Tactic Trainer Timer needs to be removed

I think the point is that the OP gets a certain level of problems correct (but perhaps takes a little too long), and finds it frustrating that he cannot "unlock" the harder problems, so to speak.

I guess the organic answer is that after enough practice, he will become more proficient with the current level, so that he will thereby discover the next layer of the onion.

Getting the solution correct and then losing points is ridiculous.   That is why I use the TT on another site... correct solution... I gain points.  Once I was interrupted and when I came back 4 mins later and solved the problem I still gained...way to go !  It should not be about speed.

It should be about learning and improving, not rating points.

Getting it correct but taking too much time helps increase the average time for other users... so you're doing a valuable service.

I think time is an important factor for this, but I also think solving vs failing is important.

wouldn't it make sense to have a cap for when you successfully solve it?
eg. if average time is 18seconds, and I take a minute and a half to solve it.. why do I get the same score as someone who got it wrong after a minute and a half?

if perfect is +10, and fail is -10.. taking to long should be maybe +2.5, or at the very least, -0?

my current experience in tactics trainer, is that I frequently get "move 10 pulled from a game" type of puzzles, where the solution seems to require either deep analysis, expereince studying opening books, or just tons of experience.  I often get these wrong, and they prevent me from gaining more points so that I can see harder or more varied problems.  as a free player, I can only see 3 problems per day, so it's pretty dissapointing when 2 of them are ones that make me think "sigh, I can either guess and have some chance at points, or I can think about it and have no chance at points"

"wouldn't it make sense to have a cap for when you successfully solve it?
eg. if average time is 18seconds, and I take a minute and a half to solve it.. why do I get the same score as someone who got it wrong after a minute and a half?"

There is a cap, 20%.

Ohhh yes.

In fact if you reach 1900+ problems they start to become so much more complicated that they "concede you" a lot more time to think.

Most of the time when I fail a problem is simply because I don't get it. Time pressure becomes less relevant in higher tactics, in my experience.

I am by no means a 'fast' player.  I was stuck in the 2100 TT rating for quite some time.  Turning off my timer and doing more problems really helped me jump to the next level.  I think understanding the problem and making correct concrete calculations is the most important thing you could be doing when playing tactics trainer, not gaining rating points.

Similar to that, one thing which really helped me was to focus not on my rating but instead on my plus score. I was frustrated that I had more incorrect than correct, and worked on this, so now I am +384, with a 52.5% success rate.

In other news, I just discovered that if you are inactive for some time, you don't show in the list of players.

I love the timer, it helps me see things that I probably wouldn't see normally, but also certainly wouldn't see quickly in a game. Even in long games I'm tempted to play one of the first moves I see, just because I don't want to take 5 minutes a move.

By being able to see the whole board quickly, it not only helps to visualize moves quickly as well, but also works as a refresher for me when I'm in a complex situation and, say, can't remember which pieces are covering what squares. I remember before I played a lot of tactics trainer, I would often move a piece away from covering a vital square and end up losing because of it. Now I can quickly brush over the board to re-analyze the whole situation I'm in, and maybe I'll see a checkmate soon or an exchange where I'll come out ahead.

I certainly think the timer is a great aid, and I don't want to sit over a puzzle for 10 minutes to make sure I get the right combination of moves just for a decent rating. I will admit there are a few times where I feel that I finished a puzzle in a reasonable time and ended up losing points though, and that is annoying. I think it just needs some work if anything, but it definitely shouldn't be removed.

EDIT: possibly, as somebody else suggested, if you get the solution right and take a ton of time, just don't gain or lose any points, +0. Also I have a lot of bad/off days as well where I either can't get a single puzzle right, or don't see a single move for a good minute or so, and I don't want to be punished on those days just because I want to get my 3 TT puzzles in and not waste them =/

But the actual game would represent the test. The hours or days in advance you have to study for it represents the Tactic trainer. You learn nothing from TT the way it is.

Unless your school just throws tests at you to see if you already know it?

Yes, I totally agree with you nikiburny! :D

I think the OP gets a lot of them correct, but also said that his rating hasn't gone up any in years even with all those patterns he's seen.

Also for a couple of the people here who say the timer helps them, I've seen their ratings go up from 1100 to 1300 (highest) and back to 1200, (which is where you start at) over a period of 5 or 10 months. If the timer is so useful, then why isn't it helping them any?

I would like to see one of the 2900 rated problems. Post a link to one please. It's the only way I'll ever see it. Especially one that has a time of 15 seconds since TT is all about recognizing patterns and not thinking as one person put it. (I'd hate to think of chess as just that. It's why I stopped doing sudoku puzzles. I had essentially solved it.)

As an NM I'd think you would be well aware of the adage that if you see a good move, look for something better. If I see a hanging queen, it might be the correct move to take it, but I'll look for something that leads to mate. Since I have the distracting yellow and red timer bar turned off, I have no way of knowing if it's a simple problem or not. So if I spend time looking for a better move I will get a lower score if there wasn't one, because most people will go for the obvious. When they are thrown a harder problem I notice the pass rate is low.

I seriously doubt that anyone has improved their chess from TT alone the way it is. And if you can't improve from it without books or other forms of help, then it's only useful as much as a test in school is. A measuring stick for how good you already are.

I once failed a problem because I found a perfect mate in 4 in seconds. Problem was that the correct answer was a mate in 3!? Having found a viable solution, I deemed it to be correct. If I choose to look for anything better, I risk getting a lower score. In an actual tournament game of chess, you might have 15 minutes to think of a move in midgame.

Chess isn't a first person shooter high speed action game. So why should I want to play it like that?

(Of course it also doesn't help that about 10% of the time, after I move a piece, the computer does nothing and the clock keeps ticking away. I have to move the opponents piece to then fail the problem, keeping my rating, and thus the types and variety of problems I'm getting, the same.)

Chessplayer11, you want to see one of the high rated problems?  No need to get your rating to 2900.  Just go here:

http://www.chess.com/tactics/problems?sortby=rating_hard

Here's a good one (at least I liked it)

http://www.chess.com/tactics/server?id=79598

Thanks. I think the problem is not so much the timer by itself, but that it's still buggy for me and a few others. About two years ago I could get a problem in 1 second. Then they changed the code up so that when I make a quick move, the timer keeps going and the computer does nothing and I get a 0% score. Getting a 100% has become nearly impossible for me, no matter how low the rating for the problem is.

There are forums on this where other people have mentioned this bug, but the staff just doesn't care. I think it's been around that time that my rating have become stagnant. I more worried by that glitch, then pressured by the timer. So the timer is the problem, but for a different reason.

However, I still don't think you necessarily get better by being timed in a learning environment.

The tactics trainer timer is required to calculate the rating, so it should stay. The bug is not so fun though so the admins should look into that.

Interesting problem. Ironic because it's the most obvious move to make in a real game, but only from a positional point of view, not tactical (in a tactic trainer). Which to me is far easier to see than trying to calculate a five move mate. I would have missed this one from what I've come to expect from TT problems.

They really need to change the name of TT to "Really, whatever is the best move trainer" :) Or maybe just call it strategic trainer, as this is by one definition, not "made or carried out with only a limited or immediate end in view."

See, I like this too because it adds variety by not being limited to just  immediate material gain or checkmate. I think the pass rate for that might be lower because people have come to expect a convoluted series of moves that would have mated the enemy before he can mate or force a draw against you. Instead of going for what's natural.

I'm assuming that positional moves are uncommon even at higher rating levels since it has a low pass rate, and I don't really believe you would need to be a world champion to have gotten this.

It's only required since it's set up that way. It's not actually required to calculate a rating.

And the rating is only used to determine what type of problems you get. In a real world game, you're not going to always be given a 1500 rated puzzle to solve. It's the choice of the site to deliver problems of similar ratings to your own. While not a bad idea, also not very interesting when you're stuck at that level because of a buggy site. (Also my cheap mouse will sometimes drop a piece even though I didn't let go of the button at the wrong spot. Not very helpful mouse.)

There are a lot of ways to lose points, and fewer ways to gain them. If I get 15 problems correct and increased my rating by +25. Missing two problems completely can set me back all the way.

One thing they really need to do is add weights to each move. I've seen problems rated at 1100 that have a pass rate of 3% because the first 5 moves are very easy and the last is at the grand master level. Due to the simplicity of the formula used here, a good problem like that stays at an odd rating level.

And it's because you're rewarded for how quickly you can get most of the moves correct, even though you shouldn't be getting a positive score because that last move you made wrong would cost you the game in a lot of situations.

A good example of why the timer aspect of TT is flawed is this one problem I remember from a long time ago (And wish I had a link to to illustrate). It ended with a rook move that would lead to a checkmate, and I got it correct, but for the wrong reason and thus causing the rating and average time to be flawed for that problem. About half the people moved the other rook instead thinking that was also good. Had they been able to look ahead a few moves they would have seen that it would have given the opponent time to checkmate you. Had they taken the time to see this, they would have probably lost points.

That problem belongs in a higher rating bracket, but since time and luck are bigger factors in TT, it stays where it is. Had I just moved on to the next problem thinking I got it correct without seeing the comments I would have never had learned that I really just got lucky.

There are lots of problems like that, and if you remove the time component, you remove luck as a deciding factor.

One more problem is that a lot of players will make an obvious move first without looking ahead to see if it leads to anything good causing the average time to be very low. These problems really irritate me, since by the time I find that it's the only correct move, the puzzle ends right there since it happens to be a one move problem. And my score is much lower than people that make knee-jerk moves and (again) get lucky.

These problems are sometimes fairly complex, but due to how TT is so reliant on time, they end up placed at low rating levels.

I think if you take away the timer, you will see problems falling into their true rating level where they belong.

I strongly disbelieve that large numbers of people will start cheating to mess up every problem since they will learn nothing from it.

Outside of the fact that some people insist that the timer is crucial, I see no evidence that it helps at all. Even the supporters of the timer keep saying that it's about pattern recognition. Does this imply that you can't learn patterns if you get more time to study the same number of problem?

Of course thanks to DavidMertz1 helpful link I can now get exactly what paying members get. Unlimited problems at any rating range with no time limits. :)