In terms of these unequal (pointwise) trades, would you play them in a game?

1. Bishop-Knight for a rook

2. Rook-Bishop for queen

3. 3 pawns for bishop OR knight

Please feel free to add more *REASONABLE* and i cant stress that enough trades that might occur in a game. Thanks and continue to post in my forums!

Trades is one of the most important features in chess. Botvinnik said something like "chess is a game of inner-related exchanges"

I have moved away from teaching value of pieces as static. I introduce their static value but stress that value is based on action. so the value of a bishop is something for each player to evaluate based on the position. Rooks in the corner are 1? 2? 3? your choice but its in relation to the activity of everything else on the board. A captured piece is worth ,...0!

The reason i think this is more beneficial than using a static value is pieces are  dynamic units. A Bishop is worth more in an open position than closed one. Rooks are more valuable on open files.

so to answer your question 

1. depends on the position

2. depends on the position

3. depends on the position

1. 2 minors are generally better in the middlegame  but a rook is generally better in the endgame. (Marin has a nice section in his learn from the champions on Tal who liked this R vs 2 minors imbalanced for that reason (i know Tal thinking of endgames huh?)
2. a lot depends on the position  is the bishop on a protected outpost? is the rook active or can get active? are there pawn weaknesses in different areas of the board (queens love those) . how safe are the relative kings? what other pieces are on the board and how active are they to coordinate with. 

3. Pawns for a minor again depends on the factors of piece activity. king safety etc. this one is tricky so you dont see it too often unless there is a sraight foward tactical idea or endgame pattern (active king + 3 pawns vs minor piece and passive king) 

:) thanks alex. I actually coach chess so have most of this written down in various training files. 

Piece "values," whether relative to each other or on some measuring scale, are a general guideline.  A bad Bishop that defends a key square can be very valuable.  A Rook in the corner, undeveloped, in a closed positions without open files or easy ways to create them may be almost worthless - but in the ending on a wide-open board, the same Rook can become a monster and control the game.

It always depends on the position - the exact position, one or two changed pieces or pawns can alter the evaluation significantly.

Also, one more. In this game, pure endgame, evaluate or analyze for fun. Who is winning.

King + 2 rooks   vs.   King + Queen

Thanks for other comments

Depends on position, because if the rooks are connected, it is a  draw, but if they aren't supporting each other, usually a win.

A piece for two pawns is another one along these lines.

Giving up pieces for pawns is a common enough idea in endgames where the threat of queening can give more than compensation.

But yeah, as others have said, it depends on the position.  You have to look at what the pieces left on the board are doing, what comes off is irrelevant.  This is a good habit to get into, and something that coaches commonly admonish their students to do :)

A good example of trades is this one. think Activity+ initative < material

just fast foward to move 12 by white....

Also look at some games by Shamkovich, in the Grunfeld, they are beautiful