It depends on the level. The big early gains to get to let's say 2200+ were basically exclusively tactics. You really have to get your board vision to quite a decent level before the other parts of learning really start kicking in, and that's going to take thousands of hours of tactics. An important part of tactics is to ensure you actually work out the entire line - instead of just playing the obvious tactical idea, like Bxh7 or whatever, and going from there.
This was paired with 'analyzing' games of classical master (Alekhine, Capablanca, and so on) and then comparing my analysis against master analysis in plain English. Older books like Alekhine's best games books, or the various master vs amateur books (Euwe's is amazing) are a great resource here. Chessbase also now has a feature 'replay training' built in that would be really good for this. Load up a game, click on replay training, and you can basically play guess-the-move with automatic feedback in terms of how your move compares to the game as played (without revealing that move), as well as the top computer move.
It's also important to start appreciating typical piece repositioning ideas - for instance the rook lift is something that isn't very intuitive at first but radically reshapes many positions. If that rook on a1 heads to a3 and then on over to g3, a quiet Italian position can go from uninspiring to an unstoppable kingside attack really fast. The same is true of all the other pieces - for instance in the typical Najdorf structure with pawns on e5 and d6, knights on f6/d7, and bishop on e7 - if that bishop can go from e7 to e.g. b6 (so long as d6 remains solid enough), it can suddenly become a monster piece.
A key is to avoid excessive opening study until much later. It's a trap because you can spend an infinite amount of time learning openings, and you will get some wins without ever even leaving book, which feels pretty dang rewarding, but in the longrun it will stunt your growth. I remember at one point, sometime around 1800, delusionally thinking that the main difference between me and Kasparov was his encyclopedic opening knowledge. A quick glance at Hikaru doing puzzle rush will emphasize that's not quite right, of course this was long before GMs streaming was a thing! On the equal but opposite side, I also would recommend avoiding 'system' openings as a means of not having to worry about openings because the ideas and plans you see and learn in classical openings help improve your understanding much more than seeing an e.g. Colle each and every game.
Very good of you to get back to me with such a generous reply.
I'm happy reading it. I'd studied and played (in a club, real tournaments, etc) pretty "seriously" for about six months a couple of years ago, before stopping completely, even though it was going quite well, when some unavoidable life things got in the way.
Progress had been steady, climbing up to almost 1600 classical (before the statistical ratings hike last year, putting me on 1750). I was maybe going to at least temporarily settle around there, or slightly higher, it looked like. I'd played for a few years when I was younger, but didn't have serious coaching, unfortunately, so this was not 6 months from scratch at all.
What I'd come up with when studying tracks with some of your main points above. I was focusing very much on tactics, because I love them and because it seemed unavoidable - even the high-rated players who don't consider themselves "tactical players" still have an extremely solid tactical reading of any position, and spot all the usual patterns with ease.
I'd even mostly ignored heavy opening work! Which, I must say, is not a popular approach. A friend who'd seen a bit of your game would come up to you after a match in a tournament and say: "I can't beleve you played that on move six against the French defence, that's not the best move!", and I'd struggle to convince them that openings were not my main area of concern aha.
Anyway, thanks again, and congrats on your chess accomplishments - the tournaments truly are wonderful to be involved in, I find.
Yeah, exactly on the tactics issue. Computers used to be quite horrible positionally, but were still extremely strong simply because their short-term tactical vision was nearly perfect. And you really need tactics to enable positional play. Petrosian was able to snuff out tactical possibilities so effectively only because he was fully aware of where they were!
Now a days I think the tactics streak offered on Lichess (and probably on chess.com as well) is a really great tactical resource. It's the untimed option, rather than the much more popular tactics rush where you have e.g. 5 minutes to do as many as you can. I think the 'woodpecker' method of tactics is a great idea. Basically you build up on a repeated series of tactical problems, until you're able to complete them all perfectly accurately at an extremely high rate of speed. This seems contradictory, because the sites have massive tactics database, but you will regularly see the same problems due to the birthday paradox. If you're looking for some achievable ballparks, on Lichess I tend to be able to hit around 50 somewhat regularly, with a high of about 90.
I think the fundamental thing with openings is that so long as you make logical moves, even if you make an objectively weak move - you will very rarely reach a losing position because of it. And from that point on both your opponent and you are both out of book, so whoever understands the position better, and plays better, will win. Many people, especially adults, get caught up in obsessive opening study because it's the one form of chess study where results can be immediately felt.
The one very good thing about opening study is seeing ideas and concepts that you may not otherwise be able to come up with on your own. Like in the Najdorf, the pawn structure with d6+e5 and d6 generally firmly blocked and on a semi-open file leaves d6 feeling like a major weakness at first. The fact that it's generally rock solid was a serious eye-opener for me! And that translates strongly to many other positions - backwards pawns are not necessarily dooming one to passive defense, and can even be a great dynamic weapon!
Summaries of these are being added to my chess-learning files :) I will be doing Lichess tactics streak, the forced element of "can't lose" sounds like exactly what I need to be that little bit more hotly engaged.
I love the Woodpecker method! I went all in on that during my six-months of study, and had (anecdotal) positive results. I made flashcards of all the positions, and worked through them all 3 or 4 times, trying to go faster and faster. I also looked up a big list of common checkmating patterns, and put them on flashcards, and gave them names and everything to make them memorable, and drilled them.
This culminated in my last tournament, a rapid 12'3'', and me beating my first ever 1800s and 1900s, and performing well above my I think 13 or 1400 rapid rating of the time. It literally happened in two or three of the games that I'd be looking at the exact pattern I'd drilled, and then looking at the 1700 or whatever opponent and going: aha, the method works, here's the pattern.
What you say about openings tracks with what I was doing, except I went a bit mad at one stage and started learning loads of ridiculous gambits and getting smashed by anyone half-decent. I like violent positions. I'd some spectacular wins, but I think it was a silly strategy. At one stage a frind from the club destroyed me after I played some dubious gambit as black, and he said "yeah, gambits are good fun, but maybe for bullet chess online".
At a certain point too, in the process of "getting good", it's my feeling that everyone must eventually accept the quiet positions, and the slow endgames, and working hard for a draw with no story to tell afterwards. I was getting to the level where I had to accept that, but still struggled... I still would be tempted to do things that I literally knew were unsound, and would say: "oh come on, it makes no sense", but I might do it anyway, or a variant of it. Anyway, my thinking would be heavily clouded and biased by this desire for winning in the middlegame, and avoiding the slowness!
You can easily make a sharp/tactical repertoire of sound openings. As black you can play the Najdorf and King's Indian Defense. As white - open sicilian, winawer french, advance caro kann, be3/qd2/f3/o-o-o stuff against modern/pirc, bd3 bd2 o-o-o against scandi. e4e5 is the toughest nut to crack, but the evan's gambit is generally sound - Kasparov even beat Anand with it. Against the petroff you can play Nc3+o-o-o stuff.
That should just about cover everything!
But yeah, one thing you have to do to really start improving alot is to always assume your opponent will play the best move. Hope chess is how you ruin your own position!
This was paired with 'analyzing' games of classical master (Alekhine, Capablanca, and so on) and then comparing my analysis against master analysis in plain English. Older books like Alekhine's best games books, or the various master vs amateur books (Euwe's is amazing) are a great resource here. Chessbase also now has a feature 'replay training' built in that would be really good for this. Load up a game, click on replay training, and you can basically play guess-the-move with automatic feedback in terms of how your move compares to the game as played (without revealing that move), as well as the top computer move.
It's also important to start appreciating typical piece repositioning ideas - for instance the rook lift is something that isn't very intuitive at first but radically reshapes many positions. If that rook on a1 heads to a3 and then on over to g3, a quiet Italian position can go from uninspiring to an unstoppable kingside attack really fast. The same is true of all the other pieces - for instance in the typical Najdorf structure with pawns on e5 and d6, knights on f6/d7, and bishop on e7 - if that bishop can go from e7 to e.g. b6 (so long as d6 remains solid enough), it can suddenly become a monster piece.
A key is to avoid excessive opening study until much later. It's a trap because you can spend an infinite amount of time learning openings, and you will get some wins without ever even leaving book, which feels pretty dang rewarding, but in the longrun it will stunt your growth. I remember at one point, sometime around 1800, delusionally thinking that the main difference between me and Kasparov was his encyclopedic opening knowledge. A quick glance at Hikaru doing puzzle rush will emphasize that's not quite right, of course this was long before GMs streaming was a thing! On the equal but opposite side, I also would recommend avoiding 'system' openings as a means of not having to worry about openings because the ideas and plans you see and learn in classical openings help improve your understanding much more than seeing an e.g. Colle each and every game.