A while ago on my blogger blog I received what I think is perhaps the best chess improvement advice I have ever received. I thought it might be of interest to Chess.com members, so I repeat it below:
For background info, I'm 35 yo. French guy, and I have been playing competition chess since I'm 20. I achieved a ~2050 FIDE rating by 25 and it didn't improve much since then, but I've accumulated some chess experience by teaching, interacting with stronger players, and observing improving players in the various chess clubs I have been playing for. I have played around 400 OTB long games during these 15 years, most of them before I was 30.
So here is what I really think are the most important things to improve :
1) Competitive drive and focus : chess is a fight, and what makes the difference when you're playing someone around your strength is how much you want to win, and how focused you are during the games. It gives you steadiness in your play and results. That's the most important part. I've lost many half points simply because I was careless, or unable to focus properly...
2) Playing regimen : the more you play focused chess (not blitz, blitz is for fun), the stronger you become. My experience is 25-50 long games a year OTB is what you really need to improve steadily over a long period of time. 10 games/year is the minimum to stay fit. I reached my rating by playing +25 games/year during 5 years.
3) Get a good mentor : it's more important than reading 100 books, because you learn faster this way. The mentor will show you what is important FOR YOU, and as the information is relevant, the learning process will be way faster. The mentor doesn't need to be a professional chess teacher. Any nice player, significantly stronger than you (I suggest rating > 2000 is a minimum) will help you a lot. What you must do with this mentor is : analyze your games, look at his games (for inspiration) and analyze any specific chess position that is of interest for you . When I started chess, I found kind mentors in my chess club : people rated +2200 elo who eagerly analyzed my games with me, even if I was a mere 1500 player - it helped tremendously ! And attending their analysis sessions too ! You won't find chess explained this way in books.
4) Analyze your games : first with your opponent, then by yourself, then with mentor, then, and only then, with Fritz & co. : the analysis doesn't need to be complete, it needs to be instructive...better learn 1 good lesson then seeing that 23.a4 improves your position by 0,23 pts. You can always come back to this game 1 year later : you'll be surprised... I analyzed thoroughly 4 of my games when I was ~22 and, believe it or not, my rating jumped 200 pts. the next year :-)
5) If you want to practice, play semi-rapid chess (not blitz) and check the opening theory right after each game - this way you learn theory bit by bit, without suffering too much (and may find that your moves, even if not "theory", may be quite good after all...), and variations will stick in your mind...
6) Don't fear your opponent's rating when playing : if it's higher than yours, then, by all means, play the pieces, be as OBJECTIVE as possible, and stir up trouble on the chessboard...As soon as I stopped thinking +2000 players were chess gods, I started attacking them on the board and beating them :-) and crossed the 2000-barrier
7) Read what you find interesting in chess books (some are very good !), but my rule of thumb is if your book studying time exceeds one half of your combined playing/analyzing time (ie. 1/3 of your total chess time), you're going to become a chess scholar, but may not improve your rating as much as you want...Most strong players don't read many chess books. They pick what they need as a given time and set positions on the board to analyze.
8) Have you chess brain work at least 10 minutes every day : meaning solving a competition, or thinking about a specific position. It doesn't need to be 1 hr. but a little chess every day prevents rust...
All of this has worked for me, and the advice is in descending order of importance : if you can't do #1, I think #4, #7 or #8 are pretty useless...
Good luck !