Ponder ON or OFF ?

Engines should not allowed to ponder while the other engine is thinking, unless each engine has been allocated half of your cores each.
It's accepted bad practice to run 8CPU v 8CPU ponder on matches on an 8 core computer.
That's just the way it is and there's no point arguing about it.

Graham Banks wrote:

Engines should not allowed to ponder while the other engine is thinking, unless each engine has been allocated half of your cores each.
It's accepted bad practice to run 8CPU v 8CPU ponder on matches on an 8 core computer.
That's just the way it is and there's no point arguing about it.

People can do what ever they want! there are No hard rules about any of it...Which ever way I run a Match it is always fair so that both or all the Engines have the same Conditions it's as Simple as that.


W.C.N

WorldChessNews wrote:

Graham Banks wrote:

Engines should not allowed to ponder while the other engine is thinking, unless each engine has been allocated half of your cores each.
It's accepted bad practice to run 8CPU v 8CPU ponder on matches on an 8 core computer.
That's just the way it is and there's no point arguing about it.

People can do what ever they want! there are No hard rules about any of it...Which ever way I run a Match it is always fair so that both or all the Engines have the same Conditions it's as Simple as that.


W.C.N

Of course you can do whatever you want, but the conditions will not necessarily be equally fair to both engines with ponder on using all cores for both engines, as some engines like to hog more of the cpu than others.
However, what you wish to do is entirely your choice. Been there, done that and know the pitfalls.

If anyone should know about running engine matches it should be Graham. As the man said  

Been there, done that and know the pitfalls.  

Or to put it another way.....it's your time to waste

Auto232 is no longer supported on the latest software releaded by chessbase. It's a shame i have to rollback to Fritz 11 to have this feature installed.

How can i setup games between two computers? null modem cable required?

https://chessprogramming.wikispaces.com/Chrilly+Donninger#Auto232

Last edited by QueensideCastler on Thu Jun 11, 2015 11:54 am; edited 1 time in total

Personally not into running engine matches, (too busy playing my own matches). But I would assume that there is other software available that you could use to run engines
matches. Aquarium or Chess Explorer, not sure about the free one's.

Graham Banks wrote:

WorldChessNews wrote:

Graham Banks wrote:

Engines should not allowed to ponder while the other engine is thinking, unless each engine has been allocated half of your cores each.
It's accepted bad practice to run 8CPU v 8CPU ponder on matches on an 8 core computer.
That's just the way it is and there's no point arguing about it.

People can do what ever they want! there are No hard rules about any of it...Which ever way I run a Match it is always fair so that both or all the Engines have the same Conditions it's as Simple as that.


W.C.N

Of course you can do whatever you want, but the conditions will not necessarily be equally fair to both engines with ponder on using all cores for both engines, as some engines like to hog more of the cpu than others.
However, what you wish to do is entirely your choice. Been there, done that and know the pitfalls.

If other Engines don't take Advantage of how to use Ponder ON like Stockfish obviously does then it's a Problem for those other Engines!


W.C.N

WorldChessNews wrote:

Graham Banks wrote:

WorldChessNews wrote:

Graham Banks wrote:

Engines should not allowed to ponder while the other engine is thinking, unless each engine has been allocated half of your cores each.
It's accepted bad practice to run 8CPU v 8CPU ponder on matches on an 8 core computer.
That's just the way it is and there's no point arguing about it.

People can do what ever they want! there are No hard rules about any of it...Which ever way I run a Match it is always fair so that both or all the Engines have the same Conditions it's as Simple as that.


W.C.N

Of course you can do whatever you want, but the conditions will not necessarily be equally fair to both engines with ponder on using all cores for both engines, as some engines like to hog more of the cpu than others.
However, what you wish to do is entirely your choice. Been there, done that and know the pitfalls.

If other Engines don't take Advantage of how to use Ponder ON like Stockfish obviously does then it's a Problem for those other Engines!


W.C.N

It would be a problem for you as well if you care for fair and balanced engine testing with reliable results. If not, then by all means test however you like.

Wiki wrote:

In chess games between two computers, pondering makes sense only if the competing chess engines use separate processors or cores.

The issue is that you have to split up the relative speed between two engines. Half of the times the program guesses right, and are wrong otherwice.

Pondering only affects "Time Managment" the way i see it.

Chess: Learning from a chess engine

https://www.youtube.com/watch?v=kDz7jyAkaNU

Last edited by WorldChessNews on Sat Jun 13, 2015 5:16 pm; edited 1 time in total

You seem to be getting away from the point of the thread (again).

I assume that which ever engines are in this engine championship, all will be playing with the best settings for each engine.

We all know you have a hang up about Komodo and Chessbase and think they have a plot to rule the engine chess world, but most normal people just accept the fact
of life that Chessbase is (and has been for many years) the main/only company in that field.    Live with it

As for the rules being there to help Komodo win etc. etc. I don't know enough (or care) about the tournament. But as Komodo seems to be going alright in many
rating lists etc, I cannot see that it would make very little difference if it wins this 'Championship' or not. At the end of the day they are really only 3 engines
(at this time) that most people would get. Stockfish, Komodo and Houdini, Two of which you have to pay for, however you are able to purchase them without
using Chessbase if you wish.

Now the question was/is Ponder On/Off.  people have given their views about this subject...I do not myself run engine matches. I have more than enough
to do playing my own matches/tournaments etc. But I do like to see what peoples opinions are......sometimes.

QueensideCastler wrote:

Wiki wrote:

In chess games between two computers, pondering makes sense only if the competing chess engines use separate processors or cores.

The issue is that you have to split up the relative speed between two engines. Half of the times the program guesses right, and are wrong otherwice.

Pondering only affects "Time Managment" the way i see it.

The Point you are Missing is that the Very Best Engines Guess the Other Moves 60% of the time! like Stockfish :) So expect the Next World Computer Chess Championship in the Netherlands 2015 to be with Ponder OFF! :( and I Wonder Why!

It may just be that Someone or Some Company Cough! Cough! (ChessBase.com) Prefer it that way to try and ensure a Komodo Win! :( $$$$$

Take a look below before I Picked up on the Possible Chess SCAM! Concerning the World Computer Chess Championship in the Netherlands 2015 where they put Komodo at the Top of the Participants! ? Why ? it's Not even Alphabetic-le or in any order that I can Fathom! Other than they Know! :(

http://www.thejoyfulprogrammer.com/qb64/forum/attachment.php?aid=567

And then look at the most Recent Version of the World Computer Chess Championship in the Netherlands 2015 also below and See how they have Deviously Changed the Order! :)

http://www.thejoyfulprogrammer.com/qb64/forum/attachment.php?aid=568

W.C.N

Training in futility as Nathanael Russell said: If you share same core you have to sacrifice speed & depth for 50-60% chance of hitting predicted moves.

Nathanael Russell wrote:

It is still an exercise in futility. Let's say the move prediction is 50%. Both of your engines would still be reaching the correct response line half of the time in the time control and in the other half will have wasted considerable time finding the wrong move for nothing to gain from it at only half of the potential depth. Your engines might agree on the same predicted reply at 15ply for example but may change course once 30 ply is reached. An exercise in futility.

QueensideCastler wrote:

Training in futility as Nathanael Russell said: If you share same core you have to sacrifice speed & depth for 50-60% chance of hitting predicted moves.

Nathanael Russell wrote:

It is still an exercise in futility. Let's say the move prediction is 50%. Both of your engines would still be reaching the correct response line half of the time in the time control and in the other half will have wasted considerable time finding the wrong move for nothing to gain from it at only half of the potential depth. Your engines might agree on the same predicted reply at 15ply for example but may change course once 30 ply is reached. An exercise in futility.

It's Only Futile if you are not using Stockfish :) because Stockfish knows how to "Hog The CPU" (Famous Quote by Nathanael Russell :) ) and uses Ponder ON to the MAX!
Imagine Stockfish has found a Deep Winning Tactical sequence of moves because it Searches the Deepest! with Ponder ON and if Ponder was OFF :( it would never have found the move :( and may even lose the Game which is what seems to happen in those Type of Stockfish v Komodo or Komodo v Stockfish Games.


W.C.N

WorldChessNews wrote:

QueensideCastler wrote:

Training in futility as Nathanael Russell said: If you share same core you have to sacrifice speed & depth for 50-60% chance of hitting predicted moves.

Nathanael Russell wrote:

It is still an exercise in futility. Let's say the move prediction is 50%. Both of your engines would still be reaching the correct response line half of the time in the time control and in the other half will have wasted considerable time finding the wrong move for nothing to gain from it at only half of the potential depth. Your engines might agree on the same predicted reply at 15ply for example but may change course once 30 ply is reached. An exercise in futility.

It's Only Futile if you are not using Stockfish :) because Stockfish knows how to "Hog The CPU" (Famous Quote by Nathanael Russell :) ) and uses Ponder ON to the MAX!
Imagine Stockfish has found a Deep Winning Tactical sequence of moves because it Searches the Deepest! with Ponder ON and if Ponder was OFF :( it would never have found the move :( and may even lose the Game which is what seems to happen in those Type of Stockfish v Komodo or Komodo v Stockfish Games.


W.C.N

I really want to try pondering to it fullest. Too bad that today's GUI don't suppoert auto232 anymore.

Seems to me the only way to be fair to each engine is to use two computers, one engine playing on each. No problem then

stagehand wrote:

Seems to me the only way to be fair to each engine is to use two computers, one engine playing on each. No problem then

You mean i have to transfear the moves back and forth throughout the whole game? if so it's very tiresome i can predict.

stagehand wrote:

Seems to me the only way to be fair to each engine is to use two computers, one engine playing on each. No problem then

On an octal, you can fairly play 4CPU v 4CPU ponder on.

stagehand wrote:

On an octal, you can fairly play 4CPU v 4CPU ponder on.

I agree for low-time controls (blitz) such as 10-15 minutes or below. If they share same core, i don't recommend long-time controls.

The general consensus indicates that PB is a direct downgrade as you sacrifice knp/s & potential depth for indirectly increase Time Management.

QueensideCastler wrote:

stagehand wrote:

On an octal, you can fairly play 4CPU v 4CPU ponder on.

I agree for low-time controls (blitz) such as 10-15 minutes or below. If they share same core, i don't recommend long-time controls.

The general consensus indicates that PB is a direct downgrade as you sacrifice knp/s & potential depth for indirectly increase Time Management.

I should add that each engine should access tablebases from a different drive in order to avoid clashes on the hard drive.

Quote: From an internet article! about making Two Computers into one :)

"I own a computer store. One day, two policemen came into the store and told that they owned a 486 and a 286. They asked if a 486 and a 286 could be assembled together into a 686. I replied to the dumb request by asking them if two 200 horsepower police cars can be used to make up a 400 horsepower Ferrari. The policemen didn't get it and replied angrily that altering car engines is strictly forbidden by law."

Back in the real World...

Joining two computers as one

http://www.techspot.com/community/topics/joining-two-computers-as-one.1670


W.C.N

I do understand the point of people/software running tournaments etc to test the different engines, and finding ways to improve them.
There are also a few people that know what they are doing that run tournaments and this helps to see how well engines perform/play.

They are many testers etc. working on Stockfish all the time and that is why it has become so strong over the past few years.  I also understand that there have been
some improvements coming along with Komodo which will make it a great deal stronger than now,  even though many consider it the best already.
But for the standard chess player I am at a loss to understand the enjoyment to had watching two machines/engines playing the game.  

As I have already said, I do not run engine matches myself.  To be frank to me it all seems rather a waste of my time and energy to do it. I am more interested in playing against
other people etc, rather than watching two engines playing the game, I just cannot see much fulfilment in that.

Of course in this day and age even if you are playing chess online most of the time engines are involved anyway....but the best players do still know how to use their brains

stagehand wrote:

I do understand the point of people/software running tournaments etc to test the different engines, and finding ways to improve them.
There are also a few people that know what they are doing that run tournaments and this helps to see how well engines perform/play.

They are many testers etc. working on Stockfish all the time and that is why it has become so strong over the past few years.  I also understand that there have been
some improvements coming along with Komodo which will make it a great deal stronger than now,  even though many consider it the best already.
But for the standard chess player I am at a loss to understand the enjoyment to had watching two machines/engines playing the game.  

As I have already said, I do not run engine matches myself.  To be frank to me it all seems rather a waste of my time and energy to do it. I am more interested in playing against
other people etc, rather than watching two engines playing the game, I just cannot see much fulfilment in that.

Of course in this day and age even if you are playing chess online most of the time engines are involved anyway....but the best players do still know how to use their brains

More than likely because you just Don't Know anything about Chess Full Stop!

W.C.N

I will give you a game my little friend ANYTIME you want. But be warned...I do not make the same mistakes as your Stockfish

stagehand wrote:

I will give you a game my little friend ANYTIME you want. But be warned...I do not make the same mistakes as your Stockfish

You could Not Play your way out of a Paper Bag!