I suppose I'll put it in my discovered check collection since the combination sets up two key discovered checks. The first is 25. Rg4+! to set up a winning pin and the second one is a key part of the mate threat 36...Qc6 37. Rf6+ Kg5 38. Rg7+ Kh4 39. Rf3# which forced Black's resignation.

However, it integrates so many interesting secondary tactics (such as clearance, decoy, deflection, double attack, mate threat, pin, Queen sacrifices and a defensive combinations) I was tempted to include it in my combined operations collection.

26. Re1 h5 27. Rg7+ Kh6 28. Rxe4
but now 28. ... de 29. Bxe4 with the uncoverable mate threat on h7 has the problem Rd1#.

26. Re1 h5 27. Rxe4 dc and the bishop is under threat and square d1 weak. What did I overlook?

Rosition after 27.Rexe4:

click for larger view

<21. Nd7!!>

This brilliant decoy clearance (sham) sacrifice exposes Black's weak and underprotected castled King to a decisive attack.

<21...Rdxd7>

Whether Black accepts or declines the decoy Knight sham sacrifice, he loses:

If 21... Qxd7, then Wins after 22. Qxh6+ gxh6 23. Bxf6+ Bg7 24. Bxg7+ Kg8 25. Bxh6+ Kh8 26. Bg7+ Kg8 27. Bf6+ Kf8 28. Bh7 Qg4 29. Rxg4 Ke8 30. Bf5 Re6 31. Rg8+ Kd7 32. Rxd8+ Kc7 33. Bxe6 ;

If 21... Nxd7??, then 22. Qxh6+ Kg8 23. Qh7# follows.

If 21... Rexd7, then White wins quickly with 22. Qxh6+ gxh6 23. Bxf6+ Bg7 24. Rxg7 Qf4 25. Rh7+ Kg8 26. Rh8#.

Declining the offer with 21... Qf4 is relatively better, but White still wins a piece with a clear winning advantage after 22. Nxf6 Bc8 23. Rf3 Bxh3 24. Rxf4 dxc4 25. Bxc4 Be6 26. Ng4 Bxg4 27. Rxg4 .

<22. Qxh6+! gxh6 23. Bxf6+ Bg7 24. Rxg7>

Here, the discovered check threat leaves Black with nothing but losing options.

<24...Re4 25. Rg4+!>

Watch out Black! There goes that first dicovered check to force your next move.

<25...Kh7 26. f3!?>

This clever pinning combination wins slow but sure, but IMO makes it more difficult than necessary.

White can win quicker and easier with the strong reply 26. Re1!, when play could continue 26...Rde7 27. Bxe4+ dxe4 28. Rg7+ Kh8 29. Rd1 Ba6 (29... Qf4 30. Rd8+ Re8 31. Rxe8#) 30. Rxf7+ Kg8 31. Rxe7 Qb8 32. Rdd7 Qf8 33. Rg7+ Qxg7 34. Rxg7+ Kf8 35. b3 Bc8 36. Rxa7 .

<26... h5 27. Rg7+ Kh6 28. fxe4 dxc4 29. Bb1 b5>

Black can put up much more resistance with 29... Rd3, but White still wins after 30. Bxd3 cxd3 31. Rg3 Bc8 32. Bg5+ Kh7 33. Rf6 Bg4 34. Rxd3 b5 35. h3 Be6 36. Bf4 Qb6+ 37. Kh2 Kg7 38. Be5 Kh7 39. Rg3 Qd8 40. Rg5 Qd1 41. Rf3 Kh6 42. Rg8 f5 43. Rh8+ Kg6 44. Rg3+ Qg4 (44... Kf7 45. Rg7#) 45. hxg4 .

<30. e5 Rd3 31. e6 Qb6+ 32. Kh1 fxe6 33. Bxd3 cxd3 34. Rd7 c5 35. Bd8 Bxg2+ 36. Kg1! 1-0>

This final move refutes Black's attempt at a double attack swindle with 36. Kxg2?? Qc6+ 37. Kg1 Qxd7 and establishes a decisive double attack threat for White (Black must surrender the Queen for decisive material loss or get mated).

So, Black resigned in lieu of facing 36... Qc6 (36... Qxd8 37. Rxd8 Bxf1 38. Kxf1 c4 39. a4 a6 40. Rd6 ) 37. Rf6+ (here's that strong discovered check tactic again, helping force mate) 37...Kg5 38. Rg7+ Kh4 39. Rf3#.

Yet White could have side stepped this complication altogether with the stronger 26. Re1!, when the win is never in doubt.

So, he could have prevented this mess he got himself into with 18...Nd7.

click for larger view

Btw, Samisch was in some ways the Topalov of his time, a strong player with a remarkable wealth of beautiful losses to his colleagues, including Nimzo (of course), Alekhine, Capablanca, and probably several others to boot.

<Rob Morrison: It's all very pretty, but white could have just played 22. Bxf6 and it's over.>

That was my preference but I am not sure that is stronger than what Samisch played - he would have had to convert a long queen versus rook and minor piece ending.

<black was lost after his third move> Good to know.