This is a nifty little experiment. Were the set differences intentionally customized for each level or were those randomized too?

using rot13.com:

BZT! Jura V tbg gb 18, V jnf tvira ubcr, yvxr znlor V pna rfpncr? Ohg gura gung jnf dhvpxyl qnfurq jura V ernyvmrq V pbhyqa'g whzc gung uvtu. Gura 14 pnzr naq vg frrzrq n yvggyr zber cbffvoyr, ohg... abcr. Naq gura 29 pnzr, naq... jryy, nf vg gheaf bhg, lbh whfg snyy bss gur fperra naq gung'f vg. YBY. Vg jnf obgu shaal naq qrcerffvat ng gur fnzr gvzr... V qba'g xabj jung V rkcrpgrq. Vg jnf cresrpg, gubhtu.

V ernyyl rawblrq bcravat nyy 50 sbe fbzr jrveq ernfba. Gunaxf.

While playing the different stages, I couldn't help but think of the first time I played Super Mario Brothers 2 and learning how the different characters were configured to jump and behave differently.

Hopefully you can use these as "test beds" for future game projects, and customize the character's behaviour based on how you want the player to feel. Maybe you could come up with your own SMB2 tribute or spiritual successor...

This was very interesting! Thanks for sharing it. :)

Running through these was cute. Nice palettes!!

I once started a project with the plan to split each level into individual exes but I wasn't able to pull it off. This is a good approach though. I like things that give you too many options at once to prevent you from easily going through in a predetermined order, like an unordered mixtape.