We all know finding tactics in chess puzzles is much different than finding them over the board.
The crucial difference is that - by the very definition of a puzzle - you know there is SOME tactics to begin with when looking at a puzzle.
What if the Training feature would sometimes serve positions where no tactical shot is actually present? And there'd be an optional answer"no tactics here" to pick once you concluded you can't find anything (obviously you'd get it wrong now and then)?
* For the sake of those who like the feature as it is, it could be entirely optional and disabled by default; just add a switch to the profile settings
* Finding positions that make good "fake puzzles" should be fairly easy to handle algorithmically - simply find highly complex positions in which all the top 2 or 3 moves have similar evaluation.
* The percentage of "fake" puzzles could be fixed at, say, 10-15% when the setting is on. More nuanced implementation would, of course, be possible, such as adjusting the rate in real time somehow; such as start throwing more red herrings if the player consistently does a good job solving puzzles
The rationale for this feature is that it removes the certainty that the tactics is even there, adding much more resemblance to a real-life game. Also I haven't seen this on any other site.
The crucial difference is that - by the very definition of a puzzle - you know there is SOME tactics to begin with when looking at a puzzle.
What if the Training feature would sometimes serve positions where no tactical shot is actually present? And there'd be an optional answer"no tactics here" to pick once you concluded you can't find anything (obviously you'd get it wrong now and then)?
* For the sake of those who like the feature as it is, it could be entirely optional and disabled by default; just add a switch to the profile settings
* Finding positions that make good "fake puzzles" should be fairly easy to handle algorithmically - simply find highly complex positions in which all the top 2 or 3 moves have similar evaluation.
* The percentage of "fake" puzzles could be fixed at, say, 10-15% when the setting is on. More nuanced implementation would, of course, be possible, such as adjusting the rate in real time somehow; such as start throwing more red herrings if the player consistently does a good job solving puzzles
The rationale for this feature is that it removes the certainty that the tactics is even there, adding much more resemblance to a real-life game. Also I haven't seen this on any other site.