| 
After blundering a pawn, white makes a speculative bishop sacrifice for a dangerous attack, ending this miniature with a nice mate.  | 
| 
 | 
1. Nf3 
 | 
My last game with the Reti Opening was weird, so I'll try it again. | 
   
    | 
|   | 
| 
 | 
1... d5 2. d4 
 | 
Transposing into a queen's pawn game. | 
   
    | 
|   | 
| 
 | 
2... c6 3. c4 dxc4 
 | 
An unusual time to accept the gambit, although the Slav Accepted does so on the fourth turn. | 
   
    | 
|   | 
| 
 | 
4. e3 
 | 
Threatening to win the pawn back; it seems to be the most common move here. | 
   
    | 
|   | 
| 
 | 
4... Qa5+ 
 | 
Probably an overly aggressive queen move; there is no real threat. | 
   
    | 
|   | 
| 
 | 
5. Bd2 Qd5 6. Nc3 
 | 
Continuing to pester the queen. | 
   
    | 
|   | 
| 
 | 
6... Qh5 7. Bxc4 
 | 
The queen hasn't done anything other than allow me to develop with tempo, and now I can retake the pawn with a significant developmental lead.  The bishop also eyes f7, although e6 easily breaks this up. | 
   
    | 
|   | 
| 
 | 
7... Nh6 
 | 
The f pawn doesn't need another defender, so why not Nf6? | 
   
    | 
|   | 
| 
 | 
8. O-O 
 | 
I'm unsure about castling kingside, but this seems the best way to continue development. | 
   
    | 
|   | 
| 
 | 
8... Bg4 
 | 
Perhaps the castling was too early; black threatens to open the g file. | 
   
    | 
|   | 
| 
 | 
9. Rc1 
 | 
I don't want to retreat my bishop to e2, and after 9...Bxf3 10. Qxf3 Qxf3 11. gxf3, it looks fine, since I still have a large lead in development and my rook can use the g file.  Therefore I develop my other rook. | 
   
    | 
|   | 
| 
 | 
9... Nd7 10. Qb3 
 | 
Unpins the knight, attacks b7, and adds pressure to f7. | 
   
    | 
|   | 
| 
 | 
10... O-O-O 
 | 
Protecting the b7 pawn. | 
   
    | 
|   | 
| 
 | 
11. h3 
 | 
I want to push d5, but I want a rook at d1 first.  I don't want the knight to be pinned, so I drive away the bishop first. Better would have been Nb5, with similar threats as later in the game; if 11...cxb5?, 12. Bxf7  wins the queen.  However, this turns out to lose a pawn. | 
   
    | 
|   | 
| 
 | 
11... Bxf3 12. gxf3 Qxh3 13. Bxf7 
 | 
If 13...Nxf7 14. Qxf7, it's still not optimal, but not too bad. If 13...Qxf3 14. Nb5 looks like a potential mating combination or significant material loss for black with 15. Nxa7 . | 
   
    | 
|   | 
| 
 | 
13... Qxf3 14. Nb5 
 | 
Sacrificing the bishop and threatening Nxa7 . | 
   
    | 
|   | 
| 
 | 
14... Nxf7 
 | 
Black could have declined the sacrifice with Kb8. | 
   
    | 
|   | 
| 
 | 
15. Nxa7+ Kc7 
 | 
15...Kb8 16. Nxc6  Kc8 (16...Ka8? 17. Qa4#) 17. Nd5  wins the queen. | 
   
    | 
|   | 
| 
 | 
16. Ba5+ Kd6 
 | 
b6 was the saving move; after 17. Nxc6, white can eventually also win the rook for the knight, evening up material. | 
   
    | 
|   | 
| 
 | 
17. Nb5+ 
 | 
I finally see a mate in two: 17...cxb5 (forced) 18. Bc7#. | 
   
    | 
|   | 
  |