lichess.org/training/63981i've seen 1700-rated puzzles harder to solve, i might have been lucky not seeing a trap, unlikely though
and rook f8 to e8 loses with 20.h7+f8 21.xf6, it's the only other line and still pretty obvious
Because it looks like Rxf6 is playable, I think.
it does and i've thought about this line but dxc4 comes quick in mind and still, the move doesn't look as powerful as queen f5 which is the very one you want to play
Rxf6 looking playable is not enough for this puzzle to be rated 2300 imo
The puzzles rating is a result of how many people got it right/wrong, opinion never enters into it. It's based on a large amount of data.
I run into this sort of thing too where I have no idea why a puzzle is rated so hard or so easy. I think it's just because people's chess brains are highly idiosyncratic.
Nojoke -> that's what i'm kind of talking about
if i'm right, 2300 in elo means that with a 2300 training rating (which is pretty strong) you will statistically miss the puzzle one out of two times
there sure is high diversity among the ways chess players think but in this case i simply don't get it, feels like an anomaly
Yeah I suspect 1 Rxf6 gets tried a lot (intending 1...gxf6 2 Qf5) but after 1...dxc4! it's complicated.
It would be cool to have chesstempo like stats on what moves people played in these tactics puzzles.
@MoistvonLipwig Love your username tho
