Prior to embarking on a comprehensive analysis of my tournament games, I had not realized either the frequency with which I had actually faced the QGD, or the difficulties inherent in playing the English against it, rather than simply transposing with d4 to the main lines. (There of course are plenty of other difficulties involved in that, including the large body of opening theory.) As a result, I've now worked out a reasonably consistent approach involving an early e3, which I'm satisfiied with (if not completely happy). This should have better practical results than essentially randomly picking from the variety of other early move choices (4. g3, 4. b3 and 4. cxd5). As I noted in Annotated Game #64, the lack of such a consistent approach made it feel like I was playing a new, unfamiliar opening each time.
Going back to the actual game, White makes a number of small errors and one significant one on move 13. The engines' recommendation of 13...a5 I found instructive, showing how Black can use that type of pawn lever against White's queenside formation when it is left underdefended. Although Black retains a noticeable advantage, thanks to White's somewhat incoherent strategic play, White is smart enough to realize it and then manages to trade down into a drawn position.