Look at the position after Black's 10th move. White has three(!) loose pieces on the board. With that much floating around--almost randomly placed, the question of which pawn capture is best is really secondary. Something is bound to pop and end disastrously...as it does.
dxc5 or cxd5 ???

Thank you very much, people.
I was thinking Nxd5 was impossible because
on d5 square the materials are 3 to 3 and
it looked for me that I wasn't able to capture voluntarily.
BUT I was misunderstanding,
one of my 3 was not a piece but a pawn,
on the other hand all of opponent's 3 was pieces.
Good lesson for me,
thank you very much for your analysises! I'm grateful
My annotation is included in the diagram above.
My question is on 11th move.
Here c-pawns and d-pawns of both sides are facing at the center, keeping the tension.
My brain hung up about what to do.
I think I have 3 options, which are,,,
(1) cxd5 (I played this)
(2) dxc5 (computer recommends this)
(3) I don't capture any and let my opponent play cxd4 or dxc4.
The reason I played cxd5 was simple, a flank pawn is less valuable than a central pawn
so I thought the exchange of my flank pawn and opponent's central pawn would
nice for me.
I believe this way of thinking is partly correct(?) but actually it was wrong in this game.
Computer recommends dxc5 but I don't understand why.
I'm wondering that whether "dxc5-or-cxd5-problem" depends on
positional aspect of tactical aspect.
As I said I thought cxd5 was nice positionally, but I can understand it was not good tactically,
because I had to recapture my d4-pawn with my Queen and
the d4-square was aimed at by the opponent's Bishop.
But I'm not confident about whether I can choose the correct one in the next game...
I'd like to know HOW I should think...
Thank you for reading.
Thank you!