Super Dash #2: Playtesting and Frustration
Playtesting is super important for game development and there’s nothing new about that. At Valve they start as early as they can and are always iterating to put their design in practice. With Super Dash I didn’t have time or testers to do it so often but the importance of playtesting never left my mind.
In my previous post, I discussed how valuable feedback completely reshaped the structure of the game. In this post I want to write about how I’m tackling one of the biggest issues in Super Dash’s design: player frustration.
Screen Sharing
Most of the playtest was done via Discord, with friends showing their screens. This is not the best way since you lose the nuances of body language and facial expressions.
Later on, I had the opportunity to see people playing in front of me, and even the way they handled the keyboard and mouse were important for valuable insights.
Asking for feedback in text is convenient, but I've found that people are sometimes hesitant to openly discuss their frustrations. Not everyone is used to feedback culture so I believe it's better to find people online in specific subreddits or Discord servers.
The Problem
Super Dash is a bullet hell, so it's chaotic by nature. On top of that, the main mechanic is dashing forward towards enemies, making it even more hellish.
Enemies are constantly firing and in the first versions players were getting hit often. The first obvious move was to make the player invulnerable to bullets while dashing.
Dashing with precision was hard, so a mechanic to stop the dash midway was added. This fixed some repositioning issues but testers were still visibly frustrated, as if getting hit was unfair rather than a result of bad decisions. I don’t want people playing it and feeling like they got hit because the game is punishing. When they fail it must be because they made a bad call or didn’t have enough reflexes.
To address these issues, we implemented a temporary shield immediately after a dash ends. This ensures the player won’t get hit by a stray bullet and will provide some time to reposition before dashing again. The shield made playtesters less scared to try aggressive moves while removing the frustration of getting hit after a high combo. While the shield makes the game easier, it prioritizes the game’s design philosophy of rewarding aggressive playstyle.
Some other non obvious things were added based on common frustration, always following the principles already defined.
| Frustration | Mechanic |
| Ending a dash with just a few pixels before the enemy and breaking a combo | Extending the dash to hit the next enemy if its too close |
| Too many unpredictable bullets on screen | enemies that shoot many bullets have easy patterns to identify |
| Hard to aim at fast enemies | Slow the time while aiming |
| Single enemies on screen are not as fun to destroy as group of enemies | The shield can be used to destroy one enemy |
Demo and Never Ending Testing
After testing with so many people I thought the demo version was good enough for a public test, so I published it on itch.io. With so many different people testing I had valuable feedback enough for months of work. Maybe it wasn’t just me that got used to the game's current state but the testers as well. It shows how important it is to vary and get fresh opinions as well.
This makes me think a lot about the role of a game designer. I always thought a good designer would write a perfect GDD from scratch and have awesome ideas that would just work. When I published the demo I thought it was in the best state, but after seeing people play in front of me, so many new ideas popped.
Next Steps
Keep seeking feedback for the demo for and updating its state
Work on the full game that will be released by the end of the year