Friday, October 5, 2012

Studio Stories: How To Get Work Done

This will either be boring, or fascinating. You deci...yaaaawn.

In my effort to sound authoritative and feel like there's actual some use to this blog for interesting game developers, I'm going to talk about the general work schedule we had in Paragon Studios. This is something most studios have, even outside of the game industry, although it something that was completely new for me when I first started. These are schedules that are important not only for producers (the guys who set them up), but also for workers to know in order to pace themselves. So, without further ado, here are some of the basics for the work schedules in a game studio! ... Why does that not sound very exciting?

Milestones

In the grand overview of things, we have milestones. Every studio is different, but every studio should have milestones. Think of these as semesters in school, only much shorter. These can be as little as a few weeks and as long as a month or two. This is the time in which you'll be getting your work done. The studio will set up a goal that a zone or such needs to be done by milestone 5 or 6, and you're on milestone 3.

It's the responsibility of both a producer and of a designer/artist/programmer to have a base understanding of how long it'll take them to do something. For example, when we were setting up the new content for Issue 24, I was confident that I could get 7 story arcs, a zone event, develop more LUA scripts, create several entities, create 30 new maps, and around 10-15 mini-maps, in 3 milestones. I had set up in my mind how long it would take me to do roughly each of those, then added buffer time for QA testing. This is the most important part to remember.

Let's say you've got a month of work for a milestone. You almost never want to schedule yourself so that it's going to be a photo finish when you're done working. Because you're never done working when you finish your last mission. You still have plenty of work to do based off of feedback from other designers and QA. Those people will definitely find bugs, even if you've combed over it a million times. No one's perfect, so don't worry, but you need to leave yourself iteration and bug-fixing time at two crucial points in milestones. This is a nice tie in to the "quality" section of milestones.

You're ten miles away from polish, don't take a wrong turn at screwed.


Rough-In, Playable, Polished

Milestones will have different goals for where content should be. You have rough-in, playable, and polished. The precise definition of what these are varies from studio to studio, but here are the broad strokes for each.

Rough-In means that you have the base framework set up for your mission. This is where things can radically change in your arcs and big additions/swaps can be made. Arcs get removed, added, etc. Your stuff doesn't necessarily need to work in this state. You just need to have the base foundation built to see what you're going to be working on next. For City of Heroes, this meant that the main encounters, maps, and characters in missions were created and put in without any grievous errors popping up. No errors appearing meant that the system was theoretically OK'ing your work, but it didn't mean it worked.

For Issue 24, this meant having the main "what's going to happen" parts of the missions done. Things like ambient spawns, music, exact details for cutscenes, etc. didn't have to be finalized. Think of it as making a blue print for building a house and laying down the initial foundation. The rest of the building of that house, the little details etc., aren't done, but you know the road you have in front of you and what you have to do.

You still have to figure out where to put the torture chamber.

Playable means that you have to take everything you did in rough-in and make it playable by a human being. There should be no SHOW-STOPPING bugs, which is another term for progressing blockers, although show-stopper is a more terrifying description. This means QA should be able to play your mission from start to finish without a problem. The level of testing here can vary. This was a stage where we had a first pass of text in, cutscenes were hooked up and working, maps were in, objectives were all set up, etc. Playable is the stage where you build the foundations and walls for the house. People should be able to walk inside and have the electricity working.

The playable milestone is where some aspects of your missions can change. No new work should be added, just building upon existing work, but it can be allowed as long as you're confident with your schedule. Producers will not be too gung-ho about it though, and you should consider yourself as walking on glass if you're adding new things to do during the playable stage.

Now, a mission can be playable, but that doesn't mean it's ready to ship. How many missions have you guys played in an MMO where it's missing things like waypoints, ambient spawns, or other things that make you go, "this needed something extra"? That's where we get into the POLISH stage.

Not that kind of Polish.
Polish is where all the fun things happen. This is where you decorate the house and actually make it look nice for people to live in. Cutscenes get better camera angles and music, environments get touched up to look pretty/textured, small side bars are added to dialog trees, waypoints are added to make missions smoother, etc. This is a stage where nothing new should be added unless you're a cowboy. There's a lot of temptation in polish to try to add completely new things because, hey, wouldn't it be cool if there was another mini-arc of 2 missions to do on the side here? The answer is yes, it would be cool, but you'd be starting from scratch, and you run a huge risk of neglecting your current work to get it done.

Polish can be the best stage and the worst stage; you need to exert control over how much you're polishing and how much you're adding. This is where QA goes through things with a fine tooth comb to squash any and all bugs. You shouldn't be changing anything to how your missions are completed; it would be like decorating a house by tearing down part of the support structure.

In Issue 24, this meant adding extra say dialog to the allies you had in your missions, sprucing up the environments in maps with extra props and NPC's, putting depth of field and music into cutscenes, adding extra side dialog options into dialog trees, etc.

By the end of polish, you should be just about done with your work. You need to leave yourself some extra time when planning the next milestone for the occasional bug fix here and there, but if you've done everything right, it shouldn't be anything major; most likely a grammar issue here, a small balance tweak there. If you've done something incredibly "hackey" and not full proof, then your 3 milestone deadline just turned into 3 1/2 4 milestones, and you'll find yourself behind on the next project.

That's all for today. Either you're asleep by now, or busy taking down notes. Next time, we'll go over the proper usage of keyboards and how the right mouse can make the difference between not crunching at all.

That's not true, don't worry.






10 comments:

  1. There's one particular mission in Issue 10 that I always remember: you're in a burning forest full of Rikti. It's nothing special; you go in and you find the boss, and you're done. But, in the Issue 10 beta, that mission had an extra element: the reason the Rikti were in that burning forest is because a drop ship crashed into the forest, and you got to go in and destroy the drop ship. That's the only time you got to see a drop ship outside the invasions, and while it was damaged and in the ground, it was wicked cool.

    Unfortunately, it was removed in a later beta build, and that's how we have the current generic mission. I believe the reason it was removed was that the downed ship still had the VFX of the one in the zones, and the designer didn't have the time to get a version of the ship that looked damaged before it was time to release the mission. So it had to go.

    This is where me as a player becomes really disappointed at the rules. That mission, without the drop ship, is generic and boring. That mission with the drop ship, even if the drop ship doesn't look exactly right, was very cool. This was a situation where I really argued for the drop ship to be left in, because it added so much to the mission. I argue that, while the "polish" phase is right in not adding anything, it also shouldn't *remove* something that adds to the mission just because it's not "perfect".

    After all, there are much worse graphical issues in the game than "the drop shop doesn't look damaged". The entire last mission of the Lady Grey TF is full of graphical glitches because the "Ultra Mode" upgrade broke the map and made all sorts of geometry disappear. I bugged several different QA people about this mission, and I even had someone tell me it had "always been broken". That's when I made this video: http://www.youtube.com/watch?v=EOPInTuOJ2I

    It feels like in that mess of milestones, the studio should dedicate some time to "fix what broke, even if it's not directly related to what we're doing right now". The Ultra Mode update was supposed to make the zones look better, so it seems like the "polish" phase concerned itself *only* with zones. If a mission like the LGTF broke, then it stayed broken because it wasn't part of the checklist. I understand there's no time to fix everything, but when a map is as seriously broken as that one in a mission that is meant to be run repeatedly by high level teams, then maybe you can make 29 maps instead of 30 for the new issue, and re-do the broken map.

    ReplyDelete
    Replies
    1. Regarding this, there's also time for upkeep that is put on everyone's schedules while we're working. So during our milestones, there is also the understood portion of, "also fix bugs while you'e working". The graphical issue sounds like it would've fallen under programming.

      The Rikti Drop Ship example I can't comment on, as I wasn't around and I don't know the circumstances behind its removal.

      Delete
  2. Also, something that maybe you can comment on in a future update: I call it "don't design for a game in a way that breaks the expectations of the player".

    This one was particularly bad during Issue 12 and Cimerora. A COH player expects to be able to fly around the entire zone until he hits a warwall or the zone is obviously over (water everywhere, huge mountain range). Most of the early missions even mention "we've set up a force field" to explain why you go into an instance and you run into a blue force field down a street you're not supposed to go to.

    Cimerora broke this expectation by having a force field at the entrance of Romulus' fortress, with enemies clearly visible on the other side. I'm a superhero. I fly. I should be able to fly over the wall and smack those enemies.

    That force field is not even justified in the mission briefings. At no point it's mentioned that Romulus has set this invisible, impenetrable wall. It's just there. The reason given was to avoid seeing the 5th Column banners all over the place and spoiling the surprise at the end of the TF. But the "zone" version in the area doesn't even have those banners!

    Glitching past the force field in Cimerora and fighting the enemies in there became sort of a sport because of that. To this day, players find glitches to get on the other side of the zone because it *makes no sense that we wouldn't be able to*.

    Similar point, but for the lore. If you're going to write a story for a game, make sure you're not wrecking the backstory to do so. This was very painful in Issue 15 with the Barracuda/Dr. Khan Task Force. The writer for the TF seemed to be under the impression that Reichsman was the leader of the 5th Column, when they had nothing to do with each other. It was latter handwaved to "Requiem rescued Reichsman to recruit him into the 5th Column", but the cutscene for when villains free Reichsman still says this:

    "A service? I think not. I serve only the 5th Column. Instead, I shall dispose of you all!"

    Which makes no bloody sense. Reichsman was the leader of Amerika Corps from another dimension, he came into Paragon City, the Freedom Phalanx beat his ass and put him into the stasis field. He shouldn't even know or care about the 5th Column.

    Those examples are from before you were hired, so I'll give a much more recent one and you can tell me if it was your fault: destroyed Galaxy City. Galaxy Girl's statue is directly between Freedom Corps and a War Wall. But in the tutorial zone, there's many buildings around her that don't exist in the real zone, and the war walls are much further back. And we see more of the statue than we should be able to unless Freedom Corps was completely demolished, at which point it makes even less sense to have those tall buildings still standing around her. Someone wasn't paying attention.

    ReplyDelete
    Replies
    1. My fault. It's interesting, however, because this is another example, no offense to you Leandro, of listening to feedback that you might not like and taking it honestly. When you wrote, "someone wasn't paying attention", it made me irritated - mainly because I spent so much time on the really fine details of the tutorial, making sure everything was right, that little easter eggs were there, etc. I probably paid *too* much attention to things in that tutorial. So, when I read someone wasn't paying attention, it makes me want to yell, "What about ALL THE OTHER THINGS that I carefully PAINTED LIKE AN ARTIST?! NO ONE UNDERSTANDS ME!"

      However, the Galaxy Girl statue, at the end of the day, is a mess up regarding where she was placed, so it's fair to say that, and it's just something to keep in mind in the future.

      Delete
    2. Well, you did paint the zone like an artist... it's just that you were painting the Mona Lisa while I expected the The Persistence of Memory. I felt it was created as a new zone, not as Destroyed Galaxy City. You wanted to show off the new players getting into the game the "new look" of the buildings, similar to the Atlas Park revamp. As a tutorial for new players it's great. But speaking as someone who started in Galaxy City and got his first badge in Galaxy City... it just *isn't* Galaxy City.

      Later in the game there's a mission with Twinshot where you go back to Galaxy City, and *that* mission feels like Galaxy City. I can peek past the force fields and see the Arena, and the "Tank" statue, and that's when it finally clicked that yes, Galaxy City was destroyed. The tutorial zone just didn't have the same impact.

      To balance your rage, I'll say that if the statue example is the worst flaw I can find of your years in Paragon, then it's proof of what an excellent job you did. You can see by the rest of the post that I'm not shy of ripping a new one to someone who messes with my precious lore :P

      Delete
  3. This has been a great blog so far. Not boring, and you update it a lot (maybe because you're bored? I dunno...).

    Hope you keep it up.

    ReplyDelete
    Replies
    1. Thanks! I would like to keep it going for as long as I can; it's also a help for me to continue thinking about these things critically and avoiding getting too comfortable/full of myself.

      Delete
  4. I agree, Crimson.
    It's been a great read so far, although I am sure it'll be a miniseries, rather than syndicated in the end. While you have time though, Sean, I'm awaiting each post eagerly.

    [@Dr_Toerag on Twitter and @Dr Toerag on Union]

    ReplyDelete
    Replies
    1. Thanks! I'm hoping it'll be syndicated until the end or until I'm cancelled.

      Delete
  5. I actually agree with your last lines - that people tend to forget how great having a wonderful (or even just plain new) keyboard/mouse can be after some time has passed.

    ReplyDelete