99% Tactics

Yes, I have been using emrald for like 4 years. 3 user names: cotoi, cotoi1, costelus. I don't know if that's the best TT available, I'd say that emrald and chess.com are the best. They complement each other: emrald emphasizes pattern recognition, chess.com puts a higher weight on calculation. As for chesstempo, I don't like their ugly interface and their selection of problems (in particular, I hate the message "your move was good, but not the computer's first choice").

I hardly have any time to study chess except for solving some tactics. Very rarely I go into an opening variation or I get to study an endgame.

As for the question if you should train other chess areas more: you know it better! For my case, I lose the overwhelming majority of my games due to tactical errors. I often play like a coward against a strong opponent and get into hopeless positions, I don't know when it's time to strike back and I don't have courage to lose a pawn let's say to get my pieces active and create threats. So I need to work more on tactics.

I used to but I didn't like it as well -- maybe they've updated since then though (it's been over a year).

People tend to look at things as "either/or," as opposites which are mutually exclusive. This is rarely the case, and particularly with the notion of positional play. Positional play is all about tactics, it's just not about the swashbuckling, all out, "I don't care if I lose because at least I'll go down in a blaze of (self-delusional) glory" type tactics. It may be more about "win a pawn here, win an exchange there, trade down into an endgame and then crush him" tactics.

I have heard others make the point that tactical play is about utilizing the initiative. Of the three types of advantage -- initiative, material, and positional -- initiative is often the most difficult to preserve. Positional play tries to use all three advantages: the initiative to improve position; position to limit the opponent's tactics while setting up for the widest variety of potential tactical shots of one's own, in an attempt to gain material if not immediate victory; and material to further increase initiative, perhaps gain even more material, and improve position thereby advancing the game toward its conclusion.

No, good positional play won't make you bullet-proof in every instance. Reliance on pure tactical ability with no regard for position is no better. Both tactical ability and positional sense are important.

Tactics make up the foundation of positional play, because without them no strategy could be executed to its fullest, and simultaneously you'd probably start losing pieces to people who are actually seeing the tactics.

Someone with basic (but present!) positional knowledge and excellent tactics can expect to win against someone with advanced positional knowledge but mediocre tactics the vast majority of the time. That's just the truth, and my experience is consistent with this. Instead of complaining about this (as someone who stylistically prefers quiet positions where strategy predominates [but again, likely tactics will happen even in this case eventually, just with perhaps lots of maneuvering first to "prepare for battle"]), I just try to compensate for this by getting as good at tactics as I can be and trying to find and make use of any dynamic opportunities I may get. They may not necessarily be good, but it's vital to see them because they easily can be good, even game changing.

"Positional play is all about tactics"

Actually, this very unpopular interpretation caught me by surprise. If you interpret things this way matters can get quite confusing. When talking about "positional" in the purest sense, this completely isolates tactics. Unless you mean positional in the way that the positional masters play, still always keeping an eye on tactics. But that's tactics and positional play intertwined, and that's the point, you combine them.

Then indeed there's the guys who do anything for tactics, etc, but the point is tactics are the foundation of any kind of player, and more tactical ability is always useful, a little more consistently than strategic knowledge, though it's still important.

The RD is an indicator of how accurate is your rating. There is a disadvantage in this respect on emrald: a very low ID means that, no matter how fast you solve problems, your rating will remain almost unchanged. That's why I have 3 accounts there: I don't like to have an ID below 20.

Rating improvements ussually take a long time to happen. In your case, it might be just a normal rating fluctuation: in some weeks you are careful, in some weeks you are tired.

chess problems are indeed helpful to improve your game because they deal with

1.tactics

2.analysing specific positions

3.pattern recognition

by the time you play and solved hundreds of chess problems, you will already

have a sharp eye for finding the winning moves or how to gain material..

I used to hate TT -- in part because I sucked at it, though also because I didn't truly appreciate the merit of a program putting pressure on you by giving you multiple winning lines, and quite limited time, to find moves that may or may not be very sophisticated. There are some very simple ones on there, yet I would, sometimes, miss them. I would complain that it's time that's a problem -- that's valid. However, the real point of improvement is not only to calculate 6 move combinations, but also to develop consistency -- to smell a knight fork 2 moves before it's there; to never miss a queen fork ever, under any circumstance (so that you don't lose games in time pressure!); you don't only want to know tactics, you want to feel them, sense them, know them very, very, well.

If you can do thousands of tactics trainer puzzles within the time allotted, and understand them, that means you will start to be able to very efficiently find tactical themes in a position when they are there. With a strong foundation of tactics, the building blocks of combinations, long combinations will come much more easily, because all they are are just long strings of tactical moves and ideas put together.

Chess is also 99.9% not giving things away. Ever analyze a complicated game of yours with an engine? Notice how many times the advantages shift? Any amateur contest is full of these shifts, inevitably due to big mistakes, and it doesn't matter when they happened -- or how well one was playing previously -- one bad move can change everything. Clearly, it is essential to not play these bad, generous moves!

one thing about the tactics trainer thing here is that it is massively overrated, as if you compare my ratings on chess.emrald.net  (same handle) and here - 2550ish(!) versus 1750ish.

chesstempo is probably best for tactics.

Well, problem ratings on TT, just like chess ratings, are based on those who solve and fail them; clearly, problems are overrated because their failers, whether or not they had the skill to, didn't have the consistency to solve them, under the conditions given by TT. Simple puzzles often get harsh time situations, but dealing with it is good training. You certainly don't want to sap up too much of your time on basic forks and pins during your actual play, as those are far from the scariest of problems you could face on the board.

Try also this free site for tactics training http://www.ideachess.com
It's very nice!