Friday, 21 August 2020

Turbo Tomato: Perfectly Balanced

Progress in Turbo land has been less than turbo-charged of late, with a collection of outside commitments inconveniently inconveniencing me and keeping me away from working on the game as much as I'd like. Even so, progress has been made.

This time, I've started on the big play through.

I'm largely happy with the world and level design in TT as it is, but the number and types of enemies placed in each individual level is (more than) a bit haphazard and unbalanced. Playing the game, the feeling was there was no real progression through the worlds with difficulty going up and down more or less randomly.

I say "feeling" because I had no real data to back this up. So over the past few weeks I've been working on rectifying this, and more.

I started last time with generating a difficulty score for each level and starting to make them progress more regularly, by first making the individual enemy difficulties more progressive themselves.

This work continued, painfully, over many interations of patterns for distributing enemy types across the levels in the game. This took a long time, and was something of a thankless slog, but I finally feel as if it's got somewhere. I have (more or less) settled on a distribution that gives the shape of difficulty curve I'm looking for - essentially, levels get more difficult as the player progresses through a world, then take a half-step back in the first level of the next world to give them a small breather.

WIP enemy distribution and the resulting difficulty curve

 

Using this pattern, with variations to include particular enemies on particular levels as well as special types, I have completed an initial balancing pass on the first two worlds. They feel less random to play, with a nice variety of enemies across the levels, which was the entire point of the exercise. I also passed out an early demo of the newly-balanced first world to a couple of testers and they agreed. So ... win? I'm seeing the light at the end of the tunnel on this part at any rate.

As I suspected might happen, I had too many enemy types to fit comfortably in to the available levels. Luckily though, I also had a few enemy types that I wasn't too fond of. Say goodbye to Apple Tina, Punkin and the Cybermarrows.

I also found time to create an asset pipeline for the first couple of end of world guardians (boss fights!), drop in early versions of the first new tilesets from Chris,  fix a few bugs and do a little more polishing. A second alpha demo found its way to testers after that.

KTOD's ALIVE!

You know, maybe things aren't too slow after all?

The immediate next steps are to continue playing and tweaking world 1 and 2 for a while longer to make sure everything is how it should be, then move on to doing the same for worlds 3 and 4. There's a level I want to redesign on World 2, and I think World 4 will need some reworking of enemies (and possibly some new ones), but I'll cross those bridges as they come up.

So, more Excel for me.

You, stay fruity!

 

No comments:

Post a comment