The problem with puzzles is that you know in advance that the solution will deliver a decisive advantage. In a real game no one tells you that in advance. You're looking at the board, and possibly thinking: now's a good time to castle and tuck away my king, or consolidate my pieces, etc. But in a puzzle, you know that won't be correct, just because that's not how puzzles are designed. An ideal trainer would not give just standard puzzles. It would randomly throw at you any possible game situation, where the best move could be a banal positional or defensive move, forcing a draw with an otherwise lost position, gaining a small advantage, etc., along with the standard decisive advantage puzzles. The latter would then be harder since you would not know in advance that you could get a win from that position.
This is a valid point, especially for higher rated players, but for lower level chess players like me, this would be a nightmare. Not only would I struggle immensely at solving puzzles such as these, but it'd take away from what I think is the point of puzzles for people like me. Doing puzzles trains you to spot potentially advantageous tactics and plans in real games, and though you're certainly right that these positions aren't always present in any given game, it helps with the mindset of always having a plan or some tactic in mind when playing.
I have no first hand experience with high rated puzzles either, but I believe many of them do begin with seemingly benign and passive move, even if the variation then gives a decisive advantage. Small choices are much more heavily punished in high rated puzzles (and games) than in low rated ones, where there's often no other move that could even seem as good as the best move given.
This is just my experience and opinion as a weaker chess player, and as someone who doesn't play much regular chess (puzzles or games), perhaps I don't have the most informed opinion.
I have no first hand experience with high rated puzzles either, but I believe many of them do begin with seemingly benign and passive move, even if the variation then gives a decisive advantage. Small choices are much more heavily punished in high rated puzzles (and games) than in low rated ones, where there's often no other move that could even seem as good as the best move given.
This is just my experience and opinion as a weaker chess player, and as someone who doesn't play much regular chess (puzzles or games), perhaps I don't have the most informed opinion.
ChessHero is able to select from any position from various games provided. There are also other puzzles than the obvious tactical ones, which allows more advanced players to evaluate minor positional and defensive themes.
The problem with the more "random" positions is that player styles become more important than the +-0.04 difference that a computer would calculate. That is, on the first move which move is best?
The problem with the more "random" positions is that player styles become more important than the +-0.04 difference that a computer would calculate. That is, on the first move which move is best?
OP has an excellent point. One that isnt available anywhere but I hear over and over.
Another thing about puzzles is that they are guaranteed to have a single correct move. So if you find 2 or more good moves then you think them out until you are left with a single correct choice, otherwise just keep thinking.
Nonetheless, I find puzzles very interesting and have spent hours solving them. I think they are a good way to polish your tactical thinking. Sometimes its too easy to miss some line, too easy to be overconfident. At other times its simply impossible to find any winning move.
Nonetheless, I find puzzles very interesting and have spent hours solving them. I think they are a good way to polish your tactical thinking. Sometimes its too easy to miss some line, too easy to be overconfident. At other times its simply impossible to find any winning move.
Having a distinct "best" move is part of the nature of the format. If 10+ moves are playable (and different engines/grandmasters might come to different conclusions), what's the point of rating your choice?
That said, I'd also like to have puzzles of the type "find the only move to hold the draw" or "find the only move that positionally punishes the opponent's mistake on their last move". Should not be too hard to implement, too.
That said, I'd also like to have puzzles of the type "find the only move to hold the draw" or "find the only move that positionally punishes the opponent's mistake on their last move". Should not be too hard to implement, too.
Same answer as the one in your other similar thread:
lichess.org/forum/general-chess-discussion/ideal-trainer-versus-standard-puzzles#2
lichess.org/forum/general-chess-discussion/ideal-trainer-versus-standard-puzzles#2
chesstempo has a mode called "mixed tactics", there 80% is a normal puzzle and 20% is a defensive move to hold the position and u dont know that beforehand.
if u pay for a membership u can also do only defensive move tactics, "guess the move" is also interesting u play through a famous master game move by move and it rates how good your guess was for every move.
if u pay for a membership u can also do only defensive move tactics, "guess the move" is also interesting u play through a famous master game move by move and it rates how good your guess was for every move.
Oh, sorry for repeating the thread, Sargon. I couldn't see the other thread so I thought that for some reason it wasn't posted.
The comments are helpful and appreciated. Of course working through standard puzzles is enormously useful; no one disputes that. But players at all levels need to learn all the different facets of the game, not only winning tactics. Moreover, some weak players develop a bad habit of always thinking there is a tactic available, when there often isn't; indeed usually there isn't. Strategy is just as important.
Jonesmh, you make a good point. However, there are non-winning best moves - defensive moves, strategic moves, etc. - which are uncontroversially correct for any player of any style. I think a program should be able to select them, e.g. by selecting moves that give a significant advantage over the next best alternative.
Regarding your rhetorical point about which is the best first move, that's actually a good question which I've thought about. Swordfish doesn't always open the same way, but I wonder if stronger programs do. If I knew one first move was best, I would work on that opening.
The comments are helpful and appreciated. Of course working through standard puzzles is enormously useful; no one disputes that. But players at all levels need to learn all the different facets of the game, not only winning tactics. Moreover, some weak players develop a bad habit of always thinking there is a tactic available, when there often isn't; indeed usually there isn't. Strategy is just as important.
Jonesmh, you make a good point. However, there are non-winning best moves - defensive moves, strategic moves, etc. - which are uncontroversially correct for any player of any style. I think a program should be able to select them, e.g. by selecting moves that give a significant advantage over the next best alternative.
Regarding your rhetorical point about which is the best first move, that's actually a good question which I've thought about. Swordfish doesn't always open the same way, but I wonder if stronger programs do. If I knew one first move was best, I would work on that opening.
binit ты случайно не русский?Скажи
This topic has been archived and can no longer be replied to.