No Bias Here.

[mwyoung]

Houdini 2.0 has been out for months, and CCRL has not had time to play enough games to rate Houdini 2.0 on their offical 40/40 rating list. They only had time to play 155 games of the 200 needed. With 155 games Houdini 2.0 is rated #1 on the complete 40/40 rating list with a rating of 3320.

Houdini 2.0 64 bit 4cpu - 155 games: 62 wins, 27 losses, 66 draws (42.6%), score: 61.3% Rating 3320


Amazingly....
Komodo 4 64-bit SSE has been public for a few days and CCRL has played 408 games at 40/40

408 games: 171 wins, 67 losses, 170 draws (41.7%), score: 62.7% rating 3211.

http://computerchess.org.uk/ccrl/4040.l ... t_all.html

Here are some newly posted results from CCRL Dated December 29, 2011

CCRL Complete 40/40 rating list rating results with number of games played.



Name Rating Games played
Houdini 2.0 64-bit 4CPU 3320 155

Houdini 1.5a 64-bit 4CPU 3301 695

Houdini 2.0 64-bit 1CPU 3243 1253

Komodo 3 64-bit SSE 3217 604

Komodo 4 64-bit SSE 3211 408


CCRL rating results show that Komodo 4 was not able to catch Houdini 2.0 at 40/40 time controls, and it shows it behind Komodo 3.

[Graham Banks]

Houdini 2.0 has been out for months, and CCRL has not had time to play enough games to rate Houdini 2.0 on their offical 40/40 rating list. They only had time to play 155 games of the 200 needed. With 155 games Houdini 2.0 is rated #1 on the complete 40/40 rating list with a rating of 3320.

Houdini 2.0 64 bit 4cpu - 155 games: 62 wins, 27 losses, 66 draws (42.6%), score: 61.3% Rating 3320


Amazingly....
Komodo 4 64-bit SSE has been public for a few days and CCRL has played 408 games at 40/40

408 games: 171 wins, 67 losses, 170 draws (41.7%), score: 62.7% rating 3211.

http://computerchess.org.uk/ccrl/4040.l ... t_all.html

Here are some newly posted results from CCRL Dated December 29, 2011

CCRL Complete 40/40 rating list rating results with number of games played.



Name Rating Games played
Houdini 2.0 64-bit 4CPU 3320 155

Houdini 1.5a 64-bit 4CPU 3301 695

Houdini 2.0 64-bit 1CPU 3243 1253

Komodo 3 64-bit SSE 3217 604

Komodo 4 64-bit SSE 3211 408


CCRL rating results show that Komodo 4 was not able to catch Houdini 2.0 at 40/40 time controls, and it shows it behind Komodo 3.

It's a lack of testers carrying out 40/40 testing of 4CPU engines. It will be addressed as soon as possible.

[mwyoung]

Houdini 2.0 has been out for months, and CCRL has not had time to play enough games to rate Houdini 2.0 on their offical 40/40 rating list. They only had time to play 155 games of the 200 needed. With 155 games Houdini 2.0 is rated #1 on the complete 40/40 rating list with a rating of 3320.

Houdini 2.0 64 bit 4cpu - 155 games: 62 wins, 27 losses, 66 draws (42.6%), score: 61.3% Rating 3320


Amazingly....
Komodo 4 64-bit SSE has been public for a few days and CCRL has played 408 games at 40/40

408 games: 171 wins, 67 losses, 170 draws (41.7%), score: 62.7% rating 3211.

http://computerchess.org.uk/ccrl/4040.l ... t_all.html

Here are some newly posted results from CCRL Dated December 29, 2011

CCRL Complete 40/40 rating list rating results with number of games played.



Name Rating Games played
Houdini 2.0 64-bit 4CPU 3320 155

Houdini 1.5a 64-bit 4CPU 3301 695

Houdini 2.0 64-bit 1CPU 3243 1253

Komodo 3 64-bit SSE 3217 604

Komodo 4 64-bit SSE 3211 408


CCRL rating results show that Komodo 4 was not able to catch Houdini 2.0 at 40/40 time controls, and it shows it behind Komodo 3.

It's a lack of testers carrying out 40/40 testing of 4CPU engines. It will be addressed eventually.

And that is not a denial. We know it is a lack of testers testing Houdini at 4cpu...

I wonder what program will be listed first Komodo 4 4cpu or Houdini 2.0 4cpu?

Larry Kaufman Quote "CCRL does 40/40 ratings, and usually a new engine will have a well-established rating on their list within a week or two."

[Graham Banks]

I wonder what program will be listed first Komodo 4 4cpu or Houdini 2.0 4cpu?

Larry Kaufman Quote "CCRL does 40/40 ratings, and usually a new engine will have a well-established rating on their list within a week or two."

Well Mark - it would nice to be that efficient, but it's not always possible unfortunately.

Which engines are tested and in what order is usually at the discretion of those doing the testing. I think that we currently have two testers who carry out 40/40 4CPU testing and both have had other priorities in recent times.

We do our best. Meantime there are many other rating lists that you can look at to get an overall picture of how things stack up.

To see how well we've tested Houdini 2.0, try looking at the various sub-lists too. Houdini 2.0 64-bit 1CPU has over 1000 games for example and Houdini 2.0 32-bit 1CPU has over 600 games.

[Graham Banks]

Code: Select all

CCRL 40/40 Rating List - Single-CPU engines
369869 games played by 1109 programs, run by 18 testers
Ponder off, General books (up to 12 moves), 3-4-5 piece EGTB
Time control: Equivalent to 40 moves in 40 minutes on Athlon 64 X2 4600+ (2.4 GHz)
Computed on December 29, 2011 with Bayeselo based on 369'869 games
Tested by CCRL team, 2005-2011, http://computerchess.org.uk/ccrl/4040/

   Rank                Engine                 Elo   +    -   Score  AvOp  Games
      1 Houdini 1.5a 64-bit                  3256  +18  -18  64.9%  -99.7  1045
        Houdini 2.0 64-bit                   3243  +17  -17  67.8% -125.1  1253
        Houdini 2.0 32-bit                   3219  +24  -23  76.3% -183.8   674
      2 Komodo 3 64-bit SSE                  3217  +23  -22  54.5%  -29.4   604
        Houdini 2.0c 32-bit                  3215  +25  -24  75.7% -179.1   640
      3 Rybka 4.1 64-bit                     3214  +17  -17  58.1%  -52.2  1133
        Komodo 4 64-bit SSE                  3211  +28  -28  62.7%  -86.4   408
        Komodo 3 64-bit                      3210  +25  -24  62.1%  -81.2   542
      4 Critter 1.2 64-bit                   3205  +15  -15  52.8%  -18.7  1404
        Houdini 1.5a 32-bit                  3203  +22  -21  74.6% -167.7   788

[tmokonen]

I don't have an agenda. I will say again, WHO the HELL gives a F***?? huh?

I happen to care if the paid version of a program is weaker than the free version. Keep your childish temper tantrums to yourself. They are wholly unimpressive.

[Adam Hair]

Houdini 2.0 has been out for months, and CCRL has not had time to play enough games to rate Houdini 2.0 on their offical 40/40 rating list. They only had time to play 155 games of the 200 needed. With 155 games Houdini 2.0 is rated #1 on the complete 40/40 rating list with a rating of 3320.

Houdini 2.0 64 bit 4cpu - 155 games: 62 wins, 27 losses, 66 draws (42.6%), score: 61.3% Rating 3320


Amazingly....
Komodo 4 64-bit SSE has been public for a few days and CCRL has played 408 games at 40/40

408 games: 171 wins, 67 losses, 170 draws (41.7%), score: 62.7% rating 3211.

http://computerchess.org.uk/ccrl/4040.l ... t_all.html

Here are some newly posted results from CCRL Dated December 29, 2011

CCRL Complete 40/40 rating list rating results with number of games played.



Name Rating Games played
Houdini 2.0 64-bit 4CPU 3320 155

Houdini 1.5a 64-bit 4CPU 3301 695

Houdini 2.0 64-bit 1CPU 3243 1253

Komodo 3 64-bit SSE 3217 604

Komodo 4 64-bit SSE 3211 408


CCRL rating results show that Komodo 4 was not able to catch Houdini 2.0 at 40/40 time controls, and it shows it behind Komodo 3.

The only reason I can see for such a post is that you found it easier to voice a supposition than to ask a question.

Let me ask you a question. Do you notice the number of games for Houdini 2.0 1CPU and for Komodo 3 (add 542 games for Komodo 3 non SSE)?
Over 100 games more for Houdini. And 40/40 testing for Houdini 2.0 started 1 to 2 weeks after it started for Komodo 3. Does not go along with your theory of bias against Houdini. What does fit to explain the situation is as Graham says. We lack manpower for testing 4CPU at 40/40.

In contrast, check this out:

CCRL 40/4 Index

Code: Select all

 Rank                 Engine                   Elo   +    -   Score  AvOp  Games
      1 Houdini 2.0c 64-bit 6CPU                3405  +12  -12  69.2% -137.4  2359
      2 Critter 1.2 64-bit 6CPU                 3338  +17  -17  67.9% -125.9  1167
      3 Rybka 4.1 64-bit 6CPU                   3337  +13  -12  57.9%  -56.1  2151
      4 Stockfish 2.1.1 64-bit 6CPU             3330  +12  -12  56.6%  -48.0  2156
      5 Komodo 4 64-bit                         3231  +23  -22  75.1% -196.4   845

Here, we have people who can test multi-cpu versions. Does it look like Houdini 2.0 is being ignored?

[mwyoung]

I wonder what program will be listed first Komodo 4 4cpu or Houdini 2.0 4cpu?

Larry Kaufman Quote "CCRL does 40/40 ratings, and usually a new engine will have a well-established rating on their list within a week or two."

Well Mark - it would nice to be that efficient, but it's not always possible unfortunately.

Which engines are tested and in what order is usually at the discretion of those doing the testing. I think that we currently have two testers who carry out 40/40 4CPU testing and both have had other priorities in recent times.

We do our best. Meantime there are many other rating lists that you can look at to get an overall picture of how things stack up.

Larry is dead on accurate in his quote when it comes to Komodo 4. You are very efficent I see in testing what you want to test..... 408 games in a few days. You are very efficient.

I have no doubt you have other priorities other then testing Houdini 2.0 at 4cpu. Since it will most likely top your 40/40 rating list.

CCRL dislikes Houdini, and I understand your computer chess rating site does have other higher priorities then testing the strongest engine on the planet, but an unbias one would not.

[Graham Banks]

I wonder what program will be listed first Komodo 4 4cpu or Houdini 2.0 4cpu?

Larry Kaufman Quote "CCRL does 40/40 ratings, and usually a new engine will have a well-established rating on their list within a week or two."

Well Mark - it would nice to be that efficient, but it's not always possible unfortunately.

Which engines are tested and in what order is usually at the discretion of those doing the testing. I think that we currently have two testers who carry out 40/40 4CPU testing and both have had other priorities in recent times.

We do our best. Meantime there are many other rating lists that you can look at to get an overall picture of how things stack up.

Larry is dead on accurate in his quote when it comes to Komodo 4. You are very efficent I see in testing what you want to test..... 408 games in a few days. You are very efficient.

I have no doubt you have other priorities other then testing Houdini 2.0 at 4cpu. Since it will most likely top your 40/40 rating list.

CCRL dislikes Houdini, and I understand your computer chess rating site does have other higher priorities then testing the strongest engine on the planet, but an unbias one would not.

I think that most members will be able to see post for what it really is - an attack on CCRL just for the sake of it.

To see how well we've tested Houdini, just take a look at the information provided by both Adam and myself.

There is no perfect rating list and that's why we've always advocated that one should look at them all to get an overall picture.

[mwyoung]

I wonder what program will be listed first Komodo 4 4cpu or Houdini 2.0 4cpu?

Larry Kaufman Quote "CCRL does 40/40 ratings, and usually a new engine will have a well-established rating on their list within a week or two."

Well Mark - it would nice to be that efficient, but it's not always possible unfortunately.

Which engines are tested and in what order is usually at the discretion of those doing the testing. I think that we currently have two testers who carry out 40/40 4CPU testing and both have had other priorities in recent times.

We do our best. Meantime there are many other rating lists that you can look at to get an overall picture of how things stack up.

Larry is dead on accurate in his quote when it comes to Komodo 4. You are very efficent I see in testing what you want to test..... 408 games in a few days. You are very efficient.

I have no doubt you have other priorities other then testing Houdini 2.0 at 4cpu. Since it will most likely top your 40/40 rating list.

CCRL dislikes Houdini, and I understand your computer chess rating site does have other higher priorities then testing the strongest engine on the planet, but an unbias one would not.

I think that most members will be able to see post for what it really is - an attack on CCRL just for the sake of it.

To see how well we've tested Houdini, just take a look at the information provided by both Adam and myself.

There is no perfect rating list and that's why we've always advocated that one should look at them all to get an overall picture.

Graham,

How about we let all the CCC members see how fast the CCRL testers can fall all over themselves when it comes to testing Komodo 4 MP when released.

It will be a High priority....

Anyone want to bet it will not take months to test Komodo 4 MP with 4CPU at 40/40