Chessical: This variation was played three times in the match with Schlechter as Black. <9..Bb7> was according to Hoffer a recommendation of Tarrasch. Tarrasch obviously believed in the set up as he played it here: S Takacs vs Tarrasch, 1922
Schlechter plays sharply, and in fact plays <g5> more readily than the famed attacker Spielmann did in the position:
Bogoljubov vs Spielmann, 1926. This avoids an exchange of Queens, but weakens his K-side whilst his King is still on its original square.
Spielmann probably did not want to defend the sacrificial line which the player themselves discussed after the game:
<12. Bg3> Nxe4 13.Nxe5 dxe5 14. Bxe5 f6 15. Qh5+ Kd7 16. Qg4+ Ke7;
Schlechter, however, has confidence in his ability to calculate the position exactly, and not to be unnecessarily frightened by the pin on the <e> file.
<24.f3?!> is not the type of compromising move you associate with Tarrasch
<13... gxh4?> damages the K-side unduly, <14. Nc3> Bxd6 15. Nxe4 O-O 16. Nd4 with a strong attack.
The players later pointed out "a curious winding-up of the game" with <18. Rad1> Rhd8 19. Rxd6 Rxd6 20. Rd1?? Nxc3 21. Rxd6 cxd6 22. bxc3 Bxf3 23. gxf3 a5; "and the Pawn marches unhindered to Queen"
<20. Nfd4?!> is inferior <20... f3> 21. gxf3 gxf3 22. Nxf3 Rf8; with advantage to Black.
<24.Nxg4> was simple and should have kept near equality which is as much as Tarrasch was entitled to here. Contemporary annotators do not comment on Tarrasch's plan with the <f> pawn, perhaps because Schlechter now seems determined to draw immediately rather than play on. Exchanging his powerful Bishop for the Knight