Slapping on a trending mechanic on top of Auto Cop: Part 2

Context

After the half-successful prototype of version 1, it’s time to try and improve the in-game numbers. The target for D1 is to take it to 35% from 23%. We did not have time to design and develop the progression system, the upgrade system of weapons. In terms of content, we had 10 levels.

Hypothesis

I made an initial plan to address these. My plan included

• Progression: An area is divided into multiple stages. Each area represents an environment where the auto cop clears stage by stage. I planned to add a screen that shows the long-term goal in a map with all the areas while transitioning from one stage to another.
• Upgrades: Buy and upgrade weapons with money earned from stages
• Adding to the ‘Power feeling’: Between stages, health will not be restored. This will allow us to put heavy health, which goes with the Auto cop. After each stage, the player can restore their health with money, or they can risk it. This gives a sense of risk-reward.
• Dismemberment: The enemies will be made up of body parts that can be dismembered when shooting -> Improving the power feel
• Others: New world. Some unique missions to break the flow. Like: Rescuing hostages

At that time, rougelite games were taking the mobile games market by storm. The north star example is survivor.io. Along with the cover shooters, our publisher was making roguelites with other studios. Roguelites had amazing playtime, but they lacked CPI. Our publishing manager asked me if we could combine the cover shooter with a roguelite mechanic. I thought it made sense and redesigned the plan around it

New plan

• Progression: An area is divided into multiple stages. Each area represents an environment where the auto cop clears stage by stage. An area is a roguelite run. Each stage is represented by a floor. The big picture of an area is that the player goes into the ground floor and reaches the top to clear the building. If the player dies, he will need to start from the ground floor again.
• Upgrades: Buy and upgrade weapons with money earned from stages
• Adding to the ‘Power feeling’: Between stages, health will not be restored. This will allow us to put heavy health, which goes with the Auto cop.
• Dismemberment: The enemies will be made up of armor parts that can be dismembered when shooting -> Improving the power feel
• Roguelite Skill Selection: After each stage, the player can pick 1 of 3 skills presented to improve themselves (similar to survivor.io)
• Others: New world. Enemy variations with and without armor. Elite enemies to break the flow of the game, boss level

Process / Experiments

This time, our priority was Auto Cop. So, along with the publisher, we decided to bring the full team into this. I shared programming responsibilities with the programmer to fast-track development. Since there were 5 people involved, I opened up the Trello board to track the development.

I programmed and designed the skill system. After some experimentation with the skill system, here was the skill selection

It was also important for me that we don’t break the flow and take the user right into the game after the skill selection. That ensures the continuous feel was there.

Finally, I sat with the lead artist, opening up Unity to fix the lighting and color issues. I also swapped materials to improve the look and visibility of the enemies and the character. The look was improved a lot. We added a new area. I also played around with the kill physics of the enemies, and I was satisfied with the final result.

Observations & Outcomes

After testing in the US, it held the CPI at $0.38; the D3 cumulative playtime was somewhat improved from 912s to 953. D1 Retention was 30% but since the playtime did not improve, it really does not make any difference.

TAKEAWAY & NEXT STEPS

The changes were bold. I made a very big assumption that the cover shooter players will like the roguelite genre. Obviously, the assumption was wrong. It was apparent from the cumulative D3 playtime. The level funnel also showed a big drop from area 1 to area 2, showing an issue with the continuation

I think the roguelite choices were not meaningful enough. In survivor.io, there are 2 main control points of the player: Movement and Shooting. Movement is controlled by the user, and the shooting is controlled by the skill choices. It was deeply meaningful because the choices carried weight. Each choice modified the player in a significant way that impacted the power. The change was visual and striking.

Whereas, in our game, the skill selection did not carry much weight. Because it was not driving any mechanic, but rather some stat numbers. Players could simply pick a random skill and keep going. The skills were not visual in any meaningful way. Thus, it resulted in a clunky mechanic that was slapped on top of a core because of a trend, but not for the betterment of the game

Thus, my conclusion was that we did make some good changes, but it was cancelled by the weird roguelite mechanic

The next step was to keep it simple and give it another go. If we follow the known progression design of our players, will it improve the in-game numbers?