adding do not know as an option and having static rating for problems

[uri blass]

The idea is simply to discourage guessing by the fact that people who click do not know are going to lose less rating points then people who click wrong move
(it is possible to decide that they lose half of the rating points that people who click the wrong move lose).

I wonder what is your opinion about it.

Another idea is to have static rating for problems.
The problem with not having static rating of problems is that people cannot know if they improve based on rating points because it is possible that the rating of the problems was changed.

Problems may have 2 rating numbers(static rating that is unchanged) and the normal rating that they have today.

The non static rating is going to be used in the decision which problems to give to users but the static rating of the problem is going to be used in the rating calculation of solvers so practically 2 users who got the same result at different time are going to get the same rating number.

Uri

[drahacikfm]

In a real game you have to make a move, even if you don't know which move is best!  People should not be rewarded (by losing only half the rating points) for "quitting" a problem.

I don't see any need for static ratings (difficult questions on how to decide that anyway, even if it was needed).

I don't see anything wrong with the current rating system where the rating of the problem reflects exactly how people have done against that problem.   The ratings of some problems go up, for other problems it goes down.  It evens out. 
There is not a steady large decline or increase in the average rating of problems here.  If there was some large-scale rating drift in one direction, it might be an interesting idea to have some "anchor" problems with fixed ratings, but to change the whole system to static ratings is not a good idea.

If you think your rating does not measure improvement, then look at your percent standing among all people here, for example "better than 80%" and see if that goes up or down.  Of course that could change unrelated to what you do if a lot of strong players or a lot of weak players join suddenly.  But in general that percent should remove any effects of drift in problem ratings, which is there is no evidence for anyway.

[andreacoda]

The idea is simply to discourage guessing by the fact that people who click do not know are going to lose less rating points then people who click wrong move
(it is possible to decide that they lose half of the rating points that people who click the wrong move lose).

I wonder what is your opinion about it.

Personally, I don't think it is a bad idea. I understand where you are coming from, and in effect I can see the value of discouraging guess work.

[tama]

Is the idea to allow the user to quite the problem if they cant see the solution and only loss half the points? If that's the idea i think its ridicules, your basically saying, ok if you cant solve it don't try!

[uri blass]

My blitz rating was most of the time 2150-2200
I stopped to solve for a week or something like that and 
in the last days it is 2100-2150 most of the time.

 

I do not know if the reason is that I performed weaker or if the rating of the problems became smaller or the average time of the solution became faster.

my place relative to other is irrelevant because it is possible that other
people improved.

I am interested mainly in blitz rating because my standard rating is relatively higher and I want to improve in the part that I am relatively weak.
 
I probably need to look at problems that I failed and analyze why I failed them to perform better.

Note that I find the fact that the squares of the last move are in different color annoying and I can probably concentrate better if I do not see different
color for the squares of the last move but the same problem was when I had rating of 2150-2200 most of the time.

Edit:
About my idea of reducing less points for people who admit that they do not know
then I do not consider it as rewarding for quitting a problem because you still lose rating points.
nobody solves 100% of the problems.

real chess is not finding winning moves and in most of the position there is no win
so do not know does not mean not going to play this position but means that the person did not find a winning move. 

Uri

[drahacikfm]

Note that I find the fact that the squares of the last move are in different color annoying and I can probably concentrate better if I do not see different
color for the squares of the last move

You can easily turn that off in the Preferences.

[uri blass]

Note that I find the fact that the squares of the last move are in different color annoying and I can probably concentrate better if I do not see different
color for the squares of the last move

You can easily turn that off in the Preferences.

Thanks.

[uri blass]

some more thoughts about this subject.

I think that it may be interesting to know how many solvers guess and how many solvers know so users will have to click guess or know before they play a move.

Hard problems can be hard problems because of different reasons.

One reason is that people have no idea what to do and another reason is that people believe that they found some winning move when practically they missed a defence of the opponent and finding that the correct move is winning is not very hard.

I will call the problems that are hard for the second reason misleading problems and I think that misleading problems should get higher rating relative to what they get because often weak players solve them by luck of not considering the move that seems to be winning and their performance against strong players is better.

Uri

[tama]

I think the biggest factor to guessing is in blitz mode. If you guess in standard then i think your not using your time very well and it would probably more beneficial for that user to do something else.

I understand the good to having this but its just not realistic and people will start only playing if they know the answer. But for lower rated players i see this as very help full for there rating.

[uri blass]

I clearly use more time in standard mode and I plan to try to solve problems in standard mode in the future but no human plan to spend all his life to try to solve a chess problem and there may be cases that after enough time people have to admit that they did not see a solution to the problem and guess some move.

The time that people may give up can be 10 minute or an hour or 2 hours
but I expect everybody to guess in part of the cases assuming that the problems are too hard.
 
Maybe you never got problems that are hard enough for the time that you are ready to spend but even in this case you are in a small minority and I believe that at least 99% of the people who use standard mode got  problems that were too hard for them to solve.

Uri

[uri blass]

I can add that I am better than 97.14% of the people here at standard mode.

I did not solve a lot of standard relative to blitz but my relative ranking at standard is clearly better than blitz.

I do not remember if there were problems at standard that I felt after seeing the solution that I could not solve it but
there are problems in blitz that I know that I could not solve in case of getting them in standard time control because I simply rejected some move in the solution as obviously illogical move and it seems to me that even thinking for hours could not change my opinion that the move is illogical and in case of hours of analysis I could probably analyze again different lines but not the correct line.

Uri

[tama]

lol I don't expect people to spend all their life on a problem either, but if you spend 10 mins or more on a problem then i think you have some ideas that seem like they might work, so you go for one, and this is guessing in a way. I still think if you put in the effort you should go for it! And i have gotten problems i couldn't or didn't solve when i was still learning and its because of those problems were i learned the most.

I think the best solution is to play UN Rated problems like i do. I made a problem set were im only served 1900+ problems, and i get the same problems i would in standard mode but i don't loss points if i fail.

Well all in all, its all up to Richard!

[uri blass]

I also learn from problems that I do not solve but do it mainly in CTS


Here is one example when I failed to find the only saving move for black
http://chess.emrald.net/chess.php?ReviewPos=28546&Mode=review

I believe that many players with rating above 2000 are going to fail here.

My strategy in CTS now is also not to guess a move.
I pick a move only if I stronglt believe that it is the best and it means that I can think even for some minutes on some easy problems but I compensate for it by thinking less time on easy problems or on problems that I remember.

Note that also in cases that I see 2 moves that seems to me the same
I try to think which one of them is good because I know that it is impossible to have 2 good moves with the same idea when it is CTS problem and not games.

guessing one of them can be better strategy for rating points but rating points is not my only target there.

My target there is to improve both my rating and my accuracy.
 
Uri