one question about knowledge chess engine match one ply ?

[Karmazen & Oliver]

one question about knowledge chess engine match one ply ?

I was thinking of making a thematic party on knowledge, using fixed depth of a ply, another with 3 plys ...

I ask:

using a stable depth would Be annulled the advantages of code optimization, code parts that anything has to do with the knowledge, but alone with the gross force and certain dose of luck when selecting many branches...

if we use those low levels of I calculate of plys, we would not find the engines that have a knowledge bigger than the positions and they choose the road but or less correct, or rather, they choose a better road that their rival, although this it is incorrect, depending on the given position.

I will prove it. go to 1 ply. if this is error conceptual i try 3, 5, 7, 9. no more.

no books. too.

It also won't prove a thing. Some programs get way more out of 9 plies than others. For example, most current programs will do 9 plies instantly, while a program like ChessMaster (the king) might take minutes (or sometimes hours) to get that deep. plies are not uniform. Some programs extend much more aggressively than other s and equal plies will see those taking a lot more time. Other programs reduce more aggressively and equal plies will see those programs take far less time. you end up with results that are meaningless...

OK you are correct. I will prove it. go to 1 ply. and I have the results... ( I don´t need more plys...)

( ok. maybe don´t exactly 1, 2 or 3. but always the same horizontl knowledge )

also. for that reason we force to the engine to think single 1 ply. so that it is not affected neither manipulated their evaluation.

the answer of 1 ply should be instantaneous. and to require to evaluate few movements. 50~500 moves, if the engine calculates more that that. then this making traps. ; -)

[Uri Blass]

one question about knowledge chess engine match one ply ?

I was thinking of making a thematic party on knowledge, using fixed depth of a ply, another with 3 plys and others with 5 plys. perhaps another with 7 plys. no more that 9 or 11.

I ask:

using a stable depth would Be annulled the advantages of code optimization, code parts that anything has to do with the knowledge, but alone with the gross force and certain dose of luck when selecting many branches...

if we use those low levels of I calculate of plys, we would not find the engines that have a knowledge bigger than the positions and they choose the road but or less correct, or rather, they choose a better road that their rival, although this it is incorrect, depending on the given position.

I will prove it. go to 1 ply. if this is error conceptual i try 3, 5, 7, 9. no more.

no books. too.

It also won't prove a thing. Some programs get way more out of 9 plies than others. For example, most current programs will do 9 plies instantly, while a program like ChessMaster (the king) might take minutes (or sometimes hours) to get that deep. plies are not uniform. Some programs extend much more aggressively than other s and equal plies will see those taking a lot more time. Other programs reduce more aggressively and equal plies will see those programs take far less time. you end up with results that are meaningless...

I can add that even in case that fixed depth is no extensions and no reductions you cannot test evaluiation by that way.

For example in that case all the leaves of the tree are going to have the same side to move so bonus for the side to move is not going to change things when practically bonus for the side to move can change things.

The only possible test is test of evaluation relative to specific tree when all the engines search exactly the same tree and even in this case the question which evaluation is better may be dependent on the type of the tree that is searched.

Uri

[Karmazen & Oliver]

( ok. maybe don´t exactly 1, 2 or 3. but always the same horizontl knowledge if we use the same engine...)

rybka human versus rybka dinamic play with the same ply numer, when you use 1 ply.

and the games are repeat always... only do two diferents games.

now I go see that positions... OK. your example don´t is understand for me.

[Uri Blass]

one question about knowledge chess engine match one ply ?

I was thinking of making a thematic party on knowledge, using fixed depth of a ply, another with 3 plys ...

I ask:

using a stable depth would Be annulled the advantages of code optimization, code parts that anything has to do with the knowledge, but alone with the gross force and certain dose of luck when selecting many branches...

if we use those low levels of I calculate of plys, we would not find the engines that have a knowledge bigger than the positions and they choose the road but or less correct, or rather, they choose a better road that their rival, although this it is incorrect, depending on the given position.

I will prove it. go to 1 ply. if this is error conceptual i try 3, 5, 7, 9. no more.

no books. too.

It also won't prove a thing. Some programs get way more out of 9 plies than others. For example, most current programs will do 9 plies instantly, while a program like ChessMaster (the king) might take minutes (or sometimes hours) to get that deep. plies are not uniform. Some programs extend much more aggressively than other s and equal plies will see those taking a lot more time. Other programs reduce more aggressively and equal plies will see those programs take far less time. you end up with results that are meaningless...

OK you are correct. I will prove it. go to 1 ply. and I have the results... ( I don´t need more plys...)

( ok. maybe don´t exactly 1, 2 or 3. but always the same horizontl knowledge )

also. for that reason we force to the engine to think single 1 ply. so that it is not affected neither manipulated their evaluation.

the answer of 1 ply should be instantaneous. and to require to evaluate few movements. 50~500 moves, if the engine calculates more that that. then this making traps. ; -)

Even if the engine does not calculate more than it there is a big difference between 50 moves and 500 moves and even if one engine searches 100 moves and the second engine 400 moves then you can expect a difference of more than 100 elo not because of evaluation.

Uri

[Karmazen & Oliver]

Even if the engine does not calculate more than it there is a big difference between 50 moves and 500 moves and even if one engine searches 100 moves and the second engine 400 moves then you can expect a difference of more than 100 elo not because of evaluation.

Uri

I referred to 1 ply of depth. depending on the positions the engine cannot calculate but of aprox 500 positions for that then would be entering in a branch... extra. that is traps if we are in 1 ply.

I referred that the evaluation should be among 50 ~ 500 positions. for that the ALONE engine should evaluate 1 answer for each movement.

not to explain to you me... it is not the same thing that you understood.

[Karmazen & Oliver]

The only possible test is test of evaluation relative to specific tree when all the engines search exactly the same tree and even in this case the question which evaluation is better may be dependent on the type of the tree that is searched.

Uri

Don´t exits any tree ¡¡ one 1 ply. and if exists a small line. always is the same line.

for that reason the engines always plays the same games, the only one that seems to use a longer branch is chesstiger, and win versus rybka, and when ct14 play versus ct15... ( same tree ¡?) then +1 0= -1. (1-1).

and rybka versus rybka using the same tree ¡... and win rybka human.

then the best selection of first plays moves selections out it rybka human, better than r-dinamic and better than r-cpu. therefore this he/she has more knowledge chess ¡

for that reason then it wins for that in the following level, ( ply 2, after this ply 3 or 7 ... ) , he/she repeats the best Selection in the first branches...

the problem this in the principle...

he/she always needed an evaluation of the position and in that it bases the following search ....

alone it is necessary testear the first seleciones in many different positions to improve the algorithm...

he/she has to be easy for the programmers

[GenoM]

Didn't the fact that Rybka searches +2 plies more than it shows in some way biased the results your test?
Some time ago I made similar test-tourneys based on a ply-depth (and without books, of course). I would say it makes sense if you test such way only different versions of the same engine
Regards,
Geno

[Karmazen & Oliver]

Didn't the fact that Rybka searches +2 plies more than it shows in some way biased the results your test?

OK. that is irrelevant. the fact that rybka calculates +1 or +2 it is not important, the important thing is that it doesn't reach more due depth to code optimizations or due to the power of the cpu.

the one numbers it is not the important thing, the interesting thing it is that it always calculates to the same depth, you can calculate which is this, only analyzing different positions with victory force #... in 1, kill in 2 and kill in 3...

this way he/she can know when rybka him this deceiving... in the depth. but I believe that all the engines has behaving well in ARENA GUI

does it always calculate the same lines, if he/she/it see the same position and have you thought that rybka to lost against chess tiger 14 and tied with chess tiger 15 ?

it is quite curious that loses against these two engines that if they seem to be using a minimum depth of 3 plys.

also, surprisingly in function of the intelligence of the other engines (f6... f9 and deep versions f7) rybka needs of more plays to finish winning.

interesting changes are also observed in the way that rybka plays against rybka, single 6 different games for the three rybka versions, dinamic, human, cpu... and r-human is the more sure, although it takes him more time-moves to win and r-dinamic is shown very aggressive with white, that is to say with the initiative. and have some victory very fast... ¡ 17-21-22 moves only ¡

Some time ago I made similar test-tourneys based on a ply-depth (and without books, of course). I would say it makes sense if you test such way only different versions of the same engine
Regards,
Geno

their experiment if it can be polluted for that depending on the engine, some consider that 4 plys has already advanced, and others consider that they have arrived to 5 plys...

evidently an advantage exists when carrying out that test. and that test will have a worse result when more plys of depth selects.

I try that that type of errors doesn't have relevance, when indicating that ALONE it calculates 1 ply, we guarantee ourselves that it calculates very little and not so much decisive relevance in the final experiment. ( 1~3 plys )

anyway it is observed at first sight when an engine this making traps, 1 ply, he/she means that the one numbers of analyzed positions he/she has to be very reduced, smaller than 1.000, ( this not is n/s... only numbre of positions) , in another case the engine would be deepening more in the search.

I also think that the KEY this in these first evaluations that makes the program for that like it is a recursive function, this it is used again in the following level, and another time in the following level.

any error of valuation in ply 1 will be dragged along the whole analysis chain when the engine thinks to full power.

I believe that many programmers should pass position test of ALONE a 1 ply evaluation, and to compare them with the rybka results... this way they can improve the programming algorithm for not wasting time of I calculate...

the most important thing is the first idea ply...

[bob]

one question about knowledge chess engine match one ply ?

I was thinking of making a thematic party on knowledge, using fixed depth of a ply, another with 3 plys ...

I ask:

using a stable depth would Be annulled the advantages of code optimization, code parts that anything has to do with the knowledge, but alone with the gross force and certain dose of luck when selecting many branches...

if we use those low levels of I calculate of plys, we would not find the engines that have a knowledge bigger than the positions and they choose the road but or less correct, or rather, they choose a better road that their rival, although this it is incorrect, depending on the given position.

I will prove it. go to 1 ply. if this is error conceptual i try 3, 5, 7, 9. no more.

no books. too.

It also won't prove a thing. Some programs get way more out of 9 plies than others. For example, most current programs will do 9 plies instantly, while a program like ChessMaster (the king) might take minutes (or sometimes hours) to get that deep. plies are not uniform. Some programs extend much more aggressively than other s and equal plies will see those taking a lot more time. Other programs reduce more aggressively and equal plies will see those programs take far less time. you end up with results that are meaningless...

OK you are correct. I will prove it. go to 1 ply. and I have the results... ( I don´t need more plys...)

( ok. maybe don´t exactly 1, 2 or 3. but always the same horizontl knowledge )

also. for that reason we force to the engine to think single 1 ply. so that it is not affected neither manipulated their evaluation.

the answer of 1 ply should be instantaneous. and to require to evaluate few movements. 50~500 moves, if the engine calculates more that that. then this making traps. ; -)

ChessMaster won't be "instantaneous" on one-ply searches. That's the problem. A program that does checks in the q-search will see more with a 1 ply search than one that does not. A program that has a qsearch will see more with a 1 ply search than one that will not.

It's been known for years that you can't compare programs using "equal plies", because plies are not the same from one program to another.

[Karmazen & Oliver]

ChessMaster won't be "instantaneous" on one-ply searches. That's the problem. A program that does checks in the q-search will see more with a 1 ply search than one that does not. A program that has a qsearch will see more with a 1 ply search than one that will not.

It's been known for years that you can't compare programs using "equal plies", because plies are not the same from one program to another.

ok. yes. chesmaster is bad example. this is a bug error program.

but.

rybka human, and rybka cpu, and rybka dinamic, have the same ply searcj.

and fritz 6,7,8,9 deep =

and chesstiger 14 and chess tiger 15 have indentical reference plys.

arasan OK

hiarcs OK

etc..

I see during the game, and a lot of engines are correct. chesmaster no. ok. don´t problem. I don´t test with chessmaster

but the conclusions evaluation are interesting.