the article wrote:
When you play Really Bad Chess, you always play against the same artificial intelligence. It doesn’t get smarter or faster the further you progress. Instead, as you move up in rank, the quality of the pieces you get changes. In the beginning you will have much better pieces than the AI, improving your chances at winning. But as you get better and win matches, the balance starts to shift in the other direction.
This reminds me of a teaching technique I heard recommended for teaching Shogi to children: the stronger player teaching the newbie will start with only their king and no other pieces! The newbie should be able to mate them pretty easily.
Then they play with the stronger player having their king plus one standard piece. After the newbie wins that, they play with the stronger player having their king plus two standard pieces. And so on.