Many thanks for sharing your games. The first game against De Jong appears particularly important. I am glad that you played 5.a3 N4a6! which I happen to prefer over N4c6 myself. Counterplay with c5 (or b5) makes sense here, at least to me. Surprisingly I find only 72 games with 5...N4a6 in the database, versus 254 with N4c6.
Just a few remarks on
De Jong - MNb. 1.e4 d5 2.exd5 Nf6 3.d4 Nxd5 4.c4 Most critical. Yet eventually we'll have to return to 4.Nf3. Playing for "two results" with 4...Bf5 must be sound, but it wouldn't help to make the variation popular. That said, I don't believe in Khalifman's antidote (in volume 3 of his Anand series, p. 325) 5.Nh4!?. It's a loss of a tempo after 5...Bd7 followed by e6 and c5 - nothing to worry about.
4...Nb4 5.a3Géza Maróczy was the first to recommend this reply in
Paul Morphy (1909): "Auf [4...] Sb4 antwortet Weiß am besten durch 5.a3. Fehlerhaft wäre es, mit 5.Da4+ auf Figurengewinn zu spielen, [...]" Funny how the Hungarian's comment, written in 1908, preceded the stem game of the system.
5...N4a6 6.b4One of the most successful continuations: 85% out of 10 games. Alternatives are 6.Nc3 (23 games, 91%), 6.Nf3 (27 games, 74%) and 6.Be3 (10 games, 80%).
6...Nd7!
I like this move. It could well represent the "solution" against 6.b4.
7.b5?! Black's idea of c7-c5 is illustrated by 7.Nf3 g6 8.Be2 Bg7, for example: (a) 9.Nc3 c5! 10.bxc5 Ndxc5 11.0-0 0-0 12.Be3 b6 13.Rc1 Ne6 and Black can be satisfied with his position.
(b) 9.Ra2 c6 10.Nc3 Nc7 11.0-0 a5 12.Qb3 0-0 13.Bf4 b5!, about =.
7..Nab8 8.Nf3 c5The immediate 8...a6 was more precise, about =.
9.Bb2White had several other options, e.g. 9.bxc6 Nxc6! 10.d5 Nce5 +=.
9...cxd4 10.Bxd4
10...f6Preparing a move that Stockfish plays at once: 10...e5!, about =.
For much of the remaining game, White had an edge.
MNb wrote on 07/21/17 at 02:11:52:
1. Every single book on the Scandinavian either dismissed it or neglected it.
2. The one exception is an obscure Swedish monograph from 1967 written by a certain Bo Jonsson.
3. Polugaevsky once said that no opening is better than one with a bad reputation enriched with new ideas.
4. About 10 years before a new idea occurred to me for the Scandinavian Marshall Gambit: if White prevents ...e5 then play ...c5.
5. I couldn't find a refutation of this idea when analyzing it with Fritz 3 (or was it 2 ?) so I assumed my opponents couldn't either.
Your first point is still true today. In 123 volumes of NIC yearbook, there are 53 articles on the Scandinavian Defence, but just five of them cover 2...Nf6, and none of those would help someone facing 4...Nb4.
I'll post more thoughts over the next days. Other duties interfere from 26-30th of July though.