@Sputnick; Thanks for putting up with this silly post.   I still do tactical training on a regular basis. I use Lev alburt's 300 pocket trainer; reinfield's 1000, and Polgar's Brick on occasion as well as CT-Art still ( I lost count on which circle I am on) I use it as a warm up before an event or before I begin evaluations of a game.

As for the comments on the game. I used my brain first as well and ran the Rybka analysis afterwards ( that's why there are some odd remarks mixed with my more "verbose" response). I've gotten into the practice of thinkign for myself first then having it checked with a better player or the computer.

I totally misplayed that position. I have a hard time when I am under attack ( both games were good examples of this) and not fall into a psychological trap. Interesting how during a game one can get tripped up and start thinking "OMG, I better play extremely defensive here." Instead, I should handle it more like " What is the position trying to tell me I need to do"  When the clock isn't running, and ratings aren't on the table, it's a whole different process.

As I read Bronstein's comments in some of the games in Zurich 1953, he comments on occasion how a particular move was "more of a psychological weapon more than a real positional threat" as the game shifts gear. Some masters fall into the same trap as a 1700 player!

After using my own brain, I like to check out my ideas with the computer. I checked the first few moves with Fritz, which does not find my move 8, 10 or 16 suggestions better than yours - they are very slightly less desireable in its eyes. Thus, the computer evidence is inconclusive on these moves. Yes, 14. Bxg7 is an error, but the variation is too complex for a most 1700s to figure out. Fritz does not think that 17. Rb1 is a huge error. It seems that you started to go wrong only after this move. 19. Nh2 was an error and 20. Re1 compounded it.

A few observations on the game.

I would prefer 8. Qc2 with the plan of putting the light square bishop on e2. This way your Bb2 is protected and the pin by black's light square bishop on g4 is not an issue.

Perhaps 10. h3 is better, forcing white to take the knight or evacuate.

I guess I miss what new chances black gets after you play 14. Bxg7. Whatever he plays, he is not getting more than two pawns for the piece. The question is what is better, taking the f-pawn or the h-pawn. I would prefer keeping the h-pawn as white.

Perhaps 16. Be2 is better, as it will allow white to play Nh2, removing one of black's key attacking pieces and giving the king the chance to hide behind the knight is necessary.

Interesting blog. Playing through old master games is certainly a more enjoyable way to improve than tactics torture. If you can stomach it, repeating tactics is probably a more efficient way to improve up to 2000, but if you are not enjoying it, it will not be effective.