Interested in some feedback

When considering, how the rating system works with tactical puzzles, I decided that I disagree with the approach.

If you play a tactical puzzle its very black and white (no pun intended) and narrow, either you are right or you are wrong. However, chess is not so simple.

Why would someone ever choose to be up an exchange and capture a rook opposed to taking a knight for free? Its a simple concept really.

Its a choice to give up the possibility to lose out on an extra pawn to take out a major piece. Why would someone ever head down that avenue?

Well, for one there might be plenty of open files and rooks are far easier to coordinate then minor pieces that don't travel long distances.

Another example of a problem a person would fail is this. Your opponent has a rook and a queen left and the pawn count is even, 3 pawns on both sides and they are locked up in the center. You have 2 rooks left and a queen. At this point you have there king in check and you are forking a rook and the king, the queen blocks the check but the rook is still hanging. You happen to also have an available move to check the king with a rook forcing them to give up the exchange. 

I am pretty sure in a position like this, Most people would opt to win the exchange opposed to winning a piece. The extra pawn simply isn't worth a queen. 

In tactics trainer, you play the move that wins the most material or leads to mate the quickest. If your move leads to mate in four but the problem has a possibility for mate in three. You fail.

I think that you shouldn't be punished for making good moves in 'tactics trainer' you should only be punished for making bad moves.

Interested in some feedback on this. 

At the tactics site chesstempo any move that keeps your position "winning" (I think the threshold is something like 2.00) you wont fail.  If it's not the best move you get a message "That move was good, but not the computer's first choice" (or something like that) and you can keep trying until you get it (or fail it with a legitimately bad move).

Yeah, this is better, but sometimes it's frustrating!  When your first 3 guesses are "try again" and you're thinking "I would have wont his game 3 times by now!!" 

But in any case, the point is to learn some pattern or to push your calculation to see more.  Better to treat it like a learning opportunity (but yeah, it's frustrating to fail one).

shell_knight 

I agree with your perspective on the whole situation and sometimes chesstempo will let you have another chance if you guess a nice move but not the best.

Do you ever get the feeling that in the blitz problems from both sites, chesstempo and chess.com, even though you are timed to play the best move, that time is really

not considered at all?

For instance, I played a blitz problem where you could simplify to a winning end game that would be fairly simple to navigate to a winning game or you could

play the move that lead to mate in 16 and included a queen and a bishop sacrifice. :) The problem was rated 1300 and you only had to play 1 move.

Well I do view 'tactics trainer' as an opportunity to learn, some things are ridiculous. The reason why this blitz problem was only 1300 is because it included

1 move. This one move that you could play was the only move that lead to a mate in 16. However, it was obvious that there were 2 possible options and everything else would fail.

The problem also had a 50% success rate.. hmm lol.

Sometimes tactics trainer also seems to hurt. +4 points rating for the one you solve and -14 for the one you fail on chess.com

Chesstempo however, depending on your RD. You could spend 20 minutes on a standard problem rated 2000 when you are 1600 rated and gain two points but if you

fail a 1600 rated problem, you will get like -10. 

Sometimes, tactical problems, seem to do more harm then good. Say you are trying to prepare for a match up by doing tactics, and you lose rating, you solve

5 strong but failed 2 weak ones. In effect, you will probably get less strong problems and more weak problems because you are losing rating. 

I have a theory, before a match, if you are playing very strong problems and maybe solving 1 or 2 out of 10.. you are way better off then solving 10 of 10

weaker problems.

What do you think?