Which IvanHoe Engine Is Strongest

[De Vos W]

I have seen many discussions and rating with other versions of IvanHoe that are supposedly not official releases. I just wanted to know which IvanHoe is strongest and where can I download the newest Ivanhoes. After many games at 40/1 time controls it seems that IvanHoe 999957a is strongest. These results are based on very fast games though. My rating list is as follows.

Code: Select all

    Program                          Elo    +   -   Games   Score   Av.Op.  Draws

  1 Houdini 1.03a                  : 3242   23  23   440    60.1 %   3170   50.7 %
  2 IvanHoe 999957a                : 3223   25  25   380    57.5 %   3170   49.2 %
  3 Houdini 1.02                   : 3212   23  23   440    55.9 %   3170   49.5 %
  4 IvanHoe 999963                 : 3197   18  18   680    54.7 %   3164   53.5 %
  5 IvanHoe 999952a                : 3195   23  23   400    53.4 %   3171   55.8 %
  6 RobboLito 0085g3               : 3191   19  19   560    51.4 %   3181   56.1 %
  7 Fire 1.31 Warrior              : 3184   26  26   360    50.0 %   3184   48.9 %
  8 Stockfish 1.9                  : 3170   28  28   360    46.8 %   3192   38.6 %
  9 Fire 1.31                      : 3166   21  21   480    47.8 %   3181   53.5 %
 10 Stockfish 1.8                  : 3157   27  28   340    44.9 %   3193   45.0 %
 11 Stockfish 1.7.1 ZugDet         : 3145   21  21   600    46.0 %   3172   44.3 %
 12 Stockfish 1.7.1                : 3124   22  22   600    43.1 %   3172   38.5 %
 13 Rybka 3.0                      : 3123   18  18   840    44.3 %   3162   43.2 %

Thanks to anybody who can help me find the other IvanHoe Engines. Also, does anybody know if there are any vast differences in strength relative to other engines at longer time controls.

Thanks.........................................................................

Take a look at this forum: http://ahmed-chess.forumotion.com/forum.htm

[David Schumaker]

I have seen many discussions and rating with other versions of IvanHoe that are supposedly not official releases. I just wanted to know which IvanHoe is strongest and where can I download the newest Ivanhoes. After many games at 40/1 time controls it seems that IvanHoe 999957a is strongest. These results are based on very fast games though. My rating list is as follows.

Code: Select all

    Program                          Elo    +   -   Games   Score   Av.Op.  Draws

  1 Houdini 1.03a                  : 3242   23  23   440    60.1 %   3170   50.7 %
  2 IvanHoe 999957a                : 3223   25  25   380    57.5 %   3170   49.2 %
  3 Houdini 1.02                   : 3212   23  23   440    55.9 %   3170   49.5 %
  4 IvanHoe 999963                 : 3197   18  18   680    54.7 %   3164   53.5 %
  5 IvanHoe 999952a                : 3195   23  23   400    53.4 %   3171   55.8 %
  6 RobboLito 0085g3               : 3191   19  19   560    51.4 %   3181   56.1 %
  7 Fire 1.31 Warrior              : 3184   26  26   360    50.0 %   3184   48.9 %
  8 Stockfish 1.9                  : 3170   28  28   360    46.8 %   3192   38.6 %
  9 Fire 1.31                      : 3166   21  21   480    47.8 %   3181   53.5 %
 10 Stockfish 1.8                  : 3157   27  28   340    44.9 %   3193   45.0 %
 11 Stockfish 1.7.1 ZugDet         : 3145   21  21   600    46.0 %   3172   44.3 %
 12 Stockfish 1.7.1                : 3124   22  22   600    43.1 %   3172   38.5 %
 13 Rybka 3.0                      : 3123   18  18   840    44.3 %   3162   43.2 %

Thanks to anybody who can help me find the other IvanHoe Engines. Also, does anybody know if there are any vast differences in strength relative to other engines at longer time controls.

Thanks.........................................................................

Take a look at this forum: http://ahmed-chess.forumotion.com/forum.htm

Hi De Vos Walter,

Thank you for all of the information. I don't think that I ever knew about that chess forum. Seems like there's lots of good stuff there like several opening books and such. I will start downloading some of the new IvanHoes and start testing.

[ernest]

I got Houdini + 20 over the strongest Ivanhoe with over 2000 games played.
R4 and Ivanhoe are very close.

My tests at 2'+1", dualcore, completely agree with that!

[Dann Corbit]

OK, I only see memory overwrites in all of the versions from here:
http://www.chesslogik.com/ivanhoe.htm

The last one I tested was 49j.

I haven't tried Ivanhoes for quite a bit of time, I see movegen is correct in the recent version. Still, if what you say is true then it would be quite easy to check it and correct it in case the problem is in the source. However, if you don't compile it yourself, and use other compiles and complain then there is not much that can be done...

It is easy to find the problems.
1. Run PC-Lint by Gimpel software against the source code base
2. Run "Analyze"/"Run Code Analysis" from the VC++ IDE
These two tasks will identify likely array bounds over-writes.
Then, for each array address in question, put in asserts for addresses less than zero or addresses larger than the biggest allowed bound of dim-1.
You will find the problems without too much trouble.

The code is improving, but I guess that the programmers working on it are smart people without a lot of programming experience. I don't post on the Russian board any more because they threw me out for being a smart alec.

[Milos]

It is easy to find the problems.
1. Run PC-Lint by Gimpel software against the source code base
2. Run "Analyze"/"Run Code Analysis" from the VC++ IDE
These two tasks will identify likely array bounds over-writes.
Then, for each array address in question, put in asserts for addresses less than zero or addresses larger than the biggest allowed bound of dim-1.
You will find the problems without too much trouble.

When I know the code I solve problems directly through disassemble (just access violation address is enough). I tend not to rely too much on Analyze since it usually finds "problems" and overwrites where there are no real overwrites.

The code is improving, but I guess that the programmers working on it are smart people without a lot of programming experience. I don't post on the Russian board any more because they threw me out for being a smart alec.

The code is really improving but since there are not many new ideas, especially not those that work, I'm not checking it often. I guess couple of original guys are quite experienced but none of them is working on the code any more.

[Dann Corbit]

It is easy to find the problems.
1. Run PC-Lint by Gimpel software against the source code base
2. Run "Analyze"/"Run Code Analysis" from the VC++ IDE
These two tasks will identify likely array bounds over-writes.
Then, for each array address in question, put in asserts for addresses less than zero or addresses larger than the biggest allowed bound of dim-1.
You will find the problems without too much trouble.

When I know the code I solve problems directly through disassemble (just access violation address is enough). I tend not to rely too much on Analyze since it usually finds "problems" and overwrites where there are no real overwrites.

When the address is Array[-7] it's an overwrite.
When the address is Array[4][65] and the array is [2][64] it's an overwrite.
The assert() is only to discover which ones are real and which ones are not.

The code is improving, but I guess that the programmers working on it are smart people without a lot of programming experience. I don't post on the Russian board any more because they threw me out for being a smart alec.

The code is really improving but since there are not many new ideas, especially not those that work, I'm not checking it often. I guess couple of original guys are quite experienced but none of them is working on the code any more.

[De Vos W]

I have seen many discussions and rating with other versions of IvanHoe that are supposedly not official releases. I just wanted to know which IvanHoe is strongest and where can I download the newest Ivanhoes. After many games at 40/1 time controls it seems that IvanHoe 999957a is strongest. These results are based on very fast games though. My rating list is as follows.

Code: Select all

    Program                          Elo    +   -   Games   Score   Av.Op.  Draws

  1 Houdini 1.03a                  : 3242   23  23   440    60.1 %   3170   50.7 %
  2 IvanHoe 999957a                : 3223   25  25   380    57.5 %   3170   49.2 %
  3 Houdini 1.02                   : 3212   23  23   440    55.9 %   3170   49.5 %
  4 IvanHoe 999963                 : 3197   18  18   680    54.7 %   3164   53.5 %
  5 IvanHoe 999952a                : 3195   23  23   400    53.4 %   3171   55.8 %
  6 RobboLito 0085g3               : 3191   19  19   560    51.4 %   3181   56.1 %
  7 Fire 1.31 Warrior              : 3184   26  26   360    50.0 %   3184   48.9 %
  8 Stockfish 1.9                  : 3170   28  28   360    46.8 %   3192   38.6 %
  9 Fire 1.31                      : 3166   21  21   480    47.8 %   3181   53.5 %
 10 Stockfish 1.8                  : 3157   27  28   340    44.9 %   3193   45.0 %
 11 Stockfish 1.7.1 ZugDet         : 3145   21  21   600    46.0 %   3172   44.3 %
 12 Stockfish 1.7.1                : 3124   22  22   600    43.1 %   3172   38.5 %
 13 Rybka 3.0                      : 3123   18  18   840    44.3 %   3162   43.2 %

Thanks to anybody who can help me find the other IvanHoe Engines. Also, does anybody know if there are any vast differences in strength relative to other engines at longer time controls.

Thanks.........................................................................

Take a look at this forum: http://ahmed-chess.forumotion.com/forum.htm

Hi De Vos Walter,

Thank you for all of the information. I don't think that I ever knew about that chess forum. Seems like there's lots of good stuff there like several opening books and such. I will start downloading some of the new IvanHoes and start testing.

Your welcome

Any time!

DVW.

[David Schumaker]

I have seen many discussions and rating with other versions of IvanHoe that are supposedly not official releases. I just wanted to know which IvanHoe is strongest and where can I download the newest Ivanhoes. After many games at 40/1 time controls it seems that IvanHoe 999957a is strongest. These results are based on very fast games though. My rating list is as follows.

Code: Select all

    Program                          Elo    +   -   Games   Score   Av.Op.  Draws

  1 Houdini 1.03a                  : 3242   23  23   440    60.1 %   3170   50.7 %
  2 IvanHoe 999957a                : 3223   25  25   380    57.5 %   3170   49.2 %
  3 Houdini 1.02                   : 3212   23  23   440    55.9 %   3170   49.5 %
  4 IvanHoe 999963                 : 3197   18  18   680    54.7 %   3164   53.5 %
  5 IvanHoe 999952a                : 3195   23  23   400    53.4 %   3171   55.8 %
  6 RobboLito 0085g3               : 3191   19  19   560    51.4 %   3181   56.1 %
  7 Fire 1.31 Warrior              : 3184   26  26   360    50.0 %   3184   48.9 %
  8 Stockfish 1.9                  : 3170   28  28   360    46.8 %   3192   38.6 %
  9 Fire 1.31                      : 3166   21  21   480    47.8 %   3181   53.5 %
 10 Stockfish 1.8                  : 3157   27  28   340    44.9 %   3193   45.0 %
 11 Stockfish 1.7.1 ZugDet         : 3145   21  21   600    46.0 %   3172   44.3 %
 12 Stockfish 1.7.1                : 3124   22  22   600    43.1 %   3172   38.5 %
 13 Rybka 3.0                      : 3123   18  18   840    44.3 %   3162   43.2 %

Thanks to anybody who can help me find the other IvanHoe Engines. Also, does anybody know if there are any vast differences in strength relative to other engines at longer time controls.

Thanks.........................................................................

There are zillions of IvanHoe engines (probably 100+ builds released).
You can find a bunch of them here:
http://www.chesslogik.com/ivanhoe.htm

and you can do a web search using google to find dozens of other builds.

Nobody knows which one is strongest. Most of the testing seems a little helter-skelter or light. Sometimes openchess lists tests, but they are generally less than 100 games and therefore not very useful except as wild guestimates.

Hi Dann,

Is IvanHoe_999949j the latest "official" version. I have searched the web for the other IvanHoes but I can't seem to find any downloads. All I ever find are just references being made about them on various forums and on various ratings lists of which there are many.

[Dann Corbit]

This link has a whole bunch of them:
http://www.chesslogik.com/ivanhoe.htm

They have both source code and binaries in them, if I recall correctly.

[De Vos W]

For other IvanHoe's this link:

http://ibnalmimon.mylivepage.com/file/index/