing if you messed up, that many teams chose to skip the fight entirely.
It’s important to constrain the difficulty of your content when you are also breaking core rules so that it doesn’t become unapproachable to players.
This encounter solidified my lessons about simplicity and clarity of purpose.
11 Responses
I haven’t played WoW, so this comment may be lacking context, but it seems like it’d be pretty difficult making portal gameplay distinct from DDR or bullet hell (both of which also sometimes struggle with giving players choices to make). It might have been interesting to try an Ikaruga-esque polarity system to make the fight more dynamic, but that’s less about the portals and more about chasing down distinct energy types.
Random question: Is 3 iterations typical for a encounter like this?
No, it was actually more like 6-7 incremental steps to hit those 3. Plus a LOT of number iteration that happens continuously thoughout. 3 Makes for a better story tho 🙂
I haven’t played WoW, but I find this interesting to read still.
That last step when you just did EVERYTHING and really over-designed is that a common mistake when you as a designer have figured out the issues you have?
I don’t know how common it is, but it feels like the natural thing to do, instead of finding one elegant solution(which is super hard, and sometimes impossible) that fixes everything, you fix all the issues one by one and then you get too much.
Your conclusion was spot-on. While current content, we killed Netherspite a couple times then skipped him in every run thereafter. It was just too complex to explain to new players, and there was too much individual responsibility. If one person stepped in the wrong beam, or the right beam at the wrong time, you could wipe the raid. One of the worst encounters in TBC. (sorry!)
Good stuff. I remember playing Naxx when it was actually relevant before stepping into the other BC Raids. It was a great instance, and honestly you hit the nail on the head with the fight analysis. It was definitely a “play by my rules” fight. The beams made the fight more of a routine. However, I will say that if you had poor execution and communication, you could easily be punished by overlapping your debuffs.
I was a little sad that you guys didn’t get to use the original model, and the story didn’t revisit that, so I’m curious if anything else came up after “Oh god no animators!” and the model having some inherent flaws being so large.
Overall, if I had to rate Netherspite on a scale of 10 based on it being in a BC bubble of bosses…. I’d say that Netherspite was a solid 6.5/10. It was definitely more memorable than most might think, and is one of the few “stand in stuff” fights that players got to fiddle with.
Thanks for writing this. Looking forward to more when you have time.
I think you’re being a little hard on yourself about the design being restricted so that he can only be killed the way you wanted. While that might be true, it was one of the few encounters in early WoW that could be defeated with a wide variety of raid groups. That can probably be attributed to the beam buffs and that it largely equalized a lot of the classes and how they contributed. You didn’t need 3 rogues, 2 warriors, etc.
I do agree with you on the spell layering, that probably could have been done more eloquently. I’m not a huge fan of telegraphing (particularly when it’s overused), and don’t really feel that any of his abilities required that. The biggest challenge was managing the beams, the rest fell onto player skill whether or not you could beat the encounter.
I think if there was one thing I personally would have done differently in designing it, would be a smoother transition back into the beam phase. It was very unforgiving and the green beam hitting him for even a couple seconds would typically result in a wipe.
I disagree, Netherspite was one of those rare fights that we killed once or twice and many, many people vowed never to return. Not too many of those. Rhyolith comes to mind. Not super fun.
How well do you think “training” the players by giving the boss’s abilities to trash mobs would have worked in alleviating the problems with this fight? It seems like the preference in designing boss fights has been to make the early players repeatedly wipe on the boss while they learn all the boss abilities, and then have those players write the guides that everyone else follows. Were there any attempts throughout the years to teach the players the boss abilities through trash mobs and then have them fight the boss on instinct, or did it not work out that well in practice? I remember small attempts at this throughout the years, especially with dungeon bosses sharing abilities of raid bosses, but it seemed more like the designers were dealing with a limited set of abilities they could use rather than any actual attempts at teaching the players how to react in boss fights.
As for the Netherspite fight, I do remember my biggest issue was with the color association of the beams. You guys put a lot of effort into making sure the players could immediately read their surroundings and understand what was going on (though that somehow went out the window for many players during the Shade of Aran fight in understanding that fire is bad), but there was none of that with regards to the beams. Sure, green is pretty well understood to be associated with healing, but I remember continuously mixing up what the red and blue beams did on this fight, even after weeks of killing him. The icons may not have existed back then, but WoW currently has a blue shield icon for tanks, a red sword icon for DPS, and a green medic (cross) icon for healers, yet Netherspite’s red beam was for tanks and his blue beam was for DPS. Was the color association not a problem that showed up during playtesting?
Also, if I may ask, at what stage of the dungeon design are the boss mechanics worked out and finetuned? Your posts give the impression that the level design itself is already finished and you’re just placing mobs into the space. Were you not given much leeway into how the area the fight took place in was designed? What happened if you ended up finding out it was just too big or too small for the encounter – were you just forced to rethink the fight?
Thanks for posting these, they’re an absolute blast to read.
I may be an exception to the rule, but all my groups (I’ve ran with 3) did kill Netherspite every week if we weren’t in some hurry. Most of the runs I was the OT (he never dropped my gloves), so I remember how we had to switch carefully to avoid moving the boss and mucking up the beams.
The fight did resemble a dance, but in a good way. It made everyone care about the positioning and alert of the mechanics. I am a huge fan of individual responsibility in a boss encounter and Netherspite delivered it in spades. There were times when rogue stepping in the tank beam (with evasion ready) or dps stepping into healing beam saved the day, so those opportunities were there just as with other encounters. I think I once “got” the healer beam as a prot paladin when our healer DC’d and threw a heal or two – felt good.
Netherspite was unique in a (good) way that is he required something completely different than usual from the group. Sure the novelty faded as execution streamlined but this problem is hardly unique to the encounter.
What also was true of Netherspite is that he made the group care about who the bad players in their ranks are (the feat later surpassed tenfold by Archimonde). If you had a bad dps or a bad healer they simply never were assigned a beam and were consequently demolished on the meters. In many encounters this is reversed (players doing complex jobs look worse on meters), and Netherspite was really cool in this regard.
Thanks for writing these series.
I think that there’s room for pure execution encounters in raids, and Netherspite was far from the last such. Even today, I’d argue that something like Paragons of the Klaxxi is pretty much down to executing everything just right to win, without a ton of latitude to change things up. Ascendent Council in the Bastion of Twilight might be another good example.
I think that “execution” fights can be satisfying and should be a part of the experience. Proper raid pacing, IMO, should feature different styles of encounters to break things up. Having some fights be “do the dance just right” and some fights be more freeform, allowing each raid team to figure out what works best for them, I think makes for a more diverse and enjoyable raid. Obviously, though, it’s a spectrum and there can be fights that fall in between those two poles and feature elements of each.
These post-mortems are great, hope to see more.