Cold streak in tactics training

Fluctuation is inevitable in tactics training, but it's still quite frustrating and I would welcome any suggestions for breaking a cold streak. I recently reached a new personal best rating (1668) after three months on the site, in which my improvement was quite steady. (I can only do 5 tactics per day.) I've watched it plummet over 120 points in the last ten days and, despite some days of stabilization, the trend has been downward. I do puzzles on other sites and in books regularly, so I don't think I'm getting worse at chess, and I recognize that improvement isn't always linear -- but what's the best way to bounce back, both psychologically and in terms of the rating? Thanks!

Tactics are frustrating. For me, to gain the queen is not less than a mate. Or a mate in three is not less than a mate in two. Sometimes I found an excellent move (after the official anaylysis), but it was not the one the enginge demands. The answer is always minus...minus...minus.

Chess.com tactics have a certain tolerance for the second or third move. But, in my opinion, it should have a certain tolerance for the first move. Chess is seldom only one golden move. The silver move very often is enough. ----the opinion of a frustrated hobby player. Proud of the silver move!

Just take your time and calculate once more after you think you found the solution. Looking at your tactics training history, you seem to spend like 20-30 seconds on average on each puzzle. That's just not enough if you aren't already a tactics master. I usually take like 2-3 minutes on each problem.

Solve for accuracy, not for speed. You'll get less points on right answers but you learn better and you should slowly improve the rating as well having a higher solving accuracy rate.

Nexim, you are absolutely right! But why is it neccessary for chess.com to solve the puzzle within a certain time? If you are too slow, you will be punished. It's not so on lichess. I cannot understand this. Even Grandmasters need a certain time. Sometimes on easy puzzles. Sometimes I think chess.com wants me to become an engine. But I am not. I am a human being with certain preferences. Often the official solutions leeds into a position I don't like. I don't master. I would prefer an easy solution, maybe a bit less professionell. The answer is always: minus...minus...minus. To be honest: I hate tactics. It's not the real chess life. It cant't be, because here in the puzzles trainer it is clear, that there is a solution. But during the real match, it's not clear.

I solved 3 in a row today and got 1 point each. Made 1 error and lost 14 points. Chess is hard.

yeah, my rating fluctuates on an average of 200 points. I'm old so I have that going against me. I had surgery a few months ago and had plans of spending my convalescence playing chess. Apparently the anesthesia damaged my brain and my tactics rating plummeted. I was missing really obvious moves. it's slowly climbing back up. The tactics confuses me sometimes though, it will have moves that nobody would really do in the real world. Not all the time of course. But there seems to be a pattern.

All the advice on here is really good.  To get better, just take your time finding the correct answer.  Realize, as someone pointed out, that the scoring system is poor.  In otb chess, winning a queen is as good as a mate.  Also, many of the tactics problems I saw rated 1600 or less are problems that no 1600 I ever played would have seen.  This is not a fault of this site, but a problem I have found on other sites ad well.

I'm a 1600. Try me...

I have never seen a 1600 rated player in otb chess play a series exchange tactic.  But I have seen Experts and 1900's play that tactic effortlessly, even though it requires a lot of organized analysis.  

Fair enough. I shouldn't pretend like I'm smarter then I really am.

Because chess.com's rated tactics are styled as blitz tactics, meaning that you are demonstrating your ability to find tactics quickly in a given position (Chess.com also has the 'learn' feature, with themed tactics, which is more like the lichess, but it isn't as fun because you don't win/lose rating points). The problem with this is that you're expected to already have the ability to recall memorized patterns or at least get a strong feeling from the position enough to calculate one main line, or else the time kills you. I think that you'd need to spend a lot of time solving tactics on a site that doesn't use a timer first to hone your skill-- but which has a rating to keep you motivated (chesstempo, ideachess, lichess, chessbase, etc --or all of the above  ) and if you're really dedicated, to keep a copy of each tactic that you fail, and revisit it preferably using spaced repetition (chessable.com). It's not about solving thousands of puzzles and that's all - there are many tacticians who've solved 10s of thousands of positions but still have a very low rating - it's about:

1. Recognizing which patterns you don't know and drilling them into your brain until seeing their shape in a tactic becomes automatic.
2. Being able to calculate 3 or 4 moves ahead in a few candidate moves BEFORE going with your first gut feeling. I know it's hard. Sometimes i sit on my hands.

There are some tactics that are flat out ridiculous too. Like this one for example:

After playing Kf3, basically everything wins for black. I saw instantly that after Rxb7, Rd1+, Kh2 and Rh1# is a familiar "hook" pattern with knight and rook, that I've practiced with various tactic books. So seeing a forced mate sequence I happily played that move, just to realize I failed the puzzle. And only then I realized that, wow - of course. Just playing Rg2# right away was a mate in one in that position. As if I'm going to look around if there are other ways to mate if I see a forced mate in two pattern...

In my opinion they should remove all puzzles that have multiple different ways to mate in 2-3 moves. I've failed a lot of tactics playing a forced variation leading to mate, just to fail because there was a one move faster mate on the board..

Also not long ago I had a bugged puzzle, where I had two rooks pointing at a piece for a mate in one after a long sequence, but the other rook was pinned and obviously unable to move. I played the mate in one with the rook that was able to move, and the puzzle said I failed and showed the right answer to be taking with the rook that was pinned to my king by a bishop... When it showed the answer, the rook even changed colours and a black rook turned to a white rook capturing it's brother...

Years ago there was a study done to see does a correlation exist between speed and solving chess tactics.  One author even put out a book on it..  I was doubtful of its validity.  I tested masters, experts, and 1500 rated players I knew.  The book really found their current  rating.  When computers came along, the programmers tried to predict ratings as well but fall short, giving too many points, too little, or discounting reasonable answers in favor of long-winded solutions that lead to a forced mate.  So if you find a forcing move that works but isnt the mate they wanted, pat yourself on the back.  In otb chess, only a win is important.  The rest is all style.

Who can help me? Once I had 2167 tactics rating, now I have zero!!! Incredible!

But that's why the scoring system is so fair. I've lost 15 points for a wrong answer, and often only get 1 point right for a correct answer. And they often seem like the same difficulty tactics. But that's exactly how a chess game works. You can do everything right, and only gain a very small advantage. But do one single thing wrong and you lose big, a queen or even the whole game.