Iterations

Hey all!

We've been rapidly iterating in order to solve some significant design problems. That lack of progress combined with vacations made this post much later than intended.

But I realized that it is perhaps even more important to talk about the struggles during development since it is that iteration and failure that makes this entire process so challenging and rewarding.

The inherent problem with our design was the fact that the syphoning from other racers for items was entirely optional. One of the goals of this game is to effectively and meaningfully bridge the racing and dating halves. As it stood, the bridging mechanic we had - of syphoning for items during races - was far too weak.

So we needed a bridging mechanic that was integral to how the player approached the racing.

The idea we had was to create a sort of fuel system wherein replenishing fuel would be tied to driving near the other racers. All racers would operate under the same rules:

- a racer's top speed is higher when the fuel gauge is higher (they can still move when at 0 fuel)

- when two willing racers are close by each other then their fuel gauges will increase

The player is shown finally catching up with other racers and refueling to surpass them.

It was a bridging mechanic that would not be optional, provide a natural rubber banding of the AI, and wasn't terribly complicated for the player to understand. On paper, this sounded like a fantastic solution (you can tell where this is going, can't you?).

However, when we tried it, it just felt awful. It felt like it got in the way of racing. We were essentially taking away speed, the most important and engaging commodity in a racing game, from the player for not fulfilling the seemingly arbitrary task of holding down a button when near another racer.

And yet there were certain qualities about it that we liked. It presented vaguely interesting behaviors from the AI racers, even if some of them were terribly game breaking. The best part was that the behaviors were deterministic (like Pac-Man), which was something we wanted to try to preserve in future iterations.

So we needed a bridging mechanic that was integral to how the player approached the racing, but didn't actively detract from the racing.

Now, one of the other problems with the original syphoning was that once the player got an item, they would use it to get ahead of the rest of the racers. And as long as they maintain their lead, they would never need to interact with the other racers for the rest of the race.

After meditating on this particular problem, City Trial in Kirby Air Ride, and The Legend of the Zelda: Breath of the Wild, I realized that the paradoxical nature of both wanting and not wanting to be ahead of other racers is only a problem if the race track is meant to be navigated in only one direction.

By completely ditching the racing staple of mono-directional tracks, we not only solve the staying ahead problem but make navigation, path-making, and (hopefully) improvisation a part of the game.

This is the current solution we are testing out. We now have larger and more complex arenas that the player races around. The goal is no longer to reach the finish line, but to be the first racer to have traveled X miles or kilometers. The paths the player takes to travel that X kilometers is up to them.

For this simple test arena map, the light gray parts are the road and certain wall sections are color coded to match the racer that drives in that part of the arena.

The AI racers would each have unique deterministic routes through the arena. Whenever two racers drive near each other they receive a small speed boost.

Since the map is one interconnected space, getting into first place won't prevent the player from interacting with the other racers. Also, since the speed boost is mutual, repeatedly driving by the same racer will improve your bond with them more and more, but it will also make them increasingly likely to get ahead of the player.

Hopefully this will promote driving strategies that involve:

- learning the different routes of each racer

- charting a course in order to drive by as many AI racer routes as possible

- inevitably tweaking those plans based on the exact location of another racer (once the player gets close enough to see them)

- showing some preference for some racers over others

- and all the while maintaining as much speed as possible

The player seen getting a small speed boost when driving by another racer.

These are the hopes and goals for this current iteration. It's looking alright so far, but we make no promises that we won't have scrapped it by the time we get to the next post!

That's all for now.

Thanks for reading!