Studio Stories: Developing Monthly Content

Sean McCann and the Search for Monthly Content

Monthly content. It's the holy grail for MMO's. It's the thing that keeps players around between big releases. I haven't played many games that have done monthly content, so I'll only be speaking about my experience when we worked on monthly content in City of Heroes - specifically the Signature Stories. This isn't covering things like additions to the Paragon Store, bug fixes, etc. In a nutshell, monthly content can be incredibly rewarding if you can pull it off, but it's also very difficult to pull off, as it requires a very focused effort on the part of nearly every department in the studio.

PS: I'll be talking about the experience of writing the signature story arcs in a different blog post. If that was the reason why you were reading, well, thanks for increasing the post count anyway!

The Content Plan

First, we make it awesome. Then, we make money.

We knew we had to get everything done ahead of time to pull this off. We wanted to release 2 arcs a month for 7 months; this means that we should be roughly 3 months ahead of each arc. This was a big deal, as I was also working on finishing up content for our free to play release, as well as getting ramped up for Dark Astoria. Now, before anyone thinks I'm saying it was too much work, I wasn't - it was the amount that I personally agreed upon. However, there was a plan. In order to avoid work overload, we decided that we would try to split the arcs between multiple designers. I would do the first episode, another designer would work on episodes 2, a third designer would do episodes 3, 4, and 5, and I would come in to finish off episodes 6 and 7. This was done to avoid overloading my schedule and allow me to focus on major releases. As I was working on episode 1, I was helping the designer working on episode 2 get caught up up for how the story should go, what the plan was for everything, etc., making sure we were all on the same page.

Unfortunately, this didn't really work out. I took over work for episode two, as the designer who was supposed to work on that was too swamped with work to do it. While I was working on episode two, I was getting another designer, up to speed on where the story was heading, what the future plans were, what beats had to be hit, etc. He started work on episode 3 while I was wrapping up episode 2 and continuing to finish up work on our free to play release.

I was already feeling a little bit of the heat/realizing what I signed up for after working on taking what was the equivalent of 3 arcs from start to complete on top of everything else; I say 3 instead of 4, as the hero and villain arcs were roughly equivalent. I was relieved to get reprieve from the other designer doing arc's 3-5, although I was worried, as I realized we had a lot of ramp up time to get him up to speed for the arcs; this was not his fault in the slightest, as he was busy pounding away at other content. I wondered if that time would've been better spent if I had just done it myself, looking at the amount of time spent getting two designers up to speed for the arcs.

One day, Matt Miller called me in to a conference room. He said he had bad news for me. The designer who was supposed to work on episodes 3, 4, and 5, was not going to have his contract renewed due to the titanic commute (almost 2 hours) he had to take to work each day, and that I would need to get those arcs done. This is roughly what I looked like when I heard the news:

I'm Jim in this picture, not Pam.

I still had to do all my other work, as that all had to get done. However, I had now inherited what were essentially 6 story arcs on top of that. I realized that these arcs, along with the ones I was working on, needed to be handled smartly. I scoped down parts of the arcs that were a little crazy and focused on using core features that wouldn't break (hopefully). I also enlisted the help of Jeff Hamilton to help with some of the systems work for episode 3. I won't lie, of course, there were a few late nights, even with this, and the occasional coming in on the weekend - I specifically came in on a Saturday to finish the last mission for arc 4.

This was how the rest of the signature story content work fell into place. The scheduling wasn't pretty, but it was done. I wrapped up signature story 7 while also working on Dark Astoria. However, one thing I didn't account for was testing.

I found that my work on the arcs were never quite done. You never truly find all the bugs in content on the first pass, and since we didn't release it to the test server before going live (something that we changed for the second signature arc series), we ended up finding bugs that weren't caught the first go around. So, while I was technically working on arcs 3 and 4, I was also working on fixing arcs 1 and 2. And on the new player tutorial, and Atlas Park, and Mercy Island, and the Death From Below trial. I'm not saying this to garner sympathy from anyone or to pin the blame on producers; none of this was anyone's fault. I'm saying this to demonstrate to everyone the realities of monthly content development. It was a constant stop-go-stop-go of development, working on arc 5, stopping to fix some small thing in arc 2, starting again to do arc 5, stopping to work on Atlas, etc. etc.

This also involved a lot of falling on my part.

Now, this is all just purely from a design perspective. Next, we'll handle a tougher part: art.

The Art Plan

Myself, David Nakayama, and Andy Maurer had implemented what we called comic book loading screens for the new tutorial. We were trying to think of a way to convey the story of the tutorial without using a cutscene, as our lumbering beast of an 8 year old engine couldn't do that. I came up with the design for a comic book loading screen, which would be a versatile system that could be used in any mission or zone. Andy Maurer implemented it in lightning fast time, and David Nakayama did the art for it.

I mention all of this because we used the same system to do our loading screens for the signature story arcs. We wanted to have a comic book cover for all the loading screens in a particular issue. There was also a request that each signature arc contain at least one new piece of art. Finally, we had the two big death cutscenes to handle - Statesman and Sister Psyche.

The main issue with art is that it can take much longer than design work. This can kill you with monthly content; a brand new final room or completely new enemy can take an artist more than a month to do. We had to use our time wisely. We formed a strike team of artists to handle the workload for signature stories.

Character art: Jay Doherty
Environment art: Chris Gregory
Animation: Colin Brown
Effects: Keetzie and Leo Braz da Cunha
Audio: Adam Kay
Composer (Statesman and Psyche's death music): Jason Graves

Doesn't have much to do with party formation, but it's Suikoden, and Suikoden is awesome.

The very first thing we did was scope the art needed for the entire series. Custom rooms and entire characters took a long time to make compared to props, and this was all work being squeezed in to current schedules.

Environment art would make a new lava geo, lava islands, and a dark obelisk for arc 1. We put a bit more into this as it was the first arc, aka our season opener. We wouldn't use environment art again until arc 4, and there we were taking a room that was made for end game content and re-texturing it. Arc 5 was the next big work for environment art. We were going to take an existing map that was supposed to exist in Roman times and "modernize" it as ruins. This was also done to save time, as not a lot of new geo would have to be made. Finally, environment art would deliver the final set piece, which was a giant floating island in space. Luckily, the very talented Ryan Roth had already made the Earth for previous content which was cut, so this saved us a lot of time.

Concept art was entirely done by David Nakayama. He, myself, Matt Miller, and Nate Birkholtz would get together and talk about our ideas for each comic book cover. David began work on the comic book covers around the same time that I would begin work on the arcs. I can't comment about how hard the pieces were for David, but I know we saved him time by never requesting more than two characters on the covers. He also gave us a ton of support on character concept art, which you'll see up ahead.

Example of David's work. More can be found here: http://pixel-saurus.deviantart.com/

Character art was a tricky beast. We wanted to do a new villain group, the Rulu-Shin, a new character, Marshal Blitz, and a final end boss, Rula-Wade. The Rulu-Shin and Rula-Wade were going to need concept art. Concept art is, in general, extremely helpful for artists. It saves them a ton of time, as they have right in front of them how the character or environment should look, so they can just get to work right away instead of doing research on what they need to do. We saved time with the Rulu-Shin by making 3 or 4 costume pieces for them instead of an entirely new character; I built the base of them with existing assets, then worked with David Nakayma  and Jay Doherty on what I thought the new pieces should look like. Jay confirmed that it was all do-able with the engine, and David went to work busting out a concept by drawing the new pieces on top of the current in-game models. Jay then banged out the Rulu-Shin in no time flat. Rula-Wade was done in a similar way. We took an already made costume, Rularuu, and altered it. Jay also prettied it up while he was there since it was a little older.

Animations and effects were focused mainly on the death cutscenes for Statesman and Sister Psyche. This had to be done way ahead of time, as this needed approval from basically every important person in the studio. Our art lead at the time, Gilbert Martinez, and myself story boarded both cutscenes on a whiteboard around the time that I was working on arc 1. We worked with Colin Brown to make sure none of the animations were too crazy, then got it approved by Important People. Once we were set on that, we put the requests in. Animation was done around the time of arc 3, and the cutscene was all hooked up when I was barely working on arc 4. Keetsie and Leo came in and did the effects once they saw how the cutscenes were supposed to go. The final bit was audio.

This was not our audio booth. But imagine if it was.

We had a contract with Jason Graves ( http://www.jasongraves.com/) to compose new music for the signature arcs. Adam Kay and I sat down and talked about what the music pieces should be like for the cutscenes. Adam then spliced together existing pieces of audio behind the cutscene music to give Jason an idea of what we were looking for. In the meantime, Adam was also working on doing the ambient sounds for the cutscenes - footsteps, rain, thunder, etc. Jason nailed the feeling of the piece on his very first try and only needed to make slight alterations to it - this was a big help, as it saved us a ton of time on iteration for both pieces. Below is the piece done for Statesman's death.

Fun fact: I had been trying to get a piano piece into the game ever since I found out I could make audio requests. Adam Kay teased me at the end of this, "All it took to get your piano piece was killing off the main character of the game. I hope you're happy."

That just about sums about the art time for the signature arcs! They involved a lot of pre-planning and smart usage of art. We had to see how we could get miles out of work that wouldn't take months, and I think we pulled off the art really well. The last part of all of this is publising.

The Publishing Plan

Signature arcs were going to run separately from our regular content updates, meaning that we had to publish them to the live servers every month, followed by patches for them, if needed. We also needed to make sure that they were up properly on the Paragon Market. I wasn't too familiar with the amount of work done for this, but I know that it was no walk in the park for our team, as the arcs wildly deviated from our regular content schedule and were their own beast to handle. Towards the end, however, I know we worked our a good system for getting them in, especially when we decided not to worry about contacts being in the game ahead of time.

Overall Lessons

This was the first time that most of us on the team attempted to do monthly content. I can only speak for myself regarding the lessons that I personally learned. A lot of these lessons were also learned talking with Melissa Bianco, Matt Miller, and Nate Birkholtz as we were finishing up the signature stories and figuring out how to move forward with the process.

Become one with the monthly content.

First, if you're going to do monthly content, you need a small, dedicated team to get it done. This team should be focused purely on monthly content. It shouldn't be a large team, however, as monthly content needs to be lean and fast. Concept art, effects work, and programming should be considered a shared resource for this team, programming especially! Adding new code can be risky, especially on a month to month basis, so you'll want to be very cautious with the amount of new code you're doing. The team should consist of one designer, one environment artist, one character artist, and one QA tester. (This is purely my opinion, by the way.)

Second, play test early, play test often. We didn't have as many play tests as I would've wanted, and that's my own fault. Your core team should be playing what you've done several times throughout the arc's life. This'll help give your arc that polished feel and bring the team together in the project.

Third, don't keep the same team forever. Developing monthly content is taxing work, as you're working on smaller projects with a faster rate of iteration and production. It's nice to be able to catch your breath and work on content that has a bit of a bigger and longer scale. This part is the one that's more opinionated than based off of experience, as there are dangers with team swapping; we had switched designers for signature series 2, but still were shuffling around artists etc.

The main takeaway with this is that the team designing the big ticket content items that are delivered every 3 to 4 months in an MMO needs to be different from the team making your monthly content. Your monthly content guys are the ones helping to keep users around month to month. Your big content guys are the ones who will release something to help generate new users to come in. Then, those users are kept around by the monthly content guys. It's a circle of life! Unless you mess up, then it's a line of death. And no one likes lines or death.

I need to stop making these blog posts so long.