Tactics trainer question

Hi. I suppose this question was discussed many times but anyway i have my curiosity. Look at my TT achievements. Rating is low although percentage is decent. I'm curious how it's possible to calculate all variations so fast. I'm solving puzzles but my rating goes down because of time. Looks like it's better not to calculate but to move as quickly as possible because it doesn't matter if you solve it correctly. On other sites even if you spend much time but solve a puzzle then you are praised with scores anyway. What do you think is the best strategy to increase TT rating? Thanks in advance.

Don't worry about your rating or the timer. Just work on solving the problem, don't make any moves until you see the solution. Your rating will likley take a hit but eventually, you'll get a little quicker as the patterns become more familiar.

They should be more or less familiar already. What have you done to get your highest rating? It's not easy to solve quickly even if i already know most patterns. And sometimes i'm late even if have come across this puzzle before. I need 3,4 times to memorize.

The site has changed this a few times since the last time I reset my TT rating. So, my highest rating here isn't close to accurate based on the current state of TT.

I don't do tactics on here as much as I used to and have a habit of not following my own advice as much as I should. However, I still see tactical patterns I don't recognize in the rating ranges I get and that keeps my ratings where they are at.

Based on your success rate, my guess is that you can calculate the tactics fine but you don't see them very quickly. The idea is that when you get really familiar with the patterns you will see them more quickly and solve them faster, getting higher ratings.

The best thing is to figure out the variations as you do, no point of doing it just for guesses. And you will recognize the patterns, that doesn't mean you will recognize them after just only one or 4 puzzles, but after some time you'll recognize most of the patterns in a rating area and then you can leave that area and work on the next one.

you get the score you deserve.

why care about the rating you get for solving a tactic if it doesn't make you better.

the only way you get better is to see the whole variation before moving.

let's say you play a 2 hours game otb. if you do not see the tactic within 2 minutes you will not see it in a game.

therefore time means everything.

if you have 5 candidates moves for a position imagine using 10 minutes for every candidate move to see if you have a tactic. your time will run out quickly if you do that for every move.

as you get better you will see the variation before moving faster.

therefore it is key that tactics are timed so you can monitor how well you react under pressure.

"Don't move before you see it".

you can win a lot of 10 or 25 minutes game if you can spot a tactic within 5 seconds.

TT rating means nothing. don't focus on the rating. just pay attention to the tactics that appear as if it was position against a player and you wouldn't just aimlessly make a move just to see it is correct. calculate as deep as you can. if you spot mate don't calucate further.

I don't know. It doesn't make sense everytime. Many combinations are really difficult to calculate till the end for short time. You just feel that it's correct but sometimes you could be wrong of course. Kasparov has said that intuition is very important. It might be a TT problem. I could see a correct move quite early but i'm not sure and spend time. I'm worried about rating because i don't want to stuck on the same level.

Omega Doom you will not get stuck at the same level as you get better. it just happens without you looking at the rating.

I don't look at my TT rating at all. I just know I have to solve the tactic that they put in front of me. nothing else matters.

what do you prefer to get 100+ rating fast and lose it because all tactics get complicated

or get 100 + slow and suddenly being able to solve higher rated tactics.

Here is a good strategy if you want to improve your tactics trainer rating:

To improve your tactics trainer rating, one plan is to not bother seeing all the lines to the end. No way. This is the first thing - don't calculate till the end, just guess. If you are stuck then just take a random punt at the first check or sacrifice which looks nice and maybe you will get some 'partially correct' even if you get it wrong and pick up extra points this way. ('Partially correct is actually 'wrong' in real life terms, but who cares?)

Also it is important to ignore your own position/defense for example - do not waste time considering if your move is safe, this wastes valuable seconds and the timer is ticking down all of the time remember, just play and hope for the best!

These points will improve your tactics trainer rating for sure. It might help wreck your chess and imbed bad habits, but you will improve a bit tactically, although only in attack, but your rating will go up.

You must remember though, when you do solve a complex and nuanced mate in 7 position in 2 seconds (and this will happen when you take the punt approach) remember to give something back to the chess.com community by writing helpful comments such as 'first' 'wayyy too easyy' or '+10 points' 'solved it in 2 seconds' and so on, it is always nice to give something back.

Of course there are alternatives to the approaches above. You can turn off the rating, ignore it or do tactics elsewhere. And when you do these things do the opposite of the advice above, that is: first consider opponent's threats, then look for candidates, then push these to the end, maybe also think of evaluating the resulting position etc, etc. This might not improve your rating, immediately, though but might just improve your chess, and rating, in the long run.

The point of having tactics with a timer is to get used to calculating under pressure, and being able to recognize pattrns quckley.  DO NOT get caught up in the score, or tactics rating. 

If you take your time, and learn the pattern, and how to solve the tactic with a minus score, and lower tactic rating, that is more beneficial than going through them quickely, getting lucky on some, learning nothing, but getting a higher tactic rating.

Ok. It still looks for me controversial. For instance there are one move puzzles which should be solved for about 20 secons or so, patterns are very simple but continuations can be several moves long. If i solve them till the end i will lose on time definitely because my mind can't calculate that fast. Unfortunately i'm not a machine. Do you suggest doing it till the end anyway? If time is given us in order to help under pressure then how can we not to hurry?

The problem is that the time allocated for each move is related to the overall time users take to solve each puzzle. So, you get the '2 second movers' bringing down the time to unrealistic and dangerous levels - 'dangerous' because these people are not making safety checks or considering the position fully and this is bringing down the average time. Even careful users are encouraged to move quickly, not accurately because of the ineffectual rating system here with TT. I have already posted the solution to this issue.

Thats what im saying.  Dont get caught up in the rating or the time.  Take as much time as you need ot solve, and fully understand the tactic.  Its better to know, and recognize the pattern, than to thave a higher rating, and not understand it. 

Again it depends. I've seen puzzles more that 20 moves long till the final kill but it needs to take just couple to solve it. These couple moves can be calculated fast but i have really big doubts about the rest ones. The same happens in a real game. For example if you see that opponent's king is heading towards the center then you don't need to calculate everything to understand that it's correct. You just need to rely on your intuition but anyway i hate such puzzles.

I've found that the situation greatly improves once you get to problems beyond 1600-1700. Partly because problems where the 'guessing approach' fails, naturally will get a rating boost due to the number of people employing it, I'd guess.

I've been fooling around with the TT here a lot during my days on this site, and I feel that most problems are reasonably rated and allow ample time for solving (but, yes, there are exceptions...). The biggest problem is that partial solutions aren't punished enough, I think.

How many puzzles do i need to solve in order to do them quicker? I've solved more than 5000 and i'm stuck at 1600 - 1700 rating. Most puzzles and patters are simple. I have come across the same ideas tons of times but i still can't spot them so fast. 50000, 500000 puzzles more?