Hi,

i'm trying to create some endgame-training, for instance white king and pawn against black king. First i enter the position, then the possible moves for white and for black. So far so good. My problem is in training the positions: if there is more than one possible/good move for white, i can't train it.

What i tried to do:
1) If there is a position for white where there is more than one white move that is correct i've marked every good white move as "B" (Best);
2) In the training section, i clicked on "all best moves" as "right moves".

So if i came to a position, where more than one white move is ok, i have trouble. Lets say Kb5 is ok, Kc5 is ok and Kd5 is ok. If i choose Kc5 CPT is accepting the move as a right move, but then positions the King on the other field (b5) and tests me from there on. So i always have to train all possible moves, but instead of this, i wish only to train one "line" and if my moves are correct than why should i train other correct moves? But if i dont enter the moves in the first place, then if i make a correct move, cpt says its wrong, because its not a canditate move.

Or to say it with other words: I have lets say a tabiya with mate in 2, see attached position/graphic:



There are 6 possible first moves for white to make a check mate in 2: Qg5, Qc7+, Qg4, Qg3, Qb8 and Kf7. I would like to "train" this Tabiya position, but it should be marked as "solved/learned" if i enter one of the possible ways for check mate in 2.
Right now, i have to train all possible moves to solve the tabiya as learned, this means i have to enter more then 50 moves (6 first moves, then some black moves, then some more white answers) in the training section to solve the tabyia...

Greetings,

Thanks for reporting this. It looks like the author didn't consider this case or better didn't realize that it can't be handled the same way as it is done for openings.

I have to check how much work it is to fix it. Probably the easist fix would be that CPT would still stick to the (first) variation it had in mind, doesn't consider the other played best move as wrong and doesn't go through all alternative moves for player's side.

So call me befuddled,, but if you are playing and training for the winning side in a certain position, why would you need to know all of the acceptable moves? Does it really add enough value to your knowledge. Isn't a win a win? Wouldn't knowing more than one acceptable move in a single end-game position be more rather than less confusing?

Just a thought.

I think he is saying there are position with 'multiple' acceptable moves...no specific 'winning try'.

I don't use this feature...use CPT just for openings training, I'm not sure many do. I see CPT as an program to 'train openings'.

@ AlekhinesCat: the person who tries to solve the tabiya has to enter only move (for example Kf7) and your right, that this person doesn't have to know/enter all of the 6 possible correct moves. But CPT has to know all the correct moves! So if I enter 6 possible moves, CPT only accepts one of the 6 moves (for example Qg4) in the training mode as correct, but not the other 5. So if the person who tries to solves the tabiya mayby enters a correct move like Kf7, CPT reports that this is wrong...

@ leavenfish: for this reason that endgame-training doesn't work in CPT, I also use it for opening-training only