lichess.org
Donate

Training puzzles: Mate in N vs Material advantage

Great point, Clarkey... these puzzles are ideally meant to improve your real-game tactical calculation and intuition. If a player cannot sense or calculate a mate over material gain, they need to work on that! (by doing more puzzles :))
Except that in training you usually have a coach to tell you. Even the great Grandmaster puzzles in books differentiate between "mate in x" and "plays and wins." Training isn't the real world, it's preparing you for the real world.

Now, that being said, I find no error in your logic Clarkey. It's just as easy to say not knowing is better because it prepares you for real world situations otb. I can't really argue with that logic. Learning to differentiate between a won position and a forced mating position might be very useful.

I could, however, argue that the reason GMs who put puzzles in books and most online puzzles coach you as to which is which because the more you see positions that are "plays and wins" vs. "mate in x" you are in fact building up a way to start recognizing the difference, systematically, just by viewing the positions and not even succeeding at the puzzles. You will see patterns that your brain identifies as "tactics can win material" and patterns that your brain identifies as "I should look for a forced mate here."

I feel as though the reason coaches have generally always distinguished between these kinds of puzzles is not because it isn't important to identify those situations otb...but rather for that very reason!
#12 is somewhat accurate :)
Lichess puzzles are not intended for puzzle training like a book would be (no curator assembles the puzzles by theme/motif, from easiest to hardest to help you learn)... so not sure you can compare them. Lichess puzzles are for puzzle 'playing'! (much like blitz is for chess playing, not opening training... you need to look elsewhere (books etc.) for this)

Also, maybe it's a beginner thing, none of the puzzle books I have (mainly chess informants and quality chess books) ever tell you whether it is mate or material. They just tell you the side to move.
static_shadow, you just need to do 50 more puzzles. First hundred for me was a real source of frustration when the only thought you have for 10 minutes is "how on earth the checkmate is possible here?" Then you come up with something and it's a perfect fail. The best move turned out to be winning the bishop in ~5 moves. But then it all gone, now I'm totally agree with Clarkey's point. If you want to learn play better then get rid of helping clutches, just find the best move in any position.
I've always been of the opinion that categorized tactical puzzles (mate in N, white to play and win a pawn, etc.) are sort of the worst of both worlds.

For instructional purposes, as presented in the #12 post, then that's much better achieved by an instructive text that presents tactics in various forms, explains the common structural themes, and gives suggestions for spotting them in a game.

For training purposes, it's much better (obviously in my opinion; there haven't exactly been rigorous studies comparing their efficacy) to just have to find the best move, without knowing what it might achieve.

That's one of the reasons chess.emrald.net is far and away my favorite trainers. Not only does it not tell you what to expect, the only criterion the puzzle has to meet is that each move must be an "only" move, in that all other moves are much worse.

The solution could even just be a move that maintains equality, if all other moves lose.

They also take time into account quite aggressively, which I find helpful (and for them it helps prevent people from cheating to get a high puzzle rating).

Either way, if you're wanting to receive some sort of instruction or education about tactics, puzzles aren't so great. Pedagogical texts are much better.

If you want to test and hone your skills for application in a game, then uncategorized puzzles are superior.

Categorized puzzles just strike me as some attempt to provide some minimal instructional content, while also being an exercise, and I think trying to do both just makes it do neither particularly well.

Obviously all the above is just my $.02. I think it's based on solid evidence and experience, but until there's a large, well-controlled study to compare effectiveness of the various training methods, we'll never really know :)
I believe if we practice something, we can only get better at it. Practice the same type of puzzle and it can only lead to mastery.
We need to learn chess the same way we learned mathematics or how we learned to play an instrument like the flute or the piano.
We learn everything one-step at a time. Keep things simple. Some of us learn by comparing things. I am sure there are other methods to learn stuff.

Let me state clear and for the record, I personally dislike chess puzzles in general. I prefer instructional resources.

That being said, I was simply stating the simple point that the likely purpose that many puzzles one might find in books (and no, it isn't "just a beginner thing") or on many other places on the internet is because a lot of chess is about pattern recognition, and if one learns to recognize mating patterns from starting positions vs. positions that there is no forcing mate available from the start of the position, one benefits from having those patterns in the database of their mind.

For the record, I never intended to state that one way or the other is somehow better. I have my own preference, which is that they are categorized. Clarkey and others have their preference, which is that they are not. I don't think either is honestly better than the other, they both have pros and cons. I was simply pointing out the pros of my preference. Also, thanks for those of the same mindset who reaffirmed my logic. :)

This topic has been archived and can no longer be replied to.