Matt Leone is Polygon’s features editor.
About once a year, I tend to get obsessed with researching a story to a degree most would consider odd.
Now, there's an element of obsession that goes with every story — a moderate, healthy amount that gets what you need but doesn't wake you up at 4 a.m. because you can't stop thinking of ways to phrase interview questions.
But it’s those semi-annual unicorns that tend to be the most fun. I just finished one of them, an oral history of the 1997 video game Final Fantasy 7. It was a big project, with 35 interviews and around 30,000 words. Perhaps most notably, it’s a story about the people who made the game much more than it is about the game itself. I spent a long time — about two years, off and on — getting into their heads and trying to piece together what it felt like to work in their office 20 years ago.
I think that’s why I got so attached to it — it was easy for me to relate to their stories, and the more I dug into it, the more I became fascinated by how everyone saw things a bit differently. In game-media-history stories, you tend to get a lot of single-source interviews, and I love showing that something that’s true for one person is less true for someone else. It helps explain that games are made by people rather than machines.
Anyway, here are some takeaways after approximately two years of reporting:
Look for opportunities
Planning a long oral history is always a bit of a gamble because you need a lot of people to participate for it to work well, and up-front you never know how many will agree to interviews. For FF7, I saw what looked like an opportunity because I knew I could get access to three or four of the bigger names involved, so I started to pull at those threads. The first few went well, so I kept on going and soon I had a big roster of subjects on board.
I think there’s huge value in not talking yourself out of a story because you don’t have a good head start, but when you have that head start it can be a great thing to lean into. Personally speaking, it also helps motivate me when I see things moving forward. Over the course of a couple years for this story, I spent a lot of time organizing interviews, transcribing and editing text, and working with photographers, video producers, designers, developers, and even game designers building a trivia game. But if I hadn’t been lucky enough to get a lot of interview subjects to participate, it would have been hard to get as excited about doing all that additional work.
Transcribe and edit interviews as you go
If you ever do an oral history in this format I would highly recommend transcribing and editing together interviews as you go, rather than conducting them all up-front and trying to stitch them together later.
It makes it far easier to see what you’re missing, and helps you know what specific points to ask about in follow-up interviews so you can get people to respond to existing quotes. I didn’t start the FF7 story as vigilant about that as I should have, but as time went on, it helped give the story a rhythm instead of just having people loosely talking about the same things.
Try to do interviews in person
Similarly, if you can afford the travel and time investment, I think it’s important to do interviews in person — or at least over the phone or with Skype. That’s common reporting advice in general, but I think it’s especially important for this format where the story lives and dies on quotes feeling casual.
Plan for extra editing time
Expect the editing to take forever, too, because it’s hard to bridge gaps when you’re working with other people’s words. It’s challenging to balance getting the best quotes in, with telling the smoothest reading version of the story, with making sure you don’t take anyone out of context.
Personally, I love it — it feels like solving a puzzle — but it can be frustrating when you need someone to say something really simple to fill a gap but you just don’t have a quote for it. Filling in those gaps and editing the FF7 story led to multiple delays on my end. Fortunately, because it was a retrospective story we didn’t have a firm time peg to hit.
Expect some inconsistencies
Another task I spent a lot of time on was coordinating with photographers to get more than 90 original images for the story. I had the idea early on that I wanted the story’s layout to feel kind of like a museum, and while we didn’t end up leaning into that theme too heavily in terms of design, we did make a big effort in that direction with our approach to the photographs.
Getting them all to look consistent was tough because you’re dealing with many different interview subjects in many different locations, where you sometimes have limited access, and are working with different photographers. For one assignment, I needed to track down a specific building address based on just an old photo, so I did a street view search for a chain coffee store that appeared in the same photo, then “walked” around each one. I eventually found it (see image below), but things like that took awhile. The photographers we hired did a great job, but it was a time-consuming process.
Toward the end of the project, we also decided to do a series of video interviews with different people in the story. We wanted to focus on four people that would have something they could show, rather than just something they could talk about. I tend to find that long video interviews — especially translated ones — can be hard to skim for those who only want to see part of them. We filmed two artists and one composer who showed us their process, and one engineer who walked us through a tech demo.
The interviews came together at the end of the reporting process with a variety of different videographers, so they’re not perfectly consistent. Our video team did a great job making them all feel cohesive with editing, but given a chance to do it all over again I’d try to plan the video assignments earlier in the process.
Feel free to get in touch if you have questions about the story, process, or for anything regarding Polygon's features section: @LattMeone. Thanks for checking this out.