Postmortems are always helpful in understand how other game developers work and what issues they brush up against. It can be very illuminating when multiple studios run into the same problems. They are also useful if you’re thinking of getting into the games industry to see if those problems are even ones you want to tackle.
The great folks at DrinkBox are always looking to be better than they were the previous year. Chris Harvey, from DrinkBox wrote a great postmortem on Guacamelee! for Indie Games. It goes into great detail and address many things that arise during production of a game.
Their art style for the game evolved from a cutesy look to a more adult look, with (surpassingly) the tech team making a first pass at some of the art.
While this success was naturally a result of the talents of the art team, it also depended on a cooperative back-and-forth between the art, tech, and production departments. In particular, over the course of Guacamelee! we changed how effects were developed, counter-intuitively putting the development more into the hands of the tech team and then letting the art team direct refinement.
He even shares internal issues that any company runs into when it grows. The studio had an informal work structure which works quite well for small teams but inevitable breaks when the company grows.
During Guacamelee!, we encountered difficulty scaling this approach. Some new members of the team struggled to find their place alongside peers when working without a clear authority structure. Lacking the confidence to navigate disagreements constructively, in frustration these team members sometimes ignored decisions or emotionally disengaged from certain tasks. As a result, we’ve begun to put a clearer authority structure in place for the team, and we’ve become more aware of the need to monitor personalities and provide ways for people to make their concerns known.
[If] you make the marketing material interesting on its own, it’s irrelevant whether it “sells” your game. Our focus was always on creating content that was on its own fun for people to experience and to be a part of, with essentially 0% of the design aimed at trying to get the game to sell.