By its nature in tactics we're always playing on the winning side. But, most of us suffer losses due to blunders, not from missing a great attack. With the available database for blunders, can't we have puzzle questions that ask the best move right before the blunder for the side that originally walked into a tactic.
As our rating gets higher on this regard it may slowly turn into don't make the lesser move, mistake, inaccuracy, perhaps.
What you're asking for is completely relative to the player and so virtually impossible, unless there were an entirely new tactics trainer built to pull only from your games to find these blunders.
The thing is, learning to prevent blunders isn't the goal of tactics training. Tactics training is designed to aid in pattern recognition and tactical ideas like pins, forks, skewers, opening lines, and ultimately mating webs.
I'd argue though that by doing these sort of puzzles, you learn to recognize the best move in many sorts of situations, and in doing so you are less likely to choose the blunder because it won't even be in your thought process.
@Fenris1066, what I'm asking is:
1.Select blunders from analysed games (just as tactics trainer does)
2.Take back one move
3.Ask the best move
So, I don't think it is impossible or player dependent. As you said tactics trainer has another goal in mind, and this is exactly the reason I'm asking for a "new" kind of tactics trainer to be added.
What I mean is that it's completely player dependent because what one player misses and makes a blunder over another player might have seen for 10 moves.
For the record, the tactics trainer already does this quite a bit for missed mates and the like in the actual game that was analyzed. I really just don't understand what you want or if what you want is even possible...
Is there anyone who understood what I was trying to explain in #3 ?
Yeah. You want puzzles made from your own games.
No, I mean the overall analysed games pool.
There are many potential blunders from most positions. Just because you go back one move from a blunder doesn't mean that's the only potential blunder, or even necessarily the worst blunder, someone could make from that same position. Indeed in many cases there could be a huge number of blunders. Where your queen is threatened and you make any move that fails to protect it for example.
It is correct that there is an abundance of blunders in mostly any given position. But, my idea is, this blunder space is filtered out by human players. So, when we look up the moves in the analysed games database that are tagged as blunders, i think it is possible to filter out most of the "too-apparent" blunders out of such blunder-space. Ofcourse, the filtering effect would be dependent on the ratings of the players that made them. For example, comparing two pools that are formed by blunders of all 1600 to 1700 rated players and one with 2100 to 2200 players, I would expect the latter would consist of blunders that can only be realized upon very close inspection - hence the potential for a sincere "dont blunder tactics" puzzle. Thank you for the valuable input by the way, I didnt think of that before.
On the surface this thread seems really pointless, but I think I know what you're actually wanting.
The point of learning tactics is to choose the best move. The point of not blundering is not to choose a sub-optimal move. Isn't it really the same thing? And as I stated before, quite a lot of those tactics puzzles, if you look at the actual games, are derived from positions where the player *did* blunder by missing a forced mate or forced winning material gain so you already have what you're asking for in many cases, which is why we're confused.
But I think the key thing you seem to be trying to point out is that you are noting that players often blunder in equal or near-equal positions, which that isn't covered by the trainer, because they are often the blunders that least to those forced tactics found in the trainer. There really is no way to design tactics puzzles like what you want, though. That's because identifying points where certain moves are obvious blunders is a different sort of training...it's //positional// training:
Examples of positional awareness:
* I can't move this piece because then that piece is unprotected.
* I can't move that pawn because it hangs a piece.
* I need to protect this piece over here because it's hanging and the opponent could easily snag it in a double attack on a stronger piece that I have to move.
These things aren't "tactics training" they are positional awareness, and that's where the majority of blunders arise from equal positions. The others are from missing forced tactics that win material or the game, which are already covered by the tactics trainer.
So maybe, more accurately, you are looking for a positional trainer? I'm not sure how one would go about making one of those.