Tuesday, December 29, 2009

Friedel's Fritz-Ulvestad Wins Again

MacKinnon - Friedel, Edmonton 2009
Black to Play and Win

I have annotated the game MacKinnon - Friedel, Edmonton International Tournament 2009, where GM Josh Friedel continued his winning ways with the Two Knights Defense, Fritz-Ulvestad Variation (5....b5), to which he has returned since his loss to Nakamura with the more traditional 5...Na5 line. The line gave him an important point on his way to a tie for first in the Edmonton International Tournament earlier this month. His opponent was 16-year-old Canadian expert Keith MacKinnon of Saskatchewan, who commented on the game at his blog: "I didn't want to get slowly outplayed by a stronger opponent in my game against GM Josh Friedel, and so I tried to follow the game that Nakamura won against him at the US Championship this year. He played a slightly different line which I had looked at (but not nearly enough to play it against a GM in such a sharp position.) I lost quickly since my intuitive thirteenth move was actually a pretty big mistake." Actually, theory suggests that it was his 12th move that was the problem, and there followed a series of small errors that made Black's win look easy.

Related Links


katar said...

check out the game Zierk-Friedel, 29 Dec 09. Friedel played 6...h6(?!) and white retreated 7.Nf3(?).

Michael Goeller said...

Thanks for the update--it's surprising to see Friedel having to face this line again so quickly. He must recognize that opponents (like MacKinnon) are preparing this sharp stuff against him to avoid just getting GMed to death, and maybe he decided to risk 6...h6 in the expectation that his opponent would not have had time to study the classic Ulvestad line. Or maybe he had something new and interesting up his sleeve? There was an equally surprising Naiditsch - Carlsen game in 2006 which makes me wonder.... Naiditsch didn't play 7.Nxf7 either. Very interesting.

oddodddodo said...

Thanks for the two posts on Friedel's games, which of course are of great interest to me! I love to see a GM playing this line.

You mentioned in your notes somewhere that I am an advocate of the piece sac 8. ... Qh4 9. Ng3 Bb7!? 10. cd O-O-O in the Fritz. Which was true, but I had two losses with that line this year. You can't completely blame the opening in either case, but the results, combined with my analysis of those games with the computer, has cooled my ardor for the piece sac a bit. I am definitely planning to play 8. ... Ne6, as you recommend, next time I go into this line.

Dana Mackenzie

katar said...

You may have already known this, but Naiditsch was apparently "inspired" by the Carlsen game and tried this h6-Ulvestad line against Vallejo, who finally stepped up the plate with Nxf7 and won a miniature despite a lot of thrashing around by black.

Mark Ginsburg said...

The Ulvestad looks highly suspect.

I ran the MacKinnon game through the computer. It shows that in what is termed the 'main line', (a ridiculous position with black's king on d8),

6...Nd4 7. c3 Nxd5 8. cxd4 Qxg5 9. Bxb5+ Kd8 10. Qf3 Bd7 11. O-O Rb8 12. dxe5 Ne3, just 13. Qh3! wins. For some reason, 13. Qg3 gets a lot of attention but Mr. Rybka is really liking 13. Qh3. What the deelio? Why is 13. Qh3 not considered?

If 13..Qxg2 (forced, right?) 14. Qxg2 Nxg2 15. d4 Nh4 16. Bg5+ wins; on other 15th moves for black white is also just up material in a should-convert ending. Goodbye, Ulvestad.

In the recent Zierk-Friedel game with 6...h6, that looked highly suspect too! The computer just thinks 7. Nxf7 is an easy white win; he does nothing special, develops, and remains material up (much like Naka-Friedel).

Philosophically I think if Karpov and others do 3...Bc5! that is probably a good third move! Black can often attack later with O-O, Kg8-h8, Nf6-moves, f5, etc. (with the Bishop tucked away on a7).

The Ulvestad is one of those "going crazy" / "burning the bridges" openings that is totally unnecessary and usually provably unsound. C'est la vie.

Michael Goeller said...

Hi Mark --
Take a look at my notes to the game, based mostly on Jan Pinski's Two Knights book and Rybka/Fritz. As I indicate, he does give 13.Qh3 as the main move and follows your line -- but then continues 13. Qh3 Qxg2+ 14. Qxg2 Nxg2 15. d4 Nh4 16. Bg5+ Be7 17. Bxh4 Bxh4 18. Nc3 Bf3! (Pinski's analysis) 19. b3 Rb6 20. Bd3 Rg6+ 21. Bxg6 hxg6 with compensation, and he carries the analysis much further. I'd be interested if you can improve for White -- or if you think Pinski is wrong. This does seem to be the absolutely critical line of the Fritz-Ulvestad, IMHO.

Mark Ginsburg said...

Looks like White wins.

18. Nc3 Bf3 is the only move.

19. Rab1 Rb6 20. Bd3! to guard g6.

(I don't know why Pinsky is saying 19. b3?)

I don't know Pinski's book (I don't know Pinski either!) but if he is saying 20...Rg6 anyway after 20. Bd3, that does not look like it works.

For example,

20. Bd3 Rg6 21. Bxg6 hxg6 22. Rfe1 prepares Ne4 to defend against the possible Rh5.

(22...Rh5 23. Ne4 is a decisive edge for white).

If 22...Bg5 (better than 22...Rh5 at least) 23. Rbd1!! is lights out.
This brilliant defensive move prepares the crushing Rd1-d3.

Black has nothing better than to take, 23...Bxd1 24. Rxd1, and white should convert the point.
For example, 24..Bf4 is met by the exceedingly strong 25. Ne2!.

Conclusion: the "main line" Ulvestad is hopelessly unsound.

Mark Ginsburg said...

I collated the findings
here. I'll debunk the 6...h6 separately.