Infinity Levels Studio, the winner of Indie Prize Best Mobile Game nomination at Casual Connect USA 2016, is a small Thai-based games studio that focuses on building differentiated gameplay and amazing artwork. Coming from a not-so well-known place to produce innovative mobile games, and due to the competitive nature of the category, Nikki Assavathorn, the head of the studio, was pretty sure they wouldn’t win anything. So she sat at the back of the room and didn’t realize her studio has won the award, and only an hour later, when she chatted with the other gamers, she found out that Blades of Revenge has won.
We are a very young studio, with most team members, including myself, having a very limited experience in this industry. I still feel a little bit overwhelmed when I come across veterans who have been in the games industry for decades. I think what differentiated us is that we approach the making of the game strategically, with a goal in mind. Games is an art business and it’s a lot more than just “fun”.
Last year, Infinity Levels got the Most Inspiring Southeast Asian Games Studio Award for another game, Ranch Run, as well as won the Evil Game Design Challenge at Casual Connect Singapore. It may not be plain luck, but we’ve made plenty of mistakes and wanted to share them as well in the Blades of Revenge story.
How The Idea Came To Life
Inspired by the success of those great puzzle RPG games like Puzzle and Dragons and Monster Strike, we wanted to create an RPG that would involve a bit of puzzle-solving thinking in each action. We came up with numerous ideas, so made a checklist to see which ones are really unique and fun.
The mechanic of Blades of Revenge is based on the premise that if there’s a grid, and enemies’ locations are random, the player should do something each turn to optimize their attack.
We wanted to make our grids different from normal grid-based RPG games, so eventually came up with the idea: what if the playable characters and their enemies lived on separate mirrored grids, instead of being in the same one? What would happen? And that was the start of Blades of Revenge version one.
Always Know Who The Game Is Made For
“Knowing the target audience is very important. One of the mistakes we made was that we didn’t clarify it for everyone on the team”, says Tanapon, our game producer. “While it needs to be on team’s mind all the time they’re working on the game.”
Game designers, programmers, artists, sound designers, testers, producers, marketing – everyone should be on the same page in terms of understanding the target audience, up to being able to imagine what our players look like in person. And make sure this person looks the same for everyone.
Expensive mistakes of art
And this is indeed a very important part, since design work isn’t done by a single person. Artists create characters just the same way as game designers do the gameplay. They must make artwork from scratch, specifically for that “target audience” person. Sound designers also create music and sound suitable for the particular audience. And testers make recommendations based on target audience, or better yet, testers are the target audience. Marketers provide targeted marketing to that exact audience the game was designed for.
Blades went through 3 versions, and these are expensive mistakes. We thought we would do it better this time, since it isn’t our first game, and yet we still didn’t get the art and game play right from the first time. What was planned for 6 months to complete ended up taking 10 months. Blades of Revenge will be launched at the end of 2016.
Playtesting with AI
Tools matter. We often neglect investing time in building game tools, but whenever we do create one, we speed up the work so much that we start wondering why we didn’t do it in the beginning.
Blades of Revenge is a game partly based on chance – like most games, since enemies randomly switch positions in each turn. With these enemies, it is much easier to write an AI that plays the game for you than to figure out the the average difficulty of a level.
If you play the game yourself and wonder how difficult it is, each play session would take 1 to 2 minutes. We needed to weed out the luck factor in this estimate: sometimes you can be really lucky and fast, and sometimes not. Here’s what we did: let’s say you need 10 games, and it would take roughly 15 minutes just to test one level with one configuration. A well-written AI that mimicks common player behaviors can playtest your game 100 times in 1 second! Or 10 seconds, if you want to watch it in hyper speed.
Not only is it useful for playtesting games like this, but also comes handy in finding bugs, both in game design data and in the code. Whenever the AI has 0 win rate, when it should be possible to win, is when you know something is wrong.
The most valuable lesson Blades of Revenge taught its creators is to trust their sense: if you feel that a custom-made tool would speed up your work in any meaningful way – go for it! Invest in creating tools to do the part of the task for you. And you can already try the result of their approach on iOS and Android.