The duolingo Oscar bot did something similar to me. It played 11 moves of book Yugoslav attack against my Sicilian Dragon, and then blundered horribly. I'm wondering if they programmed the bot to improve or to be very good at certain openings.
Oscar Bot weirdly strong

[Event "?"]
[Site "Chess.com iPhone"]
[Date "????.??.??"]
[Round "?"]
[White "?"]
[Black "?"]
[Result "*"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"]
1. c4 c5 2. Nc3 Nc6 3. d3 g6 4. g3 Bg7 5. Bg2 d6 6. Nf3 Bg4 7. O-O Bh3 8. Bxh3 Ne5 9. Bg2 Nxf3+ 10. Bxf3 f6 11. Bxb7 Rb8 12. Bg2 Rb4 13. a3 Rb8 14. Qc2 Rc8 15. b4 a5 16. b5 Ra8 17. a4 Rc8 18. Rb1 Qb6 19. Nd5 Qd8 20. b6 Nh6 21. Nc7+ Kf7 22. Bd5+ e6 23. Bxe6+ Ke7 24. Bxc8 Qxc8 25. Bd2 g5 26. Bxa5 Nf7 27. b7 Qf5 28. b8=Q Rxb8 29. Rxb8 Bh8 30. Re8+ Kd7 31. Qb2 Ne5 32. Qb8 Nf3+ 33. exf3 Qxf3 34. Qc8+ Kc6 35. Nd5# {*}

[Event "Computer Game"]
[Site "Chess.com iPhone"]
[Date "2023.11.03"]
[Round "?"]
[White "michaelpowers8"]
[Black "Oscar"]
[Result "1-0"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"]
[WhiteElo "1822"]
[BlackElo "200"]
[Termination "michaelpowers8 wins by Checkmate"]
1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 d6 6.Be3 a6 7.f3 Qc7 8.Qd2 Nc6 9.O-O-O Be7 10.g4 Nd7 11.h4 Nce5 12.g5 h5 13.gxh6 gxh6 14.Bxh6 Nc6 15.Rg1 Nd8 16.Bg5 Bxg5 17.Qxg5 Nc6 18.Qg7 Rf8 19.f4 f6 20.Nxe6 Qd8 21.Nxd8 Nxd8 22.Rxd6 Nf7 23.Bc4 Nb8 24.Qxf6 Ra7 25.e5 Nxd6 26.Qxd6 Be6 27.Bxe6 a5 28.Rg7 Rxf4 29.Qe7# {1-0}
Now I obviously did not play a great game. It was just down time playing bad bots and having fun, but this “200” ELO bot played 10 moves of opening theory in the Sicilian.
Back when I was rated 700, I knew maybe 3 moves of 2 variations of the Sicilian.
Who designed a 200 ELO to know 10 moves of theory?
If it’s just that I’m bad at chess, feel free to say that too lol. I will not have hurt feelings
By the way, here is your game in a video:
[Event "Computer Game"]
[Site "Chess.com iPhone"]
[Date "2023.11.03"]
[Round "?"]
[White "michaelpowers8"]
[Black "Oscar"]
[Result "1-0"]
[FEN "rnbqkbnr/pppppppp/8/8/8/8/PPPPPPPP/RNBQKBNR w KQkq - 0 1"]
[WhiteElo "1822"]
[BlackElo "200"]
[Termination "michaelpowers8 wins by Checkmate"]
1.e4 c5 2.Nf3 e6 3.d4 cxd4 4.Nxd4 Nf6 5.Nc3 d6 6.Be3 a6 7.f3 Qc7 8.Qd2 Nc6 9.O-O-O Be7 10.g4 Nd7 11.h4 Nce5 12.g5 h5 13.gxh6 gxh6 14.Bxh6 Nc6 15.Rg1 Nd8 16.Bg5 Bxg5 17.Qxg5 Nc6 18.Qg7 Rf8 19.f4 f6 20.Nxe6 Qd8 21.Nxd8 Nxd8 22.Rxd6 Nf7 23.Bc4 Nb8 24.Qxf6 Ra7 25.e5 Nxd6 26.Qxd6 Be6 27.Bxe6 a5 28.Rg7 Rxf4 29.Qe7# {1-0}
Now I obviously did not play a great game. It was just down time playing bad bots and having fun, but this “200” ELO bot played 10 moves of opening theory in the Sicilian.
Back when I was rated 700, I knew maybe 3 moves of 2 variations of the Sicilian.
Who designed a 200 ELO to know 10 moves of theory?
If it’s just that I’m bad at chess, feel free to say that too lol. I will not have hurt feelings