Opening, Middle-game, or Endgame, Which is most critical for improving your chess and why?

You’ve never taught, have you? I’ve been teaching since the 1980s. I’ve taught students aged 3 to 75. Subjects have ranged from chess to theology, but most has been history. Chess has been a constant on a part-time basis for a quarter century.

Lessons are tailored to the learner, but I have not yet met the student who did not benefit from instruction in endgames, including basic checkmates.

A strong focus on endgames also was instrumental in lifting me from mid-1400s to high 1900s in my 40s and early 50s.

You're jumping from "everyone needs instruction in the endgame" (not necessarily true!) to "it is therefore vital that we start off by teaching endings".

Is that high 1900s USCF? Let's see, mid 1400s is about 1350 FIDE maybe, which was about 95 BCF where we are in Britain. There were regional differences and I live in the middle of what was the strongest area for chess in the UK, back when Britain was one of the strongest countries internationally. I think that's far from being the case nowadays. Anyway, it meant that we were all undergraded and kept winning the county championships rather monotonously. I ended up by throwing away all my medals and awards, keeping only my two or three local chess club championship medals.

I was graded at 89 after the first year of playing and three years more saw me rise to 149, which will be about 1950 USCF equivalent, in my late 30s, so we have a strong equivalence. However, although I took care to learn mates like K+N+B and K+N+N vs K+p, I would think that my improvement was due mainly to opening knowledge and middle game ability. It was quite a lot later that I started to excel at endgames and I would say that it wasn't due to study but merely by working them out otb. I always had a logical and mathematical mind and good concentration.

All I'm really doing is pointing out that it isn't a general rule, that endgame study is vital. I won loads of tournaments ... honestly, too many to count and I never did count them. I remember on one streak winning a cash prize in eleven consecutive tournaments. Anyhow, I won most of my games in the middle game or the opening. Maybe 20% went to endings and when I was on form they came naturally to me, with little need to study them. I'm sure the same holds for many others and that by concentrating on teaching endings first, you might be putting off some of the more talented beginners.

This is a case in point, Ziryab, regarding why understanding development is far more important than endgame ability, because if a player cannot reach an endgame, what's the use of knowing how to play them? My opponent had a rating somewhere in the 1500s for blitz and yet he only completed his development on move 25, by which time my attack was in full swing and it was too late. Just a 5 minute game but still ...

The failure was tactical, not opening theory. Obviously, a failure to develop will create problems. All my students learn from Paul Morphy’s play. https://www.chessgames.com/perl/chesscollection?cid=1042867

The quickest way to impress upon a beginning chess player the search for truth is mastery of elementary pawn endings and simple checkmates.

I disagree that the failure was tactical. He played Marshall's Defence, which is bad. He wasn't able to develop because he didn't know how to play it.

Very often, such tactical failures are due to inability to develop. In my opinion, my argument is (or should be) convincing but if you don't find it so, at least I've set out an argument for "openings first", which many others will find to be very strong.

"Search for truth" seems overly dramatic! But a lot of people are playing Mashall's Defence just now. The popularity of the Dutch and the Slav, among people who don't know how to play them, has diminished, at least for the time being. They will play Marshall's Defence because it has a name and so is probably considered "book" by chess.com, who also don't really understand the priorities of teaching chess. "If it's book, it must be ok" but of course, it isn't ok at all.

I think that what you seem to be missing is not that he was comprehensively outplayed but that in such positions, it's much easier to outplay opponents than it is if they're sensible enough to follow sound opening principles, Ziryab.

I would say middle game is the most crucial since that's where you shape your endgame in the future, and it mostly depends on middle game if you are having a advantage or not

This is another example, Ziryab, again, just played 5 mins blitz. No attempt to develop the c8 bishop and yet he initiates tactical operations. All my moves are positionally well motivated and I'm able to set up a simple trap which looks like I'm just defending my a pawn with the bishop. Perhaps it's your insistence on "endings first" that's causing it all. You aren't a stronger player than I am ... in fact I would think it's the other way round. Using ancient GMs and the Russian school as authority for a very dodgy, doctrinaire approach isn't helping anyone play chess better. You may think you're teaching them the right way but I'm pretty sure you aren't. At least you should give it some thought.

Of course I’m not a stronger player. The rating difference means nothing.

The OP wanted to know which aspect of the game was most viral for improvement. No one said to neglect any phase, except you. You want to completely dismiss the endgame. The strongest players who have posted here, OTOH, have emphasized the endgame as most vital, while noting that tactics are also important. Of the three phases—opening, middle, and ending—openings are the least vital to the improving player, at least until one is battling masters on a regular basis. But, no phase of the game should be neglected.

My teaching the past quarter century has nurtured young players and a few adults in all phases of the game. My best students have been strong at tactics, followed sound opening principles, and have beaten higher rated players in tournament play because they were strong in the endgame. My best students grew stronger than me. That only happens when the foundation is solid. 
There is no foundation more solid than one built upon the curriculum José Capablanca outlines in Chess Fundamentals. Elementary checkmates, basic pawn endings, tactics, a few opening principles, then repeating the sequence with slightly more advanced training. This cycle has been my approach for more than twenty years. Some of my students have won state championships.

You’ll excuse me if I do not heed the advice of someone much lower rated in blitz than me, even if he is the better player.

Forget about blitz. Your last game that was a battle was a win where your tactical prowess gave you a superior ending. https://www.chess.com/game/live/78847503335 

In that game, you overcame a little opening problem that prevented you from castling and bringing your king’s rook into play.

Also, your rapid rating offers a better claim that you are a decent player, as you are in the top 1% on the site.

I'm not much lower rated than you. Completely ridiculous thing to say. Anyway, I'm a lot older than you are, by maybe about 13 years, at a time when we start to slow down. I've been 2225 in Daily on here but my mother died or something like that and I abandoned that account. I no longer have an interest in Daily but do you still play it? My rating's still about right but yours won't be!

Also Ziryab, here's another one. Again, just finished 5 mins blitz where I'm black and I complete my development on move 30. You might find this one instructional. It contains two quite decent tactical shots. On this occasion, my opponent was too keen on development, when to undevelop might have saved his game. That is, to retreat a knight, maybe. There's some indication that after a really long layoff from blitz, I'm getting at least some of my old ability back. I used to play blitz against people who were 2300 FIDE and over and I could hold my own against them. You've never been able to do that. I don't know why you keep wanting to get into arguments with me, instead of accepting that we each have our viewpoints. Stop always trying to assert yours as always being right, because it isn't!

That was a blitz game. 10 minutes each is blitz. Not my fault that chess.com don't know their socks from their mittens. We have a much ruder one than that but it would get me banned. If you think about it, 10 mins has much more in common with 5 mins than with 60 mins or 20 mins, which are real rapid speeds. 10 mins isn't rapid except to a dodo.

When they changed 10 mins to rapid without telling me, I hadn't played real rapid for months. I probably lost three games because it so happened I wasn't playing well at 10 mins. Then I noticed I'd lost over 100 points off my rapid rating for losing about three times, which is pretty disgraceful of tham really, for doing that. Anyway, I persevered with 10 mins for a while but I hadn't been playing much because my father had died and I was very busy sorting it all out, so I decided when I was less busy to play 5 mins blitz to practise playing fast chess. I've played around 5000 games in the past year and it's doing me a lot of good. Sometimes I gain a couple of hundred points in a few days and then lose them again but my average is creeping up. I'm not really a blitz player any more and always preferred slow chess. My favourite speed would be two and a half hours each for all the moves. I strongly dislike increments. I suspect they're just a fad and they'll lose their popularity.

This is me playing positionally, against the Stonewall Attack, in a three-days a move Daily game. I think the chess engine gave me 92.8% accuracy but you can ignore it. Black plays nearly perfectly. I've checked it and I know that. There was at least one occasion where the engine would have thrown away the win with its "your move was ok but mine is much better" garbage. Since when has an engine like the one they have here been able to positionally outplay a human?