A game that I liked (ChessBase 13)
[Event "Internet (Talkchess Forum)"]
[Site "?"]
[Date "2010.??.??"]
[Round "?"]
[White "Rusz, Árpád"]
[Black "(Study 42)"]
[Result "1-0"]
[SetUp "1"]
[FEN "2r5/6k1/8/5PP1/2p5/8/5R2/4K3 w - - 0 1"]
[PlyCount "27"]
[EventDate "2010.??.??"]
1. Kd2 c3+ (1... Ra8 2. Kc3 Ra3+ 3. Kxc4 Rg3 4. f6+ Kf7 5. Rf5 $18) 2. Kc2 (2.
Kd3 $2 Kf7 $1 3. Rg2 Kg7 $1 4. Rf2 Kf7 {positional draw}) 2... Kf7 (2... Rc5 3.
f6+ Kf7 4. g6+ $1 {chameleon echo} Kxg6 5. f7 $18) 3. Rg2 $3 ({Thematic try:}
3. Rf3 $2 Rc5 4. g6+ Kf6 5. Rg3 Rc8 (5... Rc7) 6. Rxc3 Ra8 (6... Re8) 7. Rc5
Ra3 8. Kd2 Rb3 9. Ke2 Ra3 $11) 3... Kg7 (3... Rc5 4. g6+ Kg7 5. f6+ $1 {
chameleon echo} Kxf6 6. g7 $18) 4. Rg3 $1 Rc5 5. f6+ Kg6 6. Rf3 $1 {the rook
finally steps on that square} Rc8 7. Rxc3 Ra8 8. Rc5 Ra3 9. Kd2 Rh3 10. Ke2 Rb3
11. Re5 $1 Ra3 12. Re7 $1 Kxg5 (12... Ra2+ 13. Ke3 Ra3+ 14. Kf4 Ra4+ 15. Re4
Ra1 16. Re5 Ra4+ 17. Ke3 Rg4 18. Rd5 $1 Ra4 19. Rd8 $1 $18) 13. f7 Ra8 14. Re8
{The main idea of the study lays in the small but very important difference
between 3.Rf3? and 3.Rg2!! followed by 4.Rg3. Only in the second case there is
later a winning rook manoeuvre (Re5!-Re7!). The positional draw is nice and I
like the echo pawn sacrifices (4.g6!, 5.f6!) too in the side variants.} 1-0