End of this Study wrong

[Paloma]

[pgn][Event "Stronghold"] [Site "StrateGems"] [Date "2017.??.??"] [Round "?"] [White "Mihai Neghina"] [Black "---"] [Result "1/2-1/2"] [Annotator "?"] [SetUp "1"] [FEN "6k1/1qr1p2p/ppN5/3p1N2/6p1/1P1P2Pp/PP2PP1P/2RK4 w - - 0 1"] [PlyCount "14"] [EventDate "3.??.??"] {Inspired by V Chekover 1947 study (FEN: 7r/p3k3/2p5/1pPp4/3P4/PP4P1/3P1PB1/ 2K5 w - - 0 1)} 1. Ncxe7+ (1. b4 {?} Rxc6 2. Rxc6 Qd7 {-+}) (1. Nfd4 e5 2. Nxe5 Rxc1+ 3. Kxc1 Qc8+ {followed by Qf8 -+}) 1... Kf8 (1... Kh8 {same as mainline}) (1... Rxe7 2. Rc8+ Qxc8 (2... Kf7 3. Nd6+ {+-}) 3. Nxe7+ Kf7 4. Nxc8 {+-}) ( 1... Kf7 2. Nd6+ Kxe7 3. Nxb7 {+-}) 2. Ng6+ (2. Rxc7 Qxc7 3. Ng6+ Ke8 {-+}) 2... hxg6 (2... Kg8 3. Nge7+ {=}) (2... Ke8 $2 {or Kf7} 3. Nd6+ {+-}) 3. Rxc7 ( 3. Nd4 $2 Rxc1+ 4. Kxc1 Qf7 {-+}) 3... Qxc7 4. Nd4 Qe5 5. e3 b5 6. a3 {White has reached the desired defensive position. This is not only a fortress, it is an entire castle...} (6. Kd2 $2 b4 {-+}) 6... Kf7 7. Kd2 {any K move inside the headquarters E1-E2-D1-D2 is fine} (7. b4 {?} Qb8 8. Kd2 Qb6 9. Ke2 a5 10. bxa5 Qxa5 {-+}) (7. a4 {?} bxa4 8. bxa4 Qb8 {-+}) (7. Kc2 {?} Qf6 {-+}) 7... Kg8 {If Black tries nothing, White does nothing} (7... Qf6 8. Ke2 {When the bQ is on the f-file, wK guards f2 by shuffling between Ke1 and Ke2}) (7... Qc7 8. Kd1 {When the bQ is on the c-file, wK guards c1 by shuffling between Kd1 and Kd2} Qa5 9. Ke2 {When the bQ is on the a5-e1 diagonal, wK guards d2 and e1 by shuffling between Kd1 and Ke2;} (9. b4 {?} Qb6 {-+})) (7... b4 8. a4 {ab blockade} a5 9. Ke1 Qd6 10. Kd1 Qa6 11. Kd2 {When the bQ is on the a6-f1 diagonal, wK guards d3 by shuffling between Kd2 and Ke2} (11. Kc2 $4 Qf6 {11#}) ) (7... Qb8 8. Kd1 a5 9. Kd2 a4 10. b4 {ab blockade} Qb6 11. Kd1 Kf6 12. Kd2 Ke5 13. Kd1 Qxd4 $2 14. exd4+ {30#} Kxd4 $1 15. Kd2 {29#; wK is always close enough to defend d3} Ke5 {von mir} 16. Kc3 {28#} g5 17. b3 {27#} axb3 18. Kxb3 Kd4 19. Kc2 {25#} (19. a4 {26#} bxa4+ 20. Kxa4 Kxd3 21. b5 {24#}) 19... Ke5 20. Kc3 Kd6 21. Kb3 Kc7 22. a4 Kc6 23. axb5+) 1/2-1/2 [/pgn]
In the last Variation 7...Qb8 is the black move 13...Qxd4 by the Author wrong.

[d]8/8/1q4p1/1p1pk3/pP1N2p1/P2PP1Pp/1P3P1P/3K4 b - - 6 13
13... Qxd4 ? 14. exd4+ 30# Kxd4 15. Kd2
Incidental i let SF calculate at move 15.Kd2 and the score was rising.
It come out there was a matt in 29 (going in the Var. and then backwards analysis)

Better seems 13...g5! which hold the draw