Tactics Trainer and Type of Problem

Can we get some way of knowing what type of problem we're supposed to be solving and, possibly, about how many moves?  In some we're looking for checkmate in one move; in some we're looking for checkmate in six or so moves (I'm looking at the 600-700 level I think); in some we're looking to get out of having a piece taken.

"Problem solved" for something -- but, for what?  I'm looking to checkmate and the problem is looking for me to keep a piece from being taken.  Or I'm supposed to look for checkmate and allow pieces to be taken all over the place and I'm looking to keep a piece from being taken.  I never know what the problem is trying to get me to do.  So, if I play to checkmate and all I'm supposed to be doing is to keep a piece from being taken, my answer is wrong; if I play to keep a piece from being taken and I'm supposed to be looking for checkmate with several pieces exchanged, I'm wrong.  Thank you!

It does not tell you this in games, so why should it in tactics?

 Because in games we can do what we choose to do -- make the strategies we wish to make and use the tactics we wish to use.  Perhaps, for instance, we have a long-range plan.  Bobby Fischer once played a game.  Two grand masters (I believe they were grand, not just masters) were in two different rooms explaining it.  Both said he was losing; he was making some very strange moves and giving up pieces for pawns.  He won.  If I can see a mate in five moves, but the computer has it in program that I take a particular piece and has its own plan for only one move (take the piece) and the problem is solved, not knowing that this would negate my plan for mate, it would (and does) tell me I made the wrong move.

The tactics trainer always looks for the best move, as should you in your games

 Okay.  Here's an example from a game I played.  I was forking a knight and a rook with a bishop.  I took neither for quite a few moves.  I reasoned that neither was going anyplace fast; there were other things at stake.  I couldn't put the king into check, but I could make things difficult elsewhere.  It wasn't until 8 or 9 moves later that I finally took one of the pieces and ended up with both of them because in the interim I'd been able to move my queen into position to do this; it also allowed me to get rid of another piece that tried to protect one of the two -- the piece that was keeping my queen from checking the king (I'm not positive of the exact details, but it was very similar to this if not exactly as I've described it).  I'm sure the Tactics Trainer would have told me at level 600-700 that I'd made the wrong move, expecting me to take the rook.  This check was the beginning of the end of the game for my opponent.

If the Tactics Trainer had said "one move" or anything of that nature, I'd have taken the rook rather than thinking down the road.

I think that the fact is that all the answers to the tactics trainer problems are the best move in the opinion of a top chess engine. Therefore you will only get it wrong if you don't play the best move.

If I don't get the best move but get some pretty decent alternative, shouldn't I get some points or something ?

I suppose this would require a reprogramming of the tactic trainer. Personally I don't do it for the rating, but for the improvement of my tactics.

If there really was forced mate in 5, tactics trainer wouldn't have marked it wrong. But if the best move is really to take the piece...

tactics trainer is teaching you to look for the best move.

you are (strangely) asking that you be either told which type of move is best beforehand, or to get partial points for getting a pretty good move versus the best move. in other words, you are asking for tactics trainer to reward you for not playing the best that is possible. kind of a strange request imho. would you feel better if instead of the "Fail" prompt coming up, it said something like "That's a legitimate move, but not the best move, sorry" - would that be satisfactory enough for you?

i strongly hope that tactics trainer does not change like you want it to.

i don't understand your point here. during this game did you know that leaving the two pieces pinned was the best move? then you can figure this out on tactics trainer as well (and become an even better, stronger player).

"I'm sure the Tactics Trainer would have told me at level 600-700 that I'd made the wrong move, expecting me to take the rook."

wrong! tactics trainer always makes the best move. if taking the rook was best than you should have taken it. if leaving the pieces pinned was the best move than you shold have left them. tactics trainer doesn't change it's opinion or look simply one move ahead. they are all tactics and therefore all moves are either the one right best move, or the multiple wrong moves that ignore the proper tactic.

EDIT: tactics trainer does not offer multiple correct moves for a problem, and if there is a problem that does, report it and it will be corrected or removed.

I think I understand this now, Trigs.  Thank you. :)

"wrong! tactics trainer always makes the best move. if taking the rook was best than you should have taken it. if leaving the pieces pinned was the best move than you shold have left them. tactics trainer doesn't change it's opinion or look simply one move ahead. they are all tactics and therefore all moves are either the one right best move, or the multiple wrong moves that ignore the proper tactic."

While were talking about tactics trainer........

My standard rating is between 1500-1600 however on tactics trainer my rating struggles to reach 1100. I know ratings don't mean a great deal but I would be interested to hear what rating other people in my standard range achieve.

The biggest problem for me is time, I'm getting on a bit and maybe don't see things as quick as I should. Do any other players in their fifties find time a bit of a problem in tactics trainer ?

Believe it or not I was wondering those very things a few minutes ago when I was working on some, and I agree completely. I don't believe in partial points or anything like that, but you should at least be told what you're expected to do in that particular puzzle.

 I'm certainly not in your range, Podge; but I'm very slow.  I'm new to chess and in my 60's.  For one thing, I prefer to double and triple check before I answer.  I use Tactics Trainer in practice mode so, hopefully, I'm not changing the overall statistics.

However, I wanted to answer the part about the discrepancy.  I've played three rated games on this site, all with the same opponent.  He asked for a draw after his first move because he moved the wrong piece.  I asked for one several moves in because I moved the right piece to the wrong square (even with the submit button I sometimes get things wrong because of eye tracking problems).  His rating just before I won the third game was 1312.  I suppose because of the two draws, mine had risen from the initial 1200 to 1251.  Weird enough.  Then, after I won the game, mine rose to 1403.  So, 1251 beat 1312 and rises about 150 rating points.  This seems a bit strange.  In no way should I be playing (when I again begin to play rather than study) anyone who is much better than I and I'm nowhere near in ability what 1403 would be on Yahoo where, if you win or lose against someone who's rating is close to yours, the numbers don't change drastically.

I think I need to play someone with a 2500 rating and, of course, lose; maybe mine will get back to where it belongs. :)

Ideally, there would definitely be a difference between making a strong but not-quite-perfect move (perhaps with a still winning game), and, say, making a move that drops your queen.  Doesn't seem like it would be that hard to program either.   Too bad it just doesn't work this way.

and now, woodshover,  let me ask you this:  in a game, are you told what you are expected to do?  (other than the rather vague instruction:  WIN  --- of course).

KittyA, your rating will 'adjust' as your games progress.  it will rise when you win, and drop when you lose.  losing to a 2500 rated player when you are about 1500 would not drop your rating much if at all.   if you want your rating to drop, lose to a 1200 rated player and it will go down.

    as you play more games, it will change less after you have played more games, but always rise when you win, and drop when you lose.

   podge52, as your familiarity with tactics rises, you will learn to spot more of them more quickly, so hang in there and keep working them.   a possibility might include studying tactics found in the right hand column, here, just scroll down.

http://blog.chess.com/batgirl/content-summary2

look for mating patterns, traps, and judy's problems.

browse to your heart's content in batgirl's stuff, she is the recognized world expert on Paul Morphy, America's first champion.

The rating changes become much more realistic the more games you have completed. Play a dozen or so turn based games so you have plenty of time to study the positions and you will see your rating settling down to your level.

Yeah -- I DID get that backwards, Ret. :)))  Thank you for the information. :)

"KittyA, your rating will 'adjust' as your games progress.  it will rise when you win, and drop when you lose.  losing to a 2500 rated player when you are about 1500 would not drop your rating much if at all.   if you want your rating to drop, lose to a 1200 rated player and it will go down."

[  EDITED name of site.  erik prohibits advertising other chess sites on his site.], if you give a decent move, they do not flag your move as a "mistake", points deducted, blah blah blah - The say "Keep Looking   ... was a good move, but not the computer's first choice. Please keep looking for another move".

There is a lot of difference between a good programming and a mediocre programming...