Chess champion Boris Spassky once expressed the thought that he would like to learn chess over again and this time learn it the right way.
The human effort to learn chess is shaped in part by its cold mathematical nature as an intrinsically difficult problem: definitive evaluation of a given position requires it be solved to all possible terminal positions.
Thus it makes sense to start learning chess at the ending. There’s the purely human consideration that it’s easier to internalize the powers of the pieces when seeing them act in solo or in small groups on an open board among reduced material. And then there’s the austere science of converting a position step by step to the final win, working back towards to the midgame in one’s studies.
General trends in teaching chess seem most askew from the game theory realities in the study of the openings.
The older notion of the openings as narrow trails mapped out in a huge, dark forest doesn’t apply to modern times. With and without the aid of computers chessplayers have broadened the ground being regularly trampled until chess is less like a mysterious forest than an overloved and over-visited public park. Opening positions are revealed to be a very broad spectrum of related positions shading off this way and that, like the iconic Hindu picture of the Transfiguration of Krishna when he is revealed as all deities fading off to the vanishing point.
Openings used to be said to “transpose”, which is fatuous, since a move order is not headed to a specific place on the strength of possessing a name assigned to the first few half-moves on the basis of some historical personage, event, or even joke among chessplayers. There are simply many paths to similar or identical positions. There certainly are an amazing number of move orders to get to the tabiya we call the Maroczy Bind, paths arising in openings as varied (in the view of our lame classification schemes) as the Sicilian, the Ruy Lopez, and the Modern. An analogous band of converging tabiyat is clustered towards the center of the Neo-Grünfeld spectrum.
Reading criticism of World Champion Carlsen’s handling of the problem of opening study suggests strongly that he takes much this view and doesn’t concern himself with classification as much as absorbing the spectrum.