bean_Fischer wrote:
Between 0 and 25. Well, they worth nothing if you are checkmated. And they worth 25 if you checkmate your opponent.
funny
Queen + Knight
Sort:
Rook-knight compound aka Chancellor is assessed to around 8.5 points.
The additional bonus 0.5 is possible due to the knight limited range !?
In comparison the Queen is some degree stronger because of diagonal access.
Ideas?
CLINTEASTW00D wrote:
Rook-knight compound aka Chancellor is assessed to around 8.5 points.
Where does that come from ?
If it's from games played by fantasy chess players, I am afraid that, err, their play is not optimal...
On wikipedia I found something about an amazon, which is a knight and queen combined (fantasy chess), there it's stated to be about 1 1/4 times a queen in value (11,25).
http://en.wikipedia.org/wiki/Amazon_%28chess%29
Also, I think fantasy chess has too many variables. I can well imagine play becoming sub-optimal when the game gets too chaotic.
Irontiger wrote:
CLINTEASTW00D wrote:
Rook-knight compound aka Chancellor is assessed to around 8.5 points.
Where does that come from ?
According this sites. It does look realible doesn't it?
http://chessvariants.wikidot.com/assigning-piece-values-with-cast
http://www.symmetryperfect.com/shots/texts/values-capa.pdf
http://www.angelfire.com/ultra/georgecm9000/GothicChess.html
| PIECE | Ed Trice-VALUES | R.Scharnagl-VALUES | My-VALUES |
|---|---|---|---|
| Pawn | 1.0 | 1.0 | 1.0 |
| Knight | 2.5 | 3.05 | 2.5 |
| Bishop | 3.02 | 3.60 | 3.02 |
| Rook | 4.73 | 5.43 | 4.86 |
| Archbishop | 6.22 | 6.65 | 6.52 |
| Chancellor | 8.13 | 8.48 | 8.28 |
| Queen | 8.73 | 9.03 | 8.66 |
| ***King | ??? | 3.72 | 3.55 (An estimation...) |
CLINTEASTW00D wrote:
Irontiger wrote:
Where [do fantasy chess values] come from ?
According this sites. It does look realible doesn't it?
http://chessvariants.wikidot.com/assigning-piece-values-with-cast
http://www.symmetryperfect.com/shots/texts/values-capa.pdf
http://www.angelfire.com/ultra/georgecm9000/GothicChess.html
PIECEEd Trice-VALUESR.Scharnagl-VALUESMy-VALUESPawn 1.0 1.0 1.0 Knight 2.5 3.05 2.5 Bishop 3.02 3.60 3.02 Rook 4.73 5.43 4.86 Archbishop 6.22 6.65 6.52 Chancellor 8.13 8.48 8.28 Queen 8.73 9.03 8.66 ***King ??? 3.72 3.55 (An estimation...)I grew suspicious about the methodology when I saw the folkloric values of knight and bishop. Seriously, more than half a point more ? it would mean that sacrificing a bishop for a knight is generally speaking worse than a knight and a pawn for a bishop - allow me to have some doubts.
And the precise calculation leaves me wondering too.
For instance in http://www.symmetryperfect.com/shots/texts/calc.pdf , in the section "attack values, ideal and practical", I cannot imagine how the six criteria were chosen.
And even if they were justified, six values is too much, it allows interpolation of known results. If I use six independant variables that have different values for each piece, even if not chess related (height of the piece, weight, ratio of volume to volume of smallest sphere in which the piece could fit, etc.), I can always choose a way to average those variables with different weigths that allows to get back on the results pawn = 1, knight = 3, rook = 5, queen = 9, which are the only possible validation of the model.
Actually, it just needs 4 parameters to have the 5 standard trade-off values (how much the piece XXX is worth in pawns) correct with the good choice of weigths. Even only 3 if the king's activity value has not been evaluated correctly because of the inability to be traded.
In a nutshell : even if the real values are approximatively correct for known pieces (where we can be confident), if too many things were put in the model, the values for fantasy pieces may well be unreasonable.
Bishops does indeed get the edge on larger board such as 10x8 and will do so as the board continuously increases.
In 10x8 gothic / capablanca - Two Bishops vs Two Knights can be considered being a rough exchange up in most practical cirumstances.
The approximate correct value for the King is 3.60 although it can be easier to recall 3.5 or 3.55, - based on average values.
CLINTEASTW00D wrote:
Bishops does indeed get the edge on larger board such as 10x8 and will do so as the board continuously increases.
In 10x8 gothic / capablanca - Two Bishops vs Two Knights can be considered being a rough exchange up in most practical cirumstances.
The approximate correct value for the King is 3.60 although it can be easier to recall 3.5 or 3.55, - based on average values.
yup
Who cares about points? Points have no practical value, they are just guideline. The practical part of queen+knight is that you should know that a queen+knight is equal or can be even stronger than a queen+bishop. The bishop dobules the queen, while the knight moves on different squares than the queen.
xxvalakixx wrote:
Who cares about points? Points have no practical value, they are just guideline. The practical part of queen+knight is that you should know that a queen+knight is equal or can be even stronger than a queen+bishop. The bishop doubles the queen, while the knight moves on different squares than the queen.
well, they give you confident for example:
you traded a rook for a pawn and a rook was worth 1 point
you would still play.
it also shows the value of the peice.
Between 0 and 25. Well, they worth nothing if you are checkmated. And they worth 25 if you checkmate your opponent.