Tactics Trainer Issues

The Tactics trainer keeps giving me the same puzzle twice in a row.  When it happens on my mobile device, I get the error message that I have exceeded my daily limit of tactics puzzles.

Hello Matt. I would recommend that you try uninstalling and then reinstalling the Chess.com application to your mobile device as this may help to solve this problem.

Thanks, Patzer.  This is happening on mobile and web. There is just a slightly different behavior on the mobile, that of telling me that my daily limit has been exceeded.

When I look in my recent problems list I am not seeing the duplicates.  So, I'm not sure how the tactics trainer selects problems for me, but I would suspect that the issue has something to do with results not properly being logged to my history.

Is the mobile limit you're experiencing on an Android or iPhone?

Hey, I'm not sure if I should just post this here, but sometimes in tactics trainer, if I drag a piece, anticipating the computer's next move, before the computer actually makes his move, TT freezes for me. I've waited for around two minutes hoping TT would unfreeze, but to no avail. A simple refreshing of the screen works, but TT just gives me a different problem. It isn't too big of a problem, but sometimes a little annoying if you had a particularly interesting problem.

Hi, kohai the problem is on Android. Just installed to a brand new device, HTC EVO. Problem won't go away.

Ok, with the droid, when playing TT are you on mobile or wifi?

I'm testing a release at the moment so would like to be able to replicate this and get it fixed before this update goes live if possible. I've tried several times so far to replicate this and its just not happening for me on mobile, which is why I asked if you were on wifi when this 'limit' error shows.

Hi guys,

I find it totally unfair when solving a problem is rewarded with minus points. For not starting a problem you get zero points so how come when solving a problem can lead to -10. The fact time is an issue sounds more like speed training than tactics training to me but okay staring at a problem for an hour shouldnt be rewarded the same way as solving the problem in 10 sec but deducting points for solving a problem within 40 sec shouldnt lead to 10 minus. I also had a problem which I failed rewarded with 2 and it wasnt particularly quick either.

So what's the thought process behind it????

Kind regards and much appreciation.

I always wondered why the score doesn't just go down to 0 points if you take too long for a given puzzle.

That is the only correct way to do it.

Failing a problem quick is rewarded over solving a problem slower than average should be out of the question.

Hi Kohai,

It happens on both wifi and mobile.

I'm having the same problem as SoulsofChess with the Tactics Trainer.  Sometimes I make a move, it is recorded on the board, but the system acts like it's still my move, but won't let me move anywhere.  I can hit refresh to continue tactics trainer, but sometimes it takes me to a new problem altogether.  This is definitely very frustrating, as the tactics training was one of the main reasons I signed up to be a member.  I'm on Windows Vista and Firefox.

I experience this phenomenon too, but very rarely (at most once every couple of days).

The way I figure:

1. The time limit is based on the average time people actually take to solve it. If you are getting it right but losing points on time, then you are obviously taking significantly longer than average to get it right. Therefore it is proper that your rating, relative to everyone else, should be lower as a result.

2. There are two sides to every transaction. If you take too long at a puzzle, then not only does your rating go down, but the rating of the puzzle goes up. This means that you are less likely to that that puzzle, and more likely to get ones that you can solve quickly. In theory, once you get enough practice with these, you'll learn to be able to solve the harder ones more easily. In addition, it means you're not continually getting problems which would take you ages and ages to solve, which would be really annoying.

3. The point of Tactics Trainer is for training. It's not like a space invaders game where the goal is to get as many points as possible. The only purpose of the rating is to select problems for you that are an appropriate level of difficulty and, more indirectly, a way for you to measure your own progress. I just don't get why some people think it's "unfair" when TT "takes points away from you", or that it's a "reward" when it "gives you points". Pun intended, but I think that kind of misses the point.

Yes, you do see it completely different than I do.

1. That time limit is based on the average time people actually take to solve it is clear. Therefore getting less points than somebody who solves a problem quicker is totally understandable but it is not about getting less points for solving the puzzle, it is about getting the same amount of points deducted as failing the puzzle completely. Average problem solving time is only relevent when everybody has more or less the same rating. Lower rated are more likely to fail so not many are not participating in time versus solving statistics and the much higher rated players you can't be compared with really. The way you see it would mean all blitz players are better players than standard chess players. Points you already get deducted for failing a puzzle so there is no need for taking away points when you solve a puzzle and apart from that it wasn't ages the problem took but 40 sec. I am already triple checking everything because sometimes taking a pawn, a knight or a queen is not the solution and sometimes it is.

2. This sound completely rediculous to me, a puzzle is a puzzle, the puzzle doesnt change so why should it be different rated today than tomorrow. Thinking about a puzzle aint the problem, I think about a problem until I got it which should be the case with chess, guessing and assuming doesnt lead to nothing and apart from that when solving a puzzle with a bit more time needed gets rewarded the same as failing in an even longer period is insane. A bit like the 100 meter dash, the fourth athlete gets disqualified and has to pay a fine on top of it cuz others were faster. I am training for otb not for bullit.

3. Exactly, Tactics Trainer is for training and not for rushing through it. You play against the clock and sometimes guess and on top of that get extra points for solving it very quick but you didnt solve it, you guessed. The goal is exactly to get as many points as possible this way you notice you get better and get more difficult puzzles, if it wasnt about the points then there wouldnt be any. Let me explain to you why it is unfair to get points taken away when you solve the puzzle, it is unfair because you solved the puzzle. It is fair to get less points or none at all when you are slower or very slow but getting less points for solving the puzzle then to fail the puzzle quick is utter nonsense. "Yes sir, I have to give you a ticket for driving through the green light later then the average, there you go!"  Any IQ based test not answering a question is better than giving the wrong answer, giving the right answer is better than not answering. How can giving the right answer be worse than not giving the answer!

Hoping I made a point.

I don't think so. Tactics are about both calculation and pattern recognition. If solve a puzzle but take too long to do it, then I think you're probably doing well on calculation but not so well on pattern recognition. Remember, in Tactics Trainer you already know there's a tactic out there somewhere so you'll be looking for it, but in a real game you won't know this, and unless you can recognize the pattern in a relatively short amount of time, then you're probably not going to spot the tactic at all. So I think being able to solve things quickly has real practical utility, because if you're playing, say, a 15 minute game and you're spending 60 seconds each move just looking for tactics, then you're going to be in trouble. Maybe you can think about if that if you didn't solve a puzzle in twice the average time (I think this is the level where you start to lose points) then you probably wouldn't have spotted it in a real game, so failing to spot the tactic entirely is basically equivalent to failing it, so losing points seems right to me.

Well, because this is how TT picks them for you. Every puzzle is rated, and TT picks a puzzle for you each time that is close to your rating. The puzzle rating gets set based on how people actually do with it, i.e. every time someone fails a puzzle the rating goes up, every time someone solves (within the average time) the rating goes down. Once 20,000 people have tried it, you'd expect the rating to be relatively stable, but there still needs to be a mechanism for the puzzle rating to change.

In the same way, everything else being equal, a puzzle that takes longer to solve is more difficult than one that can be solved quicker, so it's appropriate for it to have a higher rating, another reason why losing points on time makes sense. If your rating only went down when someone actually fails it, there'd be no mechanism for the rating of the puzzle to increase to reflect how long it's taking people.

Well, the point is to get better. Your rating may reflect your ability, but the point is still to get better, not to improve your rating. It's not costing you any money to lose points, so I don't see how it can ever be "unfair". And since "getting better" includes being able to recognize patterns and solve them quickly as opposed to slowly, I think it's right that your rating reflects this, and goes down if you take too long.

You certainly could have alternative system where time doesn't matter, or where at most you get zero points for taking too long. I just think the way TT actually does it is better, because I think taking too long is equivalent to not being able to spot it in a real game, and not being able to spot it in a real game is equivalent to failing the puzzle if the goal is to improve. If you're not recognizing the patterns quickly enough, then the remedy is to spend more time going through easier patterns to reinforce them, and this is exactly what losing ratings points on time helps to achieve.

One thing to be said  for the system is that failing a puzzle always deducts more points than solving it in insufficient time.

I noticed during my tries that I can let my rating go up through guessing and moving quickly. Pass rate goes down, but rating goes up. That's not what I want, though. I try to follow Dan Heisman's advice to always look for every possible check, capture and threat both me and the opponent has. I try to make a material count and look for the most basic features of the position before I even start calculating. If I found a good move, winning material, I try to look for a better one, possibly mate. Sometimes this is the right thing to do, but sometimes not: I always lose points on the "hanging piece" puzzles which many think are just freebies. 

I also found - to my initial surprise - that I'm more comfortable with higher rated puzzles because they (usually at least) allow me much more time to think. So if I do use the trainer when tired or drunk and my rating falls down below 1600, then the next day I have troubles getting the rating back up again because I obviously take too much time on the lower rated puzzles.

The trainer has a tendency to lead one into making quick moves. Luckily for me, I grieve much whenever I fail a puzzle, whereas I shrug it off easily when I pass and get points deducted. So, all in all, I managed to keep my pass rate >78% during a few thousand puzzles and my goal is to get it up >80% again. 

Nope, I solved a puzzle 22 sec later than average and got -10 point, I failed a puzzle and got +2.

Never mind. Still thinking that when it takes 40 sec to make a check mate move isnt a long time, when I play otb all moves take longer than 40 sec. so why should I move all of a sudden unnecessary quick.

Anyways I had a look at the top twenty of chess.com and almost everybody sucks at it or doesnt play TT, that's good enough for me.

Bubatz makes a good point, it would be better when puzzles not only come from the current rating but mixed with higher and much higher rated problems and lower rated problems too.

And then there is the change of graph which wasnt an improvement.

I just keep on training and forget about time, I fail on time, no problem, it is the TT that really fails, TT fails to see that taking more time is beneficial to me in most games. A mistake is quickly made, better think twice.

JT, I have to disagree with what you're saying and side with Goch.  The main point is that in TT, you know that there is a tactic to be had.  If you can't recognize the pattern quickly enough, then you are going to miss it in a game, when there is nothing inherently telling you there is a tactic available.  If you say you mainly play slow games, then you should be spending that extra time looking for even deeper tactics, rather than having to use it up on tactics which could have been quickly recognized with better pattern recognition, leaving you less time for the deeper tactics.

In some sense a whole chess game can come down to how deeply you can think about a position per amount of time, since your opponent is given the same amount of time as you.  As they say, in the majority of grandmaster games there are few tactics, simply because both sides are actively preventing each others tactics.  If you are using a longer amount of time to solve a given problem than your opponent, then on average you are going to allow deeper tactics for your opponent than he will for you, and thus lose.

Basically, no matter what time controls you play OTB, time is an issue.

Also, since puzzles have different ratings, it seems perfectly plausible that you lost 10 points even though you solved a certain problem, but lost less points for solving a different problem in an even longer time.

Also, Bubatz, I think that guessing quickly in an effort to boost your rating is a fallacy, as you might get lucky and guess correctly in the short term and have a rating boost, but soon you will be getting tougher problems and those quick guesses won't be cutting it, losing you points.