Uhum. Please read IM Palliser in my previous post and not IM Cox.  

My comments were rather tongue in cheek of course. Frankly I think Black should memorize just the first 9 moves (evt. with the improvement 9...b5). After that Black's play is quite simple. It is up to White to untangle his position.

Wasn't this game published completely in Lane's column?
OK, here is the full score - with a few thoughts.

Winsemius - MNb, corr NBC 1994
1.Nc3 d5 2.e4 dxe4 3.Nxe4 Nd7
With the intention to play an improved version of the Caro-Kann: compare 1.e4 c6 2.d4 d5 3.Nc3 dxe4 4.Nxe4 Nd7.

4.Bc4 Nf6 5.Bxf7+
Here I got my pants scared off. I had never seen this coming. Big luck for me, I remembered a similar line in the Philidor: 1.e4 e5 2.Nf3 d6 3.d4 Nf6 4.Nc3 Nbd7 5.Bc4 Be7 6.Bxf7+ etc. So I took up Pachman's excellent Moderne Schachtheorie I. Offene Spiele (1980), page 193 and decided to model my play to his lines.

5...Kxf7 6.Ng5+ Kg8 7.Ne6 Qe8 8.Nxc7 Qg6 9.Nxa8 b6
End of the similarity: 9...Qxg2 10.Qf3. But I also remembered some lines in the infamous Dracula-Frankenstein Variation, were White takes a rook on a8 and Black plays 9...b6 (thanks, Mr. Harding, for writing a book on the Bishop Opening in the early 70's). Simple guy as I am, I never considered 9...b5.

10.Nxb6 axb6 11.g3 Bb7 12.f3 e5 13.d3 Bc5
Again just simple chess. I already knew, that it is often a good idea to put your pieces on natural squares.

14.Qe2 h5
Activates my last piece and moreover: I have a plan! Really, since a couple of moves I felt very comfortable.

15.Be3 Bxe3 16.Qxe3 h4 17.g4 Nxg4 18.Qe2 Qe6
I hardly know anymore why I played this one; probably with the idea to have Rh8-h6-f6 available.

19.Nh3
The beginning of a remarkable detour.

19...Nf6 20.Ng5 Qd5 21.c4 Qa5+ 22.Kf2 Nh5
This one I understand again: the knight wants to go to f4. It does not happen.

23.Ne6 Rh6 24.Nd8 Ba8 25.d4 b5
0-1 as the knight drops.
I hardly can imagine, that IM Cox finds this game useful. Still I would love to read his opinion. Please remember two things:
1) the game was played in the lowest class of the NBC competition. In these days I (and my opponent) understood even less of chess then we do now. I played in absolute beginner fashion: my notebook does not have one single variation analyzed. I only tried to place my pieces as active as possible.
2) the game was played before silicon era.

IM Cox obviously does not read Dutch magazine Schaaknieuws, where I had published the game before. Since 1994 I think this line an excellent independent way to meet 1.Nc3 and also very suitable for players of the Scandinavian: 1.e4 d5 2.Nc3. I was happy to read, that 1.Nc3 guru DD van Geet agrees, that this exchange sac is correct. Of course he pointed out, that he had met this 30 years before. But hell, what did I know? Van Geet prefers 5.d3 and as far as I know, so does Keilhack. It does not strike me as particularly dangerous.

IM Cox? 

Anyone know how this game's progressing please? Like Keilhack, I believe that theory has been wrong to condemn this dangerous exchange sacrifice, as played by MNb. Incidentially, why is 1 Nc3 so rare, compared with the correpondence world, otb? It strikes me that White can often gain an aggressive and not particularly theoretical game. Furthermore, Keilhack's coverage is truly excellent.