Which of these trades are worthwile

[Perfection]

Which of these trades do you think are generally beneficial?

 

[punkbass2000]

Are these trick questions?

 

[Perfection]

Nope, just gettong some opinions on the more difficult decisions

 

[ummmm........]

By no means a chess expert, I don't understand what I could use to make these decisions other than the piece values. My answers would change in different situations, but in the general case, it seems like it's just arithmetic.

 

[Perfection]

By no means a chess expert, I don't understand what I could use to make these decisions other than the piece values. My answers would change in different situations, but in the general case, it seems like it's just arithmetic.

I'd like to see that arithmetic

 

[punkbass2000]

Well, generally speaking, the rule of thumb is:

Pawn = 1
Knight = 3
Bishop = 3 (sometimes 3.25)
Rook = 5 (sometimes 4.5)
Queen = 9

Of course, there's more to a game of chess than simple arithmetic, but without context I would consult this standard without question, and, like ummm....., answered accordingly. Personnaly, I was expecting to see arithmetically equal yet ambiguous trades, which do come up sometimes. For example, I would, assuming we're approaching the endgame and I have no particular plans for my pieces, trade rook for two pawns and a bishop, but not two pawns and a knight, generally. I would also trade my queen for any three minor pieces (bishops or knights). All of your examples, without qualification, are pretty straightforward, IMO.

EDIT: Interestingly, IIRC, ummmmm....... and I voted identically, so it is easy to deduce what the third voter voted for. All '3's we all voted for, all '0's none of us voted for, all '2's are from ummmm..... and I and all '1's are from the anonymous third voter. I'd be rather interested in the rationale of the votes that don't line up with ours.

 

[ummmm........]

Yup, that was my math, except that I've never seen a rook valued at anything other than five. Again, I'm not holding myself out as an expert.

I can't think of a whole lot of situations where trading two knights for a rook is a good call. Maybe in the endgame. It looks like the third poster is overvaluing rooks generally, if you look at the vote pattern.

EDIT: The third guy also said he'd give up both rooks for a queen and a bishop, but not a queen and a knight. Maybe he values his rooks at six?

EDIT EDIT: I can't figure it out. I think he's valuing his rooks and more than five and his knights at less than three. But then there's something else going on with the both rooks v. queen and knight/bishop thing. Plus there's several where he didn't pick either. I'm sure it makes sense, but I just don't have the math skills at this hour on a Saturday.

 

[ummmm........]

I just saw the first two versions of this poll. Sometimes you're the windshield . . .

 

[Richard Cribb]

Generally speaking, it would be like this:

Trading a queen for two bishops = Unfavourable
Trading two bishops for a queen = Good
Trading a rook for two bishops = Good
Trading two bishops for a rook = Bad
Trading two knights for a rook = Bad
Trading a rook for two knights = Favourable
Trading a rook for a knight and a bishop = Good
Trading a knight and a bishop for a rook = Bad
Trading a knight for two pawns = Bad
Trading two pawns for a knight = Good
Trading a queen for a rook and a bishop = OK
Trading a rook and a bishop for a queen = OK
Trading a queen for a rook and a knight = OK
Trading a rook and a knight for a queen = OK
Trading both rooks for a queen and a knight = Good
Trading a queen and a knight for both rooks = Bad
Trading both rooks for a queen and a bishop = Good
Trading a queen and a bishop for both rooks = Bad
Trading a king for a radioactive monkey = Rubbish
Trading a radioactive monkey for a king = More rubbish

Usually trading the queen for both rooks is also a favourable deal.
However I can't emphasize strongly enough that these are just rules, and all of them has exceptions.

 

[Aphex_Twin]

Evaluation:

Here is a brief description of the heuristics used in the positional
evaluator of the GNU Chess program. Many heuristics are functions of the
stage of the game which is based on the total non-pawn material remaining
for both sides.


PAWNS
The material value of a pawn is 100 points. Isolated pawns get a
penalty depending on which file they occupy:
(12,14,16,20,20,16,14,12) for files (a..h).
Doubled pawns (which are not also isolated) get a penalty of 12
points. Backward pawns (defined simply as not being defended by a
pawn with the square in front also not defended by a a pawn) are
penalized 6 points. A 4 point penalty is also invoked for each attack
by the opponent to a backward pawn and for a backward pawn on a
half-open file. Pawn Advancement in the centre is given a bonus of
about 4 points per rank in the opening increasing to about 8 points
per rank in the ending. Advancement on the edges is given a lower
bonus. Pawns on the e and d files and on the 2nd rank are given a 10
point penalty. An additional penalty of 15 points is invoked if these
pawns are also blocked. Pawns within 2 squares of the king are given
a 10 point bonus. Passed pawns are given a bonus for increasing rank
which is a function of stage of the game and of whether the opponent
blocks or attacks one or more squares in front of the pawn or if the
opponents king is in the square of the pawn. This bonus ranges from
about 15 points for a pawn on the second rank up to about 300 points
for a passed pawn on the 7th rank which can't be stopped from
queening.


KNIGHTS
The material value of a knight is 330 points. The main heuristic for
knights is a bonus for proximity to the centre. This varies from 0
points in the corners to 30 points in the centre. Knights are also
given a bonus for being within 2 squares of each enemy piece. This
bonus is a function of the stage of the game, equalling 4 points in
the end game. A penalty of 1 point per square is given for distance
from either king. A bonus of up to 8 points (depends on stage) is
given for knights which can't be driven away by enemy pawns.


BISHOPS
The material value of a bishop is 330 points. Bishops are given a
bonus as material falls off the board equalling 10 points in the end
game. Bishops get a bonus for mobility and Xray mobility thru pieces
but not pawns. This bonus ranges from -4 points for a totally blocked
bishop up to 18 points for a bishop attacking 12 or more squares.
Xray attacks on an enemy R,Q,K or any undefended piece are given an 8
point bonus. Bishops are given a bonus of 14 points if they lie on
the edge of the board up to 22 points if the lie in the centre. A
bishop is given a bonus of up to 5 points for each attack to a square
adjacent to the enemy king.


ROOKS
The material value of a rook is 520 points. Rook mobility is handled
similiarly to bishops with a bonus of 0 points if blocked up to 20
points if attacking 12 squares or more. A bonus of 8 points for Xray
attacks is handled as it is for bishops. Rooks are given a bonus of
10 points for occupying a file with no friendly pawns and a bonus of
4 points if no enemy pawns lie on that file. After the opening Rooks
are penalized slightly depending on "taxicab" distance to the enemy
king.


QUEENS
The material value of a queen is 980 points. The only heuristic for a
queen is that after the opening it is penalized slightly for
"taxicab" distance to the enemy king.


KINGS
Kings are given a penalty for proximity to the centre in the opening
and a bonus for proximity to the centre in the endgame. The penalty
is about 24 points for being in the centre in the opening with a
bonus of about 36 points for being in the centre in the endgame.
Except when the otherside has only pawns. Then the bonus is turned off.
Kings are penalized for lying on an open or half-open file or if the
adjacent file closest to the corner is open or half-open. This
penalty is up to 23 points in the opening and goes to zero in the end
game. The King is penalized up to 8 points if there are no pawns
immediately adjacent. A penalty is invoked depending on the number of
"safe" checks available by the opponent. This penalty ranges from 6
points for one such check to 50 points for 4 or more. Depending on
game stage, Kings are given up to 10 points for castling and a
penalty of up to 40 points for moving before castling.


SPECIAL
If more than one piece is "hung" (attacked and not defended or
attacked by an enemy piece of lower value) an extra penalty of 10
points is invoked for that side and the search may be extended one
ply. Pinned or trapped pieces are treated similarly. A special mating
routine is used if one side has only a king and the other has mating
material.

This is what the program I was studying a while back worked at the most basic level.

@luceafărul
An input from an experienced chess player would be welcome. Do the evaluations above seem sensible enough?


GnuChess is a freeware, open-source chess engine that can play at a level Elo > 2000 (I haven't tested it on my current hardware yet).

 

[Bluemofia]

It depends on the situation.

I voted for the ones that usually would be benificial.


Always do everything you can to get the king. Even if it means radioactive monkey.

 

[Richard Cribb]

This is what the program I was studying a while back worked at the most basic level.

@luceafărul
An input from an experienced chess player would be welcome. Do the evaluations above seem sensible enough?


GnuChess is a freeware, open-source chess engine that can play at a level Elo > 2000 (I haven't tested it on my current hardware yet).

Yes it seems OK to me, even if as I already mentioned, there are always situation where other factors can change the verdict. I remember for instance somebody, I think it was Chernev or perhaps Clarke, wrote that after having played through some of Petrosian's games one starts to wonder if a rook is necessarily more valuable than a knight or bishop...
I don't know GnuChess personally, but I have heard some good things about it.

 

[josephstalin]

That really depends on situation. Sometimes I give a pawn and a bishop for pawn and win in several turns. Numbered values above are good only for learning chess players.

 

[Aphex_Twin]

Well, there was one trade I had been wondering about for some time.

If all things are equal, what is better to have: a pair of knights or a knight and a bishop?

 

[Tomoyo]

If all things are equal, they are worth the same, I guess, since all things are equal.

 

[bahman]

Obviously it depends on situation. You might even be forced to exhange your queen for a pown. The only relevant and tricky exchanges in an almost balanced situation are 2 rocks for 1 queen and 1 knight for 1 bishop and that again depends on situation.

 

[Theoden]

Seems like I voted as the vast majority did.

I usually value them as
pawn: 1
bishop: 3
knight: 3.5
rook: 5
queen: 9
The knight being higher than the bishop is a result of my play style.

But then again position on the board and current needs makes the values go up and down. And trading a queen for a pawn can win you the game in the right situation.

 

[Sprites]

ususlly prefer knights to bishops but all things are position-dependant

 

[Tank_Guy#3]

I chose "Rook and Bishop for Queen" and "Rook and Knight for Queen". I use Queens way more than I should.

 

[Ribannah]

It's never the trade that matters. What is traded is no longer in the game and can therefore not influence the outcome.
The only thing that matters is what is left after the trade.