Game 5: Both are out of book very early. The game develops to a closed position, the first pawn exchange is at move 26. It seems to end in draw by move repetition, but the Challenge avoids it because it evaluates its position as a little better. The risc-taking is then punished by the Rebell with 55... Qe4! and white is in trouble.
The funny thing is that the Challenge evaluates its position about -0.3 to -0.4 until 63... exf2. Probably a horizon effect?
General remark: A problem (bug?) I saw before: even when it thinks it has the slightest advantage, it tries to avoid move- or position-repetition at almost all cost, and so deliberately makes a less good move that sometimes even leads to a loss.
On the other hand, I also saw games where some kind of self-learning developed: when one way of developing the advantage got stuck and move repetition threatened, it tried another one, until it did find the winning strategy.
Game 6: Again both are early out of book. Black handles the opening phase a little better, but the game stays rather equal till move 30. Then white makes two weak moves, and black suddenly starts a decisive attack with 32... Dc2! White must lose material, but with 33.Kg2 he chooses again not the best move, and 33... Ng5 immediately decides the game.
Krypton Challenge (Tibono settings) - Mephisto Rebell 5.0 3 - 3
(so far all games ended in a win for black!?)
[Event "Test Tibono settings"]
[Site "Eindhoven"]
[Date "2020.03.20"]
[Round "5"]
[White "Krypton Challenge, 9/10/12/8/8/11/9."]
[Black "Mephisto Rebell 5.0"]
[Result "0-1"]
[ECO "D03"]
[Annotator "Wiselius,Paul"]
[PlyCount "130"]
[EventDate "2020.??.??"]
1. d4 Nf6 2. Nf3 e6 3. Bg5 d5 {**} 4. e3 {**} h6 5. Bh4 Bb4+ 6. Nbd2 O-O 7. c3
Be7 8. Bd3 Nc6 9. O-O Bd7 10. Qb3 Na5 11. Qc2 a6 12. Ne5 Nc6 13. Ndf3 Nxe5 14.
Nxe5 Bb5 15. Bxb5 axb5 16. Qd3 Qe8 17. a3 b6 18. Rfe1 c6 19. Qe2 Rc8 20. Bg3
Ne4 21. Bf4 g5 22. Bg3 Nxg3 23. hxg3 f6 24. Nd3 Qg6 25. e4 Rfe8 26. exd5 exd5
27. Qe6+ Qf7 28. Qf5 Bd6 29. Nb4 Bxb4 30. axb4 Rcd8 31. b3 Kh8 32. Rad1 Ra8 33.
Rxe8+ Rxe8 34. Ra1 Kg8 35. Qf3 f5 36. Qd3 Qe6 37. Qb1 Kh8 38. Qa2 Qe7 39. Qc2
Qf7 40. Qd3 Qe6 41. Qf3 Qf7 42. Kh2 Kh7 43. Ra6 Re1 44. g4 Qc7+ 45. g3 f4 46.
gxf4 Kg7 47. Kg2 gxf4 48. Kh2 Qe7 49. Kh3 Qb7 50. Ra2 Qf7 51. Kg2 Rb1 52. Ra3
Rc1 53. Ra6 Qb7 54. Ra2 Qe7 55. Ra8 {
? probably the decisive mistake. 55.Dxf4 was needed.} (55. Qxf4 Rxc3 56.
Ra6) 55 ... Qe4 {! and black takes its chance immediately.} 56. Ra7 + Kg6 57. Qxe4 + dxe4 58.
d5 cxd5 59. Rd7 Kg5 60. Rxd5 + Kxg4 61. Rxb5 f3 + 62. Kh2 e3 63. Rxb6 exf2 {
the striking thing is that white did not see a major disadvantage until this moment. Presumably
a horizon effect?} 64. Rg6+ Kf5 65. Rf6+ Kxf6 0-1
[Event "Test Tibono settings"]
[Site "Eindhoven"]
[Date "2020.03.21"]
[Round "6"]
[White "Mephisto Rebell 5.0"]
[Black "Krypton Challenge, 9/10/12/8/8/11/9."]
[Result "0-1"]
[ECO "E01"]
[Annotator "Wiselius,Paul"]
[PlyCount "120"]
[EventDate "2020.??.??"]
1. d4 d5 2. c4 e6 3. Nf3 Nf6 4. g3 Be7 5. Nbd2 {**} c5 {**} 6. cxd5 exd5 7.
dxc5 Bxc5 8. Bg2 Bg4 (8... Bxf2+ 9. Kxf2 Ng4+ 10. Ke1 Qb6 11. e3 Nxe3 12. Qe2)
9. O-O Nbd7 10. Nb3 O-O {
?! surprising that black allows the exchange of it's bishop} 11. Nxc5 Nxc5 12.
Ne5 Bh5 13. Bf4 Rc8 14. Rc1 Qa5 {black has a light advantage} 15. Bh3 Ne6 16.
Rxc8 Rxc8 17. Bxe6 fxe6 18. a3 Qb5 19. Nd3 a6 20. Be5 Nd7 21. Bc3 Bg4 (21... e5
{?!} 22. Nxe5 Nxe5 23. Qd4 Rxc3 24. Qxe5 Bxe2 25. a4 Qc6 (25... Qd3 26. Qe8#)
26. bxc3 Bxf1 27. Kxf1 {=}) 22. Re1 Qc4 23. Nf4 Qc6 {?} 24. Qd4 {now white is in charge}
Nf6 25. e3 Qd7 26. Nd3 Qc7 27. Rc1 ({better} 27. Ne5) 27... Ne4 28. Nc5 Bf5 29.
Nb3 Qd7 30. Bb4 {? not good} Rxc1+ 31. Nxc1 Qc7 32. Nd3 {?} (32. Bc3 {
was necessary}) 32... Qc2 {!} 33. Kg2 {
not the best, but material loss is unavoidable anyway} Ng5 {
and suddenly white's position is hopeless.} 34. Nf4 Be4+ 35. Kf1 Nf3 {
threatens Nxh2 with a quick mate} 36. Qxe4 Qc1+ 37. Kg2 dxe4 38. Nh3 Qxb2 39.
Bd6 Ne1+ 40. Kf1 Nc2 41. a4 Qa1+ 42. Kg2 Qd1 43. Bc7 Ne1+ 44. Kf1 Nd3+ 45. Kg2
Qxa4 46. Bb6 Qb5 47. Bc7 Qc5 48. Bd8 b5 49. Nf4 Nxf4+ 50. gxf4 b4 51. f5 {
here Rebell resigned, but I preferred to give the Challenge the satisfaction
of mating him ;-)} b3 52. fxe6 b2 53. Bh4 b1=Q 54. e7 Qh5 55. e8=Q+ Qxe8 56.
Bg3 Qeb5 57. Kh3 Q1f1+ 58. Kg4 a5 59. Kh4 Qf5 60. h3 Q1xh3# 0-1