Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Stobor vs. Crafty

9 views
Skip to first unread message

Tom C. Kerrigan

unread,
Jan 18, 1997, 3:00:00 AM1/18/97
to

A *lot* of people have asked me for these, so I'll post what I have. A few
things to note:

* Stobor is always white (it plays black starting tonight...)
* Stobor always wins, except for game 4, which was a draw
* time control is 40 moves/2 hours, then sudden death in an hour
* hash tables are 2mb
* pondering is off
* books are generated from world champions' PGN archives on the ftp site
* Crafty's book learning is on (Stobor has "Tom Learning" <grin>)
* machine is a p5/100
* if you notice Stobor doing funky things, feel free to tell me :)

About the machine: Stobor is generally 4x as fast as Crafty on a p5 and
only 2x on a p6, so Stobor winning all these games may not mean *that*
much...

[Event ""]
[Site ""]
[Date ""]
[Round ""]
[White ""]
[Black ""]
[Result ""]

1. e4 c6 2. d4 d5 3. Nc3 dxe4 4. Nxe4 Nf6 5. Nxf6+ gxf6 6. Bc4 e5 7. Be3
Rg8 8. dxe5 Qxd1+ 9. Rxd1 Be7 10. exf6 Bxf6 11. Kf1 Bxb2 12. Re1 Rg4 13.
Bc5+ Be6 14. Bxe6 fxe6 15. Rxe6+ Kd8 16. Be3 Ra4 17. Rh6 Rxa2 18. Rxh7 Nd7
19. Ke2 a6 20. Nf3 Ra1 21. Rxa1 Bxa1 22. g4 Rb8 23. Bg5+ Kc8 24. Bf4 Ra8
25. g5 b6 26. g6 Nf6 27. Rf7 c5 28. g7 Ng8 29. Nd2 a5 30. Ne4 Ra6 31. Rf8+
Kd7 32. Rxg8 Bxg7 33. Rxg7+ Ke6 34. Nxc5+ bxc5 35. Rg6+ Kf5 36. Rxa6 Kxf4
37. Rxa5 Ke5 38. Rxc5+ Ke6 39. c4 Kf6 40. Rd5 Ke6 41. Rd3 Kf5 42. c5 Ke5
43. c6 Kf4 44. c7 Kg4 45. c8=Q+ Kg5 46. Qf8 Kh5 47. Rg3 Kh4 48. Qh8#

[Event ""]
[Site ""]
[Date ""]
[Round ""]
[White ""]
[Black ""]
[Result ""]

1. e4 d5 2. exd5 Nf6 3. d4 Nxd5 4. Nf3 Bg4 5. c4 Nb6 6. c5 N6d7 7. h3 Bxf3
8. Qxf3 Nc6 9. Bc4 e6 10. Be3 Qf6 11. Qxf6 Nxf6 12. Bb5 Be7 13. Nd2 Nd5
14. Bxc6+ bxc6 15. Nc4 a5 16. Ke2 a4 17. Ne5 Ra6 18. f4 O-O 19. Nc4 Rb8
20. h4 Ra7 21. g3 Rab7 22. Rhe1 Bf6 23. Kf3 Rd8 24. Red1 Rb5 25. Rd2 Rdb8
26. h5 Rb4 27. Rc2 Ne7 28. g4 Nd5 29. g5 Be7 30. Rb1 R4b5 31. Ne5 Nb4 32.
Rg2 Nxa2 33. Nxc6 Re8 34. b4 Nc3 35. Ra1 Kh8 36. Ra3 Nd5 37. Rxa4 Rb7 38.
Rga2 Bf8 39. Ra8 Rxa8 40. Rxa8 Kg8 41. f5 exf5 42. Rd8 h6 43. gxh6 Nxe3
44. Rxf8+ Kh7 45. Kxe3 Kxh6 46. Rb8 Rxb8 47. Nxb8 Kg5 48. c6 f4+ 49. Kf3
Kf6 50. b5 Ke6 51. b6 Kd6 52. b7 Kd5 53. Na6 Kxc6 54. b8=Q Kd5 55. Nxc7+
Kxd4 56. Qb4+ Kd3 57. Nb5 Kc2 58. Na3+ Kd1 59. Qb1+ Kd2 60. Qc2+ Ke1 61.
Qe2#

[Event ""]
[Site ""]
[Date ""]
[Round ""]
[White ""]
[Black ""]
[Result ""]

1. e4 e6 2. d4 c5 3. d5 d6 4. Nc3 exd5 5. Bb5+ Bd7 6. Qxd5 Bxb5 7. Nxb5
Qd7 8. Qc4 a6 9. Nc3 Nf6 10. Nf3 Be7 11. O-O O-O 12. e5 b5 13. Qf4 dxe5
14. Nxe5 Qe6 15. Qf3 Nbd7 16. Nxd7 Nxd7 17. Bf4 Bf6 18. Nd5 Bxb2 19. Rae1
Be5 20. Nc7 Qd6 21. Nxa8 Bxf4 22. h3 Qb8 23. Kh1 Bd2 24. Re7 Qd6 25. Qe2
Bg5 26. Re8 Rxe8 27. Qxe8+ Nf8 28. a4 bxa4 29. Qxa4 Bd8 30. Rb1 h6 31.
Qg4 Ne6 32. Rb7 a5 33. Qf5 f6 34. c4 Qd1+ 35. Kh2 Qd6+ 36. g3 a4 37. Qd5
Qe5 38. Kg2 Kh7 39. Qxe5 fxe5 40. Ra7 Kg6 41. Rxa4 Kf5 42. Ra6 e4 43. h4
Nd4 44. Nb6 Bxb6 45. Rxb6 Ke5 46. h5 Nf5 47. Rb7 Kd4 48. Kh3 Kxc4 49. g4
e3 50. fxe3 Nxe3 51. Rxg7 Kc3 52. g5 Nf5 53. Rf7 Nd6 54. Rf3+ Kd2 55.
gxh6 Ne4 56. Kg4 Ke2 57. h7 Nd2 58. h8=Q Nxf3 59. Qe8+ Kf2 60. Qf8 Ke3 61.
Qf4+ Kd3 62. Qxf3+ Kc4 63. h6 Kb5 64. h7 Kb4 65. h8=Q Ka4 66. Qb8 c4 67.
Qfa8#

[Event ""]
[Site ""]
[Date ""]
[Round ""]
[White ""]
[Black ""]
[Result ""]

1. e4 e5 2. Nf3 d6 3. d4 exd4 4. Nxd4 Nf6 5. Nc3 Be7 6. Bc4 O-O 7. Bf4
Nbd7 8. O-O Ne5 9. Bb3 Bd7 10. Qe2 Qe8 11. Rad1 Bd8 12. Rd2 Nh5 13. Be3
Ng4 14. Ba4 Nxe3 15. Bxd7 Qxd7 16. Qxe3 Qg4 17. f4 Re8 18. Qf3 Qxf3 19.
Rxf3 Kf8 20. a3 Rc8 21. Rf1 a6 22. g3 h6 23. Kf2 Bf6 24. h4 Kg8 25. Kf3
Bxd4 26. Rxd4 Nf6 27. Rfd1 Re7 28. g4 Ne8 29. Nd5 Re6 30. f5 Re5 31. Kf4
Rd8 32. b4 Kf8 33. c4 c6 34. Nb6 c5 35. bxc5 Rxc5 36. R1d3 Rc6 37. Rb3 Nc7
38. a4 Re8 39. a5 Re7 40. Ke3 Kg8 41. Nc8 Nd5+ 42. cxd5 Rxc8 43. Rdb4 Re5
44. Kf3 Rc1 45. Rb1 Rc3+ 46. Kf4 Ra3 47. Rb6 Rxa5 48. Rxb7 Rc5 49. R7b4
Re8 50. Rb8 Rf8 51. R8b6 a5 52. Rxd6 Rc4 53. Rd7 Re8 54. Re1 f6 55. Kf3
a4 56. Ra7 Kf8 57. d6 Rd4 58. Rd7 Re5 59. Re3 Rc5 60. Rd8+ Kf7 61. Re2
Rc3+ 62. Kf4 Rd1 63. Rf2 Rh3 64. Rf3 Rhh1 65. Rd7+ Kf8 66. h5 Rhe1 67.
Rc3 Rf1+ 68. Kg3 Rg1+ 69. Kh4 Rh1+ 70. Rh3 Rhe1 71. Rd8+ Kf7 72. Rc3 Rh1+
73. Kg3 Rhg1+ 74. Kf4 Rgf1+ 75. Rf3 Rfe1 76. Ra8 Rd4 77. Re3 Rf1+ 78. Kg3
Rg1+ 79. Kf2 Rc1 80. Ra7+ Kf8 81. Re2 Rc8 82. Ra2 Rxd6 83. R2xa4 Rc2+ 84.
Ke3 Kg8 85. Ra2 Rc3+ 86. Kf4 Rd1 87. Ra1 Rd2 88. Ra8+ Kh7 89. R8a2 Rdd3
90. Rf2 Rd7 91. Ra4 Rc1 92. e5 fxe5+ 93. Kxe5 Re7+ 94. Kd6 Rec7 95. Kd5
Rc8 96. Rb2 R8c7 97. Rab4 R7c5+ 98. Ke6 Rc6+ 99. Ke5 R6c5+ 100. Kf4 R5c3
101. Re4 Rf1+ 102. Ke5 Rc8 103. Ra4 Re8+ 104. Kd6 Rd1+ 105. Kc7 Re7+ 106.
Kc6 Rc1+ 107. Kd6 Rec7 108. Kd5 R1c3 109. Kd4 Rg3 110. Ke5 Re3+ 111. Kd6
Rec3 112. Rab4 R7c6+ 113. Kd5 R6c5+ 114. Kd4 Rc8 115. Rd2 R8c7 116. Ke4
Re7+ 117. Kf4 Ree3 118. Rb7 Rf3+ 119. Ke5 Rce3+ 120. Kd6 Re4 121. Rg2
Rd3+ 122. Kc5 Rc3+ 123. Kd5 Rec4 124. Kd6 Kg8 125. Ke5 Rc5+ 126. Ke6 R3c4
127. Kd6 Rc6+ 128. Ke7 Rf6 129. Rb8+ Kh7 130. Rbb2 Re4+ 131. Kd7 Rd4+
132. Kc7 Rc4+ 133. Kb7 Rfc6 134. Rbd2 Rc7+ 135. Kb6 R7c6+ 136. Kb5 R6c5+
137. Kb6 Rc6+ 138. Kb7 Rc7+ 139. Ka8 Rc8+ 140. Ka7 R8c7+ 141. Kb6 R7c6+

[Event ""]
[Site ""]
[Date ""]
[Round ""]
[White ""]
[Black ""]
[Result ""]

1. e4 e5 2. Nf3 d6 3. d4 exd4 4. Nxd4 Nf6 5. Nc3 Be7 6. Bc4 O-O 7. Bf4
Nbd7 8. O-O Ne5 9. Bb3 Bd7 10. Qe2 Qe8 11. Rad1 Bd8 12. Rd2 Nh5 13. Be3
Ng4 14. Ba4 Nxe3 15. Bxd7 Qxd7 16. Qxe3 Qg4 17. f4 Re8 18. Qf3 Qxf3 19.
Rxf3 h6 20. Nf5 b5 21. Nxb5 Rxe4 22. g3 Re1+ 23. Kf2 Rb1 24. c4 Nf6 25.
Nc3 Rc1 26. Re2 a5 27. h3 a4 28. Rd3 h5 29. Nd4 Ra6 30. c5 g6 31. Nc2 Nd7
32. Ne4 Rb1 33. Nb4 Ra8 34. Nc3 Rc1 35. cxd6 cxd6 36. Rxd6 Bb6+ 37. Kf3
Bc5 38. Rxd7 Bxb4 39. Ne4 Rb8 40. Ra7 h4 41. Rxa4 hxg3 42. Ra7 Rg1 43.
Nf6+ Kg7 44. Nd7 Rc8 45. Ne5 Rf8 46. Nd3 Bd6 47. Rd7 Bb8 48. h4 Kf6 49.
Re1 Rxe1 50. Nxe1 Ke6 51. Rb7 f6 52. Kxg3 g5 53. Ng2 gxf4+ 54. Kf3 Kf5 55.
h5 Kg5 56. a4 Rd8 57. h6 Kxh6 58. Nxf4 Bxf4 59. Kxf4 Rd4+ 60. Kf5 Rxa4
61. Kxf6 Kh5 62. b4 Kg4 63. Ke5 Kf3 64. b5 Re4+ 65. Kd5 Rf4 66. b6 Rf6
67. Rb8 Rf5+ 68. Ke6 Rb5 69. b7 Ke4 70. Kd6 Kd3 71. Rd8 Rxb7 72. Kc6+ Ke4
73. Kxb7 Ke5 74. Kb6 Ke4 75. Kc5 Ke3 76. Re8+ Kf4 77. Kd5 Kf3 78. Kc6 Kf4
79. Kd6 Kf5 80. Re5+ Kf4 81. Ke6 Kf3 82. Kf5 Kg3 83. Re3+ Kg2 84. Kg4 Kf2
85. Kf4 Kg2 86. Re2+ Kf1 87. Kf3 Kg1 88. Re1+ Kh2 89. Rd1 Kh3 90. Rh1#

Chris Whittington

unread,
Jan 18, 1997, 3:00:00 AM1/18/97
to

--
http://www.demon.co.uk/oxford-soft

Tom C. Kerrigan <kerr...@merlin.pn.org> wrote in article
<5bqq69$8...@merlin.pn.org>...


> A *lot* of people have asked me for these, so I'll post what I have. A
few
> things to note:
>
> * Stobor is always white (it plays black starting tonight...)
> * Stobor always wins, except for game 4, which was a draw
> * time control is 40 moves/2 hours, then sudden death in an hour
> * hash tables are 2mb
> * pondering is off
> * books are generated from world champions' PGN archives on the ftp site
> * Crafty's book learning is on (Stobor has "Tom Learning" <grin>)
> * machine is a p5/100
> * if you notice Stobor doing funky things, feel free to tell me :)
>
> About the machine: Stobor is generally 4x as fast as Crafty on a p5 and
> only 2x on a p6, so Stobor winning all these games may not mean *that*
> much...

What happened in the blitz games against CSTal, round at czub's place ?

Chris Whittington

Robert Hyatt

unread,
Jan 20, 1997, 3:00:00 AM1/20/97
to

Tom C. Kerrigan (kerr...@merlin.pn.org) wrote:
: A *lot* of people have asked me for these, so I'll post what I have. A few
: things to note:

: * Stobor is always white (it plays black starting tonight...)
: * Stobor always wins, except for game 4, which was a draw
: * time control is 40 moves/2 hours, then sudden death in an hour
: * hash tables are 2mb
: * pondering is off
: * books are generated from world champions' PGN archives on the ftp site
: * Crafty's book learning is on (Stobor has "Tom Learning" <grin>)
: * machine is a p5/100
: * if you notice Stobor doing funky things, feel free to tell me :)

: About the machine: Stobor is generally 4x as fast as Crafty on a p5 and
: only 2x on a p6, so Stobor winning all these games may not mean *that*
: much...

I'm not particularly excited about this, because it doesn't match with
reality in any sense. Crafty's played stobor dozens of games on ICC, on
equal P6/200 machines, and the results are absolutely nothing like the
above. If you want to post games, they ought to mean something. Meaning
show up on ICC, or have Hornets show up and simply play Crafty there. It
won't play oddball openings like the games you posted, and you'll be getting
results against a real program.

There's several things wrong with what you posted:

1. such a record implies you are 400+ rating points better than Crafty.
That's hogwash, based on ICC ratings, Crafty's results at the last WMCC,
and Stobor's results at the event before the last one. The results make
no sense, and are so outlandish they simply aren't believable.

2. In the games I have between ratbert and Crafty on ICC, ratbert wins
about 60% of the bullet games... it is very fast, and at bullet, that
makes a big different. At blitz games, Crafty is winning about 70%...
Blitz being several different time controls, from 2 12 to 5 5 and 5 0.

3. in searching ICC's database, there is one standard game, crafty won.
Dated 9/26/96... in blitz ICC's database has Crafty winning 64, losing 10.
In bullet, games are not stored, but as I said, my database of *all* games
played on ICC shows ratbert winning more than it loses..

4. Crafty has played hundreds of games against Genius, WchessX, CM5000,
Fritz, Rebel, MchessPro, and none of 'em are winning at the rate you
posted. In fact, there's not one that I know of that is playing well
enough to even win three out of four games against Crafty.

5. Chris and I played a match a few months back that was pretty close to
even as I recall. I think crafty won two, lost 1 and drew two, but I'm
not really sure.

In short, I'd simply brand this as either (a) pure bullshit; or (b) a very
badly controlled experiment. The easy way to confirm or disprove this is to
play some games in public on ICC, not in private using a version of Crafty
that's unknown, with a book that's questionable, with settings that appear
to be grossly off.

Try the real McCoy. If you can win 20 in a row, I'll be a believer. If not,
at least the question is solved. However, when you start trying to convince
me that the gravitational constant for planet earth is something other than
32 feet per second per second at sea level, I'm suspicious. *very* suspicious.
Fortunately, it's easy to prove or disprove... Crafty doesn't hide from
anyone...


Peter Herttrich

unread,
Jan 20, 1997, 3:00:00 AM1/20/97
to

In rec.games.chess.computer Robert Hyatt <hy...@cis.uab.edu> wrote:
: Tom C. Kerrigan (kerr...@merlin.pn.org) wrote:
: : A *lot* of people have asked me for these, so I'll post what I have. A few
: : things to note:

: : * Stobor is always white (it plays black starting tonight...)
: : * Stobor always wins, except for game 4, which was a draw


Stupid question: Who or what is STOBOR ???

Is this a commercial or free program? Never heard about it.

Thanks.

Cheerio
peter

--
--
_____________________________________________________________________________
Peter Herttrich email: dh1...@inss1.etec.uni-karlsruhe.de
Universitaet Karlsruhe Tel. +49 721 6083747 FAX +49 721 6086071
Institut fuer Nachrichtentechnik ..life outside caves is complicated ...
_____________________________________________________________________________

brucemo

unread,
Jan 20, 1997, 3:00:00 AM1/20/97
to

Robert Hyatt wrote:
>
> Tom C. Kerrigan (kerr...@merlin.pn.org) wrote:
> : A *lot* of people have asked me for these, so I'll post what I have. A few

> : things to note:
>
> : * Stobor is always white (it plays black starting tonight...)
> : * Stobor always wins, except for game 4, which was a draw
> : * time control is 40 moves/2 hours, then sudden death in an hour
> : * hash tables are 2mb
> : * pondering is off
> : * books are generated from world champions' PGN archives on the ftp site
> : * Crafty's book learning is on (Stobor has "Tom Learning" <grin>)
> : * machine is a p5/100
> : * if you notice Stobor doing funky things, feel free to tell me :)
>
> : About the machine: Stobor is generally 4x as fast as Crafty on a p5 and
> : only 2x on a p6, so Stobor winning all these games may not mean *that*
> : much...

I never got the original post, so I'm not quite sure what is going on here.

Questions for Tom:

1) How many nps is Stobor getting on the P6, typically?

2) How many nps was Crafty getting on that machine?

3) Is this approximately the same Stobor as last year, or are you making changes?

4) Is this version the one that has been running on ICC as Ratbert recently?

5) What are the details of this match? How many games were played and what was
the score?

> 1. such a record implies you are 400+ rating points better than Crafty.
> That's hogwash, based on ICC ratings, Crafty's results at the last WMCC,
> and Stobor's results at the event before the last one. The results make
> no sense, and are so outlandish they simply aren't believable.

I doubt if any micro program is 400 points better than Crafty.

Stobor had some bad bugs at that WMCCC. For instance, Tom had his king safety
upside-down, so the program wanted to put its king in front of the k-side pawns
rather than behind them. After he came back from Paderborn there was a period
during which it became much better.

> 2. In the games I have between ratbert and Crafty on ICC, ratbert wins
> about 60% of the bullet games... it is very fast, and at bullet, that
> makes a big different. At blitz games, Crafty is winning about 70%...
> Blitz being several different time controls, from 2 12 to 5 5 and 5 0.

I've seen the same thing. It is stronger at bullet than it is at blitz. It's
nothing to sneeze at at blitz though. You and I have both been pretty mad at Tom
upon occasion, but make no mistake about it, his program is strong.

> Try the real McCoy. If you can win 20 in a row, I'll be a believer. If not,
> at least the question is solved. However, when you start trying to convince
> me that the gravitational constant for planet earth is something other than
> 32 feet per second per second at sea level, I'm suspicious. *very* suspicious.
> Fortunately, it's easy to prove or disprove... Crafty doesn't hide from
> anyone...

Tom is welcome to play against mine, too. It has timestamp now.

bruce

Tom C. Kerrigan

unread,
Jan 20, 1997, 3:00:00 AM1/20/97
to

Robert Hyatt (hy...@cis.uab.edu) wrote:

> I'm not particularly excited about this, because it doesn't match with
> reality in any sense. Crafty's played stobor dozens of games on ICC, on

I'm sorry, but I did not post this so you could get particularly excited
about it. I posted this because I've had at least 5 requests for the
games, and I thought that warranted a post.

> In short, I'd simply brand this as either (a) pure bullshit; or (b) a very
> badly controlled experiment. The easy way to confirm or disprove this is to

It isn't bullshit. It isn't even an experiment. I'm simply playing these
games at night to see what might go wrong at the tournament in February.
If anything, I want Stobor to *lose*. (Unfortunately, it hasn't done this
yet. Stobor's first game as black was a win, too.)

> Fortunately, it's easy to prove or disprove... Crafty doesn't hide from
> anyone...

Prove or disprove *what*? That these are the games that have taken place
on my machine over the past week? Good luck...

I'm not threatening your manliness, Bob! Relax! Go take a stiff drink and
ride around in your boat, or whatever...

Cheers,
Tom

P.S. The version of Stobor that Ratbert plays with is 5 months old. For
all you know, I've rewritten every single line of code.

Robert Hyatt

unread,
Jan 21, 1997, 3:00:00 AM1/21/97
to

Tom C. Kerrigan (kerr...@merlin.pn.org) wrote:
: Robert Hyatt (hy...@cis.uab.edu) wrote:

: Cheers,
: Tom

Yep. but I *know* that what you are running is not a "real" crafty. On
my P5/120 notebook, crafty searches about 26-28K nodes per second as a
lower bound... That's about 2-2.5X faster than the number you posted for
your P5/100... something's wrong. Either you didn't compile it correctly,
didn't optimize it correctly, or whatever.

In any case, Stobor isn't 400+ rating points better. That's the sort of
claim I'd expect from Vincent at times. You ought to know better. As I
said, you can solve this by playing Crafty on ICC... any time control you'd
like. If you only have a P5/100, I'll log my P5/120 notebook on as Scrappy,
although Crafty is really optimized more toward the P6.

However, when someone requests data from you that is suspect at best,
emailing is the least offensive way to handle that. Should I post the 60
games Crafty won on ICC and the 10 it lost playing Stobor? At least those
games were played with the best crafty 6 months ago vs the best Stobor 6
months ago, on equal machines... You are running crafty in an unknown state,
with a book that's unknown, obviously not using the start.pgn file from the
ftp machine since crafty does *not* play 1. ... c6 in response to 1. d4...
In short, you publish games where you have a book that suites your program,
and use that same book against Crafty... We don't know how you compiled
Crafty but it certainly seems slow... In fact, I've not seen numbers that
low excepting for one P5/60 I know of running on ICC... My old P5/75
notebook was faster than that with *no* L2 cache. With a bad book, a slow
executable, it really sounds like a "fair" test doesn't it???

Simply don't post garbage... it's easy to avoid. Because it is often
mistaken for the truth. When it is really just garbage...


Robert Hyatt

unread,
Jan 21, 1997, 3:00:00 AM1/21/97
to

Peter Herttrich (dh1...@inss1.etec.uni-karlsruhe.de) wrote:
: In rec.games.chess.computer Robert Hyatt <hy...@cis.uab.edu> wrote:
: : Tom C. Kerrigan (kerr...@merlin.pn.org) wrote:
: : : A *lot* of people have asked me for these, so I'll post what I have. A few
: : : things to note:

: : : * Stobor is always white (it plays black starting tonight...)
: : : * Stobor always wins, except for game 4, which was a draw


: Stupid question: Who or what is STOBOR ???

: Is this a commercial or free program? Never heard about it.

: Thanks.

: Cheerio
: peter

Non-commercial program written by Tom Kerrigan. It's often run on ICC under
the name Ratbert, although you can never be sure. Sometimes Ratbert is Crafty
sometimes it is Stobor. It's difficult to know when it's what except if you
happen to see Ratbert play another computer, because Crafty will automatically
start kibitzing analysis against certain programs...

Tom C. Kerrigan

unread,
Jan 21, 1997, 3:00:00 AM1/21/97
to

Good question, actually. :)

Stobor is the name of my chess program. It's been around since June or
July '95. It got 25th at the '95 WMCCC, which isn't bad for a first
tournament, but I think it would have done a lot better if I had another
month to find and correct a few of the more horrible bugs. I hope to do
better this February. It's also the guts for "Ratbert" on ICC, which seems
to be doing well (although it uses a very dated version). It isn't
commercial, but this may change when I get back to the US (this summer).

Cheers,
Tom

Peter Herttrich (dh1...@inss1.etec.uni-karlsruhe.de) wrote:
> In rec.games.chess.computer Robert Hyatt <hy...@cis.uab.edu> wrote:
> : Tom C. Kerrigan (kerr...@merlin.pn.org) wrote:

> : : A *lot* of people have asked me for these, so I'll post what I have. A few
> : : things to note:

> : : * Stobor is always white (it plays black starting tonight...)
> : : * Stobor always wins, except for game 4, which was a draw

> Stupid question: Who or what is STOBOR ???

> Is this a commercial or free program? Never heard about it.

> Thanks.

> Cheerio
> peter

> --

Tom C. Kerrigan

unread,
Jan 21, 1997, 3:00:00 AM1/21/97
to

brucemo (bru...@nwlink.com) wrote:

> Questions for Tom:

> 1) How many nps is Stobor getting on the P6, typically?

Ratbert gets around 120k, if memory serves. I've made my guy faster, but
not much more than, say, 25%...

> 2) How many nps was Crafty getting on that machine?

Crafty gets somewhere between 8 to 15k on this p5/100.

> 3) Is this approximately the same Stobor as last year, or are you making changes?

Last year as in Paderborn? Besides having king safety half backwards, that
program basically had no hash table, [almost] randomly pruned moves in the
quiescence search, and had to finish a ply before switching the PV. These
are all the major bugs that I can remember, but there are probably more.
Basically, I'm astounded that it did as well as it did. Then again, it was
just over two months old...

> 4) Is this version the one that has been running on ICC as Ratbert recently?

Nowhere close. The book code, time control code, extension handling code,
and endgame code have been totally rewritten. The rest of the program has
been changed fairly radically, along with some quasi-serious bug fixes.
I'd like to think the program is a bit stronger now. :) Unfortunately,
updating Ratbert is a world-class pain in the ass for various reasons. I
will work on this after February.

> 5) What are the details of this match? How many games were played and what was
> the score?

So far, 9 games have been played. Stobor won 6, drew 2, and lost 2. I
posted the first 5, where Stobor scored 4.5.

I'm happy to report that I fixed the bug that caused one of the losses
this afternoon. :)

> Tom is welcome to play against mine, too. It has timestamp now.

I want to do this, but it's fairly impossible right now. When I get back,
certainly. Perhaps earlier.

Cheers,
Tom

Tom C. Kerrigan

unread,
Jan 21, 1997, 3:00:00 AM1/21/97
to

Tom C. Kerrigan (kerr...@merlin.pn.org) wrote:

> So far, 9 games have been played. Stobor won 6, drew 2, and lost 2. I
> posted the first 5, where Stobor scored 4.5.

Dammit. 5 wins. I can add. Really. :)

Cheers,
Tom

Peter W. Gillgasch

unread,
Jan 21, 1997, 3:00:00 AM1/21/97
to

Robert Hyatt <hy...@cis.uab.edu> wrote:

> Simply don't post garbage... it's easy to avoid. Because it is often
> mistaken for the truth. When it is really just garbage...

Wah, guys, why do you post this "mine is bigger than yours" type of
stuff... Fight it out at the next tourney and until then keep up
working... I won't post any Chess Demon data. Simply will look what
happens at the next tourney...

-- Peter

May God grant me the serenity to accept the things I cannot change,
courage to choke the living shit out of those who piss me off,
and wisdom to know where I should hide the bodies...

Robert Hyatt

unread,
Jan 22, 1997, 3:00:00 AM1/22/97
to

Peter W. Gillgasch (gil...@ilk.de) wrote:
: Robert Hyatt <hy...@cis.uab.edu> wrote:

: > Simply don't post garbage... it's easy to avoid. Because it is often
: > mistaken for the truth. When it is really just garbage...

: Wah, guys, why do you post this "mine is bigger than yours" type of
: stuff... Fight it out at the next tourney and until then keep up
: working... I won't post any Chess Demon data. Simply will look what
: happens at the next tourney...

: -- Peter

I hope my post was not a "mine is bigger than yours" as I specifically
avoid that sort of comparison here. I simply responded to some data
that doesn't make a lot of sense, based on what I know... It's better
to post real data, from real games, against real opponents. Rather than
games played offline, by unknown opponents with unknown conditions. The
data can be taken as true or false, and either could be correct. If
someone wants to use Crafty as an opponent, please do so. But use
"crafty"... which is well-known. Don't use something less than "crafty"
which is simply the program that runs on ICC and/or chess.net all the
time. It's a known opponent. With a known skill level...


brucemo

unread,
Jan 22, 1997, 3:00:00 AM1/22/97
to

Tom C. Kerrigan wrote:
>
> brucemo (bru...@nwlink.com) wrote:

> > 1) How many nps is Stobor getting on the P6, typically?
>
> Ratbert gets around 120k, if memory serves. I've made my guy faster, but
> not much more than, say, 25%...

Crafty should be getting 70-100K middlegame perhaps, so if it's getting 30K
(since you said it's 1/4 as fast as Stobor on the P6), something is wrong
with your configuration.

>
> > 2) How many nps was Crafty getting on that machine?
>
> Crafty gets somewhere between 8 to 15k on this p5/100.

Sounds low for Crafty, but Bob would know better.

Are you running both programs on the same machine? Yours is a Windows
program, right? Are you polling in your message loop or something? Does
Crafty go any faster if you let it run by itself?

> > 3) Is this approximately the same Stobor as last year, or are you making changes?
>
> Last year as in Paderborn? Besides having king safety half backwards, that
> program basically had no hash table, [almost] randomly pruned moves in the
> quiescence search, and had to finish a ply before switching the PV. These
> are all the major bugs that I can remember, but there are probably more.
> Basically, I'm astounded that it did as well as it did. Then again, it was
> just over two months old...

Last year as in before you went off to Germany to study. But I think you
answered this later in the post.

bruce

Tom C. Kerrigan

unread,
Jan 23, 1997, 3:00:00 AM1/23/97
to

Chris Whittington (chr...@demon.co.uk) wrote:

> What happened in the blitz games against CSTal, round at czub's place ?

Only one blitz game was played, and Stobor beat the tar out of CST. :)

We played two games where the programs averaged a move a minute. Stobor
got up in those, and then CST made a "sacrifice" in each and got ahead by
a pawn or two. Stobor managed to draw both.

I say sacrifice in quotes because they looked like sacrifices but Stobor
saw it was in trouble almost instantly. It isn't clear my guy wouldn't
have played the same way. It almost certainly would on a [much?] faster
machine...

Cheers,
Tom

Tom C. Kerrigan

unread,
Jan 23, 1997, 3:00:00 AM1/23/97
to

Robert Hyatt (hy...@cis.uab.edu) wrote:

> that doesn't make a lot of sense, based on what I know... It's better
> to post real data, from real games, against real opponents. Rather than

Which I did... unless you don't consider your program a real opponent?

> games played offline, by unknown opponents with unknown conditions. The

I posted all sorts of stuff about the conditions. I did forget one thing:
I'm using Crafty 11.13 without tablebases. Compiled straight out of the
"box". Almost. I added around 15 lines of code for the autoplayer.

> data can be taken as true or false, and either could be correct. If
> someone wants to use Crafty as an opponent, please do so. But use
> "crafty"... which is well-known. Don't use something less than "crafty"
> which is simply the program that runs on ICC and/or chess.net all the
> time. It's a known opponent. With a known skill level...

I want to play a game every night at IPCCC time controls. Why do I need a
known opponent with a known skill level? Besides, I have no automatic ICC
interface and the only time I can get on ICC is between midnight and 1 AM.
What do you suggest *now*?

Hyatt, questions:

1) If all the games I posted were losses for Stobor, would you imply that
I probably improved Crafty, bitch about ICC statistics, tell people
to take the results with a grain of salt, etc.?

2) If I posted Stobor vs. GNU Chess games, would you reply?

Now tell us your post was not a "mine is bigger than yours."

BTW, Stobor now has 7 points out of 11. I will post the new games, too, if
there's interest.

Cheers,
Tom

Tom C. Kerrigan

unread,
Jan 23, 1997, 3:00:00 AM1/23/97
to

brucemo (bru...@nwlink.com) wrote:

> Crafty should be getting 70-100K middlegame perhaps, so if it's getting 30K
> (since you said it's 1/4 as fast as Stobor on the P6), something is wrong
> with your configuration.

I think maybe you misread a bit. I don't have a p6. Crafty is a 1/4 as
fast on the p5/100 that I do have.

> Are you running both programs on the same machine? Yours is a Windows
> program, right? Are you polling in your message loop or something? Does
> Crafty go any faster if you let it run by itself?

This is being done with LINUX. The programs sleep instead of ponder, so
they each get 100% CPU time.

> Last year as in before you went off to Germany to study. But I think you
> answered this later in the post.

Yes.

Cheers,
Tom

Chris Whittington

unread,
Jan 23, 1997, 3:00:00 AM1/23/97
to

--
http://www.demon.co.uk/oxford-soft

Peter W. Gillgasch <gil...@ilk.de> wrote in article
<1997012121501819366@[194.121.104.134]>...


> Robert Hyatt <hy...@cis.uab.edu> wrote:
>
> > Simply don't post garbage... it's easy to avoid. Because it is often
> > mistaken for the truth. When it is really just garbage...
>
> Wah, guys, why do you post this "mine is bigger than yours" type of
> stuff...

Weren't Hyatt and Kerrigan the two guys who got themselves into the 'my
dick's bigger than your dick' argument last time.

And I jest not, it really was about phallus size :)

Keep it up, guys, as the actress said to the bishop :)

Chris Whittington


Robert Hyatt

unread,
Jan 24, 1997, 3:00:00 AM1/24/97
to

Tom C. Kerrigan (kerr...@merlin.pn.org) wrote:
: Robert Hyatt (hy...@cis.uab.edu) wrote:

: > that doesn't make a lot of sense, based on what I know... It's better
: > to post real data, from real games, against real opponents. Rather than

: Which I did... unless you don't consider your program a real opponent?

: > games played offline, by unknown opponents with unknown conditions. The

: I posted all sorts of stuff about the conditions. I did forget one thing:
: I'm using Crafty 11.13 without tablebases. Compiled straight out of the
: "box". Almost. I added around 15 lines of code for the autoplayer.

: > data can be taken as true or false, and either could be correct. If
: > someone wants to use Crafty as an opponent, please do so. But use
: > "crafty"... which is well-known. Don't use something less than "crafty"
: > which is simply the program that runs on ICC and/or chess.net all the
: > time. It's a known opponent. With a known skill level...

: I want to play a game every night at IPCCC time controls. Why do I need a
: known opponent with a known skill level? Besides, I have no automatic ICC
: interface and the only time I can get on ICC is between midnight and 1 AM.
: What do you suggest *now*?

Because you posted results against a "known" opponent, when, in fact, the
results were not against a "known opponent". You have apparently compiled
it in a way that makes it very slow, somehow, because there's at least one
crafty on ICC on a P5/60 that is as fast or faster than what you were
reporting... on my P5/120 I typically see at least 24K nps and usually more.
You used an oddball book, apparently without the standard start.pgn file, and
apparently without reading the documentation. This lets "crafty" play any
off-the-wall opening it chooses. If you play a thousand games, it will learn to
play reasonable openings, but you reported only a few. the correct start.pgn
will avoid the cramped positions I saw in two of the games. In short, you
had the "crafty" source, but that's about it, and that's what I meant when I
said your experiment was "flawed"... because Crafty was not set to play its
best, while I assume Stobor was... if you run against the one of ICC, you get
the best I know how to do at present... Crafty, Halcion, Computer, are three
good choices, fitter is a little slower (P166+) and there are some on slower
machines to boot.

: Hyatt, questions:

: 1) If all the games I posted were losses for Stobor, would you imply that
: I probably improved Crafty, bitch about ICC statistics, tell people
: to take the results with a grain of salt, etc.?

I wouldn't imply anything. I played lots of games way back against Stobor
when I had it running here. I didn't report *any* results, for the reasons
given above. I still have the results if you'd like to see them in public or
in private... I don't think I have the actual crafty log-files any longer
however.

: 2) If I posted Stobor vs. GNU Chess games, would you reply?

No, because GNU Chessis not something I'm responsible for. I am responsible
for Crafty. And the results you reported still don't sound reasonable, although
certainly 7 out of 11 doesn't sound nearly as bad as your "Stobor has won every
game so far" that you originally posted. If I could count the number of times
I've played matches against GnuChess here and started off 10-0, and endedup 30-22
or some such, I'd have a big number.

: Now tell us your post was not a "mine is bigger than yours."

: BTW, Stobor now has 7 points out of 11. I will post the new games, too, if
: there's interest.

: Cheers,
: Tom

As I said, the offer still stands. Play the *real* crafty... not one with a
crippled opening book (which can result in losses before the game gets started)
and not one that is obscenely slow (for whatever reason.) If you win 'em all,
I'll be a believer. If you win 7 out of 11, I'll still be a believer. Right
now I'm looking at contaminated data.


brucemo

unread,
Jan 24, 1997, 3:00:00 AM1/24/97
to

Chris Whittington wrote:

> Weren't Hyatt and Kerrigan the two guys who got themselves into the 'my
> dick's bigger than your dick' argument last time.
>
> And I jest not, it really was about phallus size :)

Yes, they were involved, but it was a one-sided argument from Kerrigan, I
believe.

The problems that I've seen have tended to start with Kerrigan making a
very authoritative-sounding statement that makes Bob furious. Bob replies
very vehemently and Kerrigan replies with all possible violence.

I usually agree with Bob, although I wouldn't write the replies quite the
way he does. Some people may have the tendency to applaud Tom, though. I
often write replies that are essentially "shut up, Tom", although I
generally have the sense to discard them before sending.

It's interesting the way we have alliances in here. I usually lump you
and Thorsten together, although I do realize that he has a lot more snakes
and lizards and lightning bolts coming out of his ears than you do.

bruce

Tom C. Kerrigan

unread,
Jan 25, 1997, 3:00:00 AM1/25/97
to

Robert Hyatt (hy...@cis.uab.edu) wrote:

> Because you posted results against a "known" opponent, when, in fact, the
> results were not against a "known opponent". You have apparently compiled

I get it! Your entire problem with what I posted is that I called your
program Crafty. Next time I post such games (if there is a next time) I'll
be sure to just call it "some program Hyatt wrote." You could have just
explained this earlier, instead of going off about how I screwed
something up...

> it in a way that makes it very slow, somehow, because there's at least one
> crafty on ICC on a P5/60 that is as fast or faster than what you were
> reporting... on my P5/120 I typically see at least 24K nps and usually more.

Bummer. I know this is a sucky p5/100, but I didn't realize it was that
sucky. Unfortunately, with less than a month left until the tournament, I
feel no pressing need to optimize somebody else's program... Perhaps after
February...

> given above. I still have the results if you'd like to see them in public or
> in private... I don't think I have the actual crafty log-files any longer

With all the changes I've made to Stobor, I think this data is too
"outdated" to be of value. Thanks for the offer, though.

> As I said, the offer still stands. Play the *real* crafty... not one with a

And as I've explained, this is very unlikely to happen before I get back
to the US...

Fortunately, my memory is not so bad that you have to repeat the offer
every post you make. :)

BTW, current score: 8/12...

Cheers,
Tom

brucemo

unread,
Jan 25, 1997, 3:00:00 AM1/25/97
to

Tom C. Kerrigan wrote:

> Bummer. I know this is a sucky p5/100, but I didn't realize it was that
> sucky. Unfortunately, with less than a month left until the tournament, I
> feel no pressing need to optimize somebody else's program... Perhaps after
> February...

It's probably not the program. It could be the hash table settings, the
operating system you are using, or something else as trivial.

Or it could be the computer itself. I have a Zeos Pantera P5/90. It's a
reasonable machine for most things. It runs other chess programs nice and
fast. But it can't run mine for beans. I tried three different operating
systems, numerous hash table settings, and several compilers, and always
slow, slow, slow.

It's way slower than the Gateway P5/66 I had at work at that time. For
other programs, not slow, but for mine, slow.

I'm pretty sure it had to do with the configuration of the hardware itself,
as Ferret is more memory-intensive than other programs.

I finally gave the thing to my kid and got another Gateway.

So first, your comment about optimization is probably wrong, given that
there are other machines of that class that run Crafty better.

Additionally, I don't understand why you need to make rude comments about
Crafty, and by extension, about Bob. Your comment absolutely drips sarcasm
and condescension, and is targetted directly at the quality of work Bob has
been doing for the past year and a half or so.

It's even worse because Bob has helped you design your program, and this is
the kind of thing you give back.

It should be very obvious to you the effect this kind of comment will have
upon Bob, and unless your main purpose here is to infuriate people, can't
you make an effort to be a more diplomatic?

bruce

Joe Stella

unread,
Jan 25, 1997, 3:00:00 AM1/25/97
to

kerr...@merlin.pn.org (Tom C. Kerrigan) wrote:

>[...]


>Bummer. I know this is a sucky p5/100, but I didn't realize it was that
>sucky. Unfortunately, with less than a month left until the tournament,

>[...]

It may have nothing to do with your computer itself. It may just
be your compiler, or the compiler options you used. I compiled
Crafty for my 486/DX4-100 Linux box, and it does 7,000 nps. I also
downloaded "crafty.linux" from Bob's site, and that does 9,500 nps
on the same system.

My compiler is gcc 2.7.2-9. It doesn't understand the options
"-mpentium" (no matter since the system isn't a pentium) as well as
"-fswap-for-agi", "-frisc-const", and "-fschedule-stack-insns".

Clearly Bob has optimized Crafty for the compiler he uses, and
if you do anything different from what he does then your mileage will
vary.

Joe Stella

mclane

unread,
Jan 25, 1997, 3:00:00 AM1/25/97
to

brucemo <bru...@nwlink.com> wrote:

>Chris Whittington wrote:

>bruce

Thanks , maybe it has to do with the fact that we are younger than
youn are. Chris is older than I am.
bob may be older than you. Tom is the youngster.

Let us play chess..... right !!!!

A nice tounament here in the newsgroup....


Robert Hyatt

unread,
Jan 26, 1997, 3:00:00 AM1/26/97
to

Tom C. Kerrigan (kerr...@merlin.pn.org) wrote:
: Robert Hyatt (hy...@cis.uab.edu) wrote:

: > Because you posted results against a "known" opponent, when, in fact, the
: > results were not against a "known opponent". You have apparently compiled

: I get it! Your entire problem with what I posted is that I called your
: program Crafty. Next time I post such games (if there is a next time) I'll
: be sure to just call it "some program Hyatt wrote." You could have just
: explained this earlier, instead of going off about how I screwed
: something up...

Nope. you still don't "get it"... What got my "attention" was the claim
that you were beating crafty *every* game... as in your original post. ANd
then you gave the details:

1. your own book. without crafty's "start book" for whatever reason, so that
it got into positions that it didn't like in every game in the group you
posted. What's the point in doing this? the simple start book is trivial
to build, start.pgn (input for this) is a tiny file. It's explained fully in
the read.me and crafty.doc files... So playing with Crafty using a funky book
doesn't make for a "real good" test case.

2. Crafty seems broken, in that the speed you quote for a P5/100 is simply
way too slow. If it's a crappy machine, that's one thing. If it is a normal
P5/100, then something you did in building crafty resulted in a 2x speed
penalty. In tests I run here, giving one version 2x the time of another will
produce really lop-sided results... Another problem then that makes your data
hard to understand.

3. Other parameters, such as transposition table size, pawn hash table size,
and so forth were equal? or default? Crafty defaults to a very small size by
request of the macintosh users. 2 hour games with 16,000 hash table entries
doesn't sound so hot.

So, what "bothered" me was a post of a "definitive" match, where it seems clear
to me that somethings not as it should be. Lonnie plays lots of standard games
vs Crafty... He doesn't win 'em all using any of the programs he has, on equal
machines. I can dig up his record vs crafty, although it's impossible to figure
out what he was running when. Bottom line, however, is either you are way better
than any commercial or amateur program, or your data's bad. While I happen to
think you've done great work on Stobor, I don't think that at present (Deep Blue
excepted) that *any* program I've seen is way out in front of anybody else. I'm
willing to go toe-to-toe with them. Bruce is. Steffan is... Kittinger is, and
so forth.

I simply want a claim that doesn't sound like the current claim Vincent is making,
yet that's exactly how it came across to me.

One nice thing about Crafty is that I can't fake any data, nor report on Crafty
doing something to genius, because I can't get away with it. Others have the
most recent version, and access to a P6/200, and can quickly figure out whether
what I claim is true or false. To keep it simple, I keep it honest. Before I'd
report such a result, I'd carefully check my data to be sure I didn't do something
that was unfair to Genius, or whomever. That's all I want to see... reasonable
data with reasonable accuracy. For example, let me compile crafty, and let me
compile stobor, and I'll play any time control you want, and take any odds you
want to give, because a poor compile can easily cost a factor of 2-3-4 or more.
My Sun CC with no -O option slows crafty down by a factor of 5 over full
optimization. That'd be a big speed handicap. And it wouldn't prove a thing
either...


chessman

unread,
Jan 26, 1997, 3:00:00 AM1/26/97
to


just getting back to the original post with my comments:


On 18 Jan 1997 16:28:09 +0100, kerr...@merlin.pn.org (Tom C.
Kerrigan) wrote:

>A *lot* of people have asked me for these, so I'll post what I have. A few
>things to note:

who are these *people* ....a lot ?

>
>* Stobor is always white (it plays black starting tonight...)
>* Stobor always wins, except for game 4, which was a draw

~~~~~~~~~~~~~~~~~~
inflammatory - stated as fact and truly has its roots the my * is
bigger than yours genre

>* time control is 40 moves/2 hours, then sudden death in an hour
>* hash tables are 2mb
>* pondering is off
>* books are generated from world champions' PGN archives on the ftp site

at the public crafty site..games are from the pitt.ftp site ..which at
the time and probably still does contain thousands of *trash* games
..false GM games and in some cases a complete collection of games
from the Ohio area of one player moving up from 1400 or so to 1900 or
so ..over 1000..not exactly the best set of games to make a book from.
I have provided Bob with a *clean* copy of GM games derived from the
FTP site that I have asked him not distribute since the amount of time
and effort to clean up these games were quite extensive..Bob has
respectfully honored my request.


>* Crafty's book learning is on (Stobor has "Tom Learning" <grin>)
>* machine is a p5/100
>* if you notice Stobor doing funky things, feel free to tell me :)
>

the <grin> is accurate ...your post misleading and, quite rightfully,
is recognized as such....more appropriately posted in
rec.games.chess.computer.trashtalk

cheers

Tom C. Kerrigan

unread,
Jan 26, 1997, 3:00:00 AM1/26/97
to

brucemo (bru...@nwlink.com) wrote:

> The problems that I've seen have tended to start with Kerrigan making a
> very authoritative-sounding statement that makes Bob furious. Bob replies
> very vehemently and Kerrigan replies with all possible violence.

Yes, you have criticized me for not giving people a graceful way out of an
argument, and I should have taken this to heart. Bummer. Maybe I'll do
better next time.

Cheers,
Tom

Tom C. Kerrigan

unread,
Jan 26, 1997, 3:00:00 AM1/26/97
to

Robert Hyatt (hy...@cis.uab.edu) wrote:

> I simply want a claim that doesn't sound like the current claim Vincent is making,
> yet that's exactly how it came across to me.

Oh, fuck the damn claim.

*I AM NOT MAKING A CLAIM.*

You have invented your own claim and attributed it to me.

People who read this thread ask for games. I post games. What is so bloody
hard to understand about this?

If you want me to tell the world that I have no idea how Stobor would do
against an 'official' Crafty, on ICC or in a tournament:

I have no idea how Stobor would do against an 'official' Crafty, on ICC or
in a tournament.

Probably much worse than 4.5/5, is my guess.

I have said all of this before, in one form or another. This post just has
swearing and capital letters. I hope these make my point clearer. I have
no desperate need to continue repeating myself, so don't expect replies to
anything else you post to this thread.

(Actually, I would stop reading this thread altogether, but Chris makes it
worthwhile... :)

Cheers,
Tom

Tom C. Kerrigan

unread,
Jan 26, 1997, 3:00:00 AM1/26/97
to

Chris Whittington (chr...@demon.co.uk) wrote:

> Weren't Hyatt and Kerrigan the two guys who got themselves into the 'my
> dick's bigger than your dick' argument last time.
> And I jest not, it really was about phallus size :)

Yes, after a rather drawn-out argument here, Hyatt send me a message on
ICC telling me that he's some huge-ass bodybuilder Karate teacher guy, and
proceeded to censor me on every Internet forum... I think my response was
nearly as immature... :)

Cheers,
Tom

Tom C. Kerrigan

unread,
Jan 26, 1997, 3:00:00 AM1/26/97
to

brucemo (bru...@nwlink.com) wrote:

> Tom C. Kerrigan wrote:
> > Bummer. I know this is a sucky p5/100, but I didn't realize it was that
> > sucky. Unfortunately, with less than a month left until the tournament, I
> > feel no pressing need to optimize somebody else's program... Perhaps after
> > February...
> Or it could be the computer itself. I have a Zeos Pantera P5/90. It's a
> reasonable machine for most things. It runs other chess programs nice and
> fast. But it can't run mine for beans. I tried three different operating
> systems, numerous hash table settings, and several compilers, and always
> slow, slow, slow.

Actually, thinking about this a bit more, it's probably the machine. I
assume what Hyatt uploads is what he uses, and I think we still use
basically the same operating system, so theoretically the problem is not
software related. I also know that my neighbor's four year old p5/90 runs
Stobor a solid bit faster than this p5/100. C'est la vie.

> Additionally, I don't understand why you need to make rude comments about
> Crafty, and by extension, about Bob. Your comment absolutely drips sarcasm

Damn... I think this is another misunderstanding... What I meant was not
that I would actually edit the source code to improve the program, but
instead that I won't waste time tinkering with Makefiles and whatever else
to see if I can't get it to run as fast as it "should." As far as I know,
I have a perfectly functional version of Crafty that simply runs a bit
slow, and I don't need anything beyond that for what I'm doing now.

> It's even worse because Bob has helped you design your program, and this is
> the kind of thing you give back.

I have a love-hate relationship with Bob. Yes, he did help me with my
program at the beginning, and I appreciate that, but I get no respect from
him on here and that makes for a very unfortunate situation, socially.

Our last fight was when he decided I needed a programming lecture and gave
it to me in less than polite terms. I pointed out he was off-topic, and it
turned out he wasn't even sure quite what the topic was (he missed the
post that started the thread). I got quite a bit of e-mail from people who
thought he was being an asshole, too, so I'm certain this was not entirely
my fault (although I should have dropped out of the fight much, much
earlier).

In fact, his replies to a few of my more recent posts to different threads
show he only reads enough of them to find faults.

I post games here and instead of telling me, in civil terms, how I could
"optimize" Crafty, or pointing out that "crafty" on ICC is much stronger,
he says he doesn't "believe" the games. He's accusing me of lying, while
ignoring the intent of the post and the fact that I have no motive to lie.

How the *hell* am I supposed to get along with somebody who behaves like
this??

My current policy of not replying to inflammatory posts might work. I'll
just go back to answering questions when I can.

Cheers,
Tom

Chris Whittington

unread,
Jan 26, 1997, 3:00:00 AM1/26/97
to

--
http://www.demon.co.uk/oxford-soft

brucemo <bru...@nwlink.com> wrote in article <32E92A...@nwlink.com>...


> Chris Whittington wrote:
>
> > Weren't Hyatt and Kerrigan the two guys who got themselves into the 'my
> > dick's bigger than your dick' argument last time.
> >
> > And I jest not, it really was about phallus size :)
>

> Yes, they were involved, but it was a one-sided argument from Kerrigan, I

> believe.
>

> The problems that I've seen have tended to start with Kerrigan making a
> very authoritative-sounding statement that makes Bob furious.

This wasn't how it started last time.

Kerrigan (who is 17 years old) said something Bob didn't like, and Bob (who
is a 47 year old professor) made a very disparaging remark, absolutely
guaranteed to set a 17 year-old (with more testosterone than he knows what
to do with) into a rage.

Bob knew it, it was calculated. Kerrigan fell into the trap.

Someone can go and get the posts back from Deja Vu. I thought Bob was being
highly manipulative, just to get Kerrigan to make a prat of himself.

> Bob replies
> very vehemently and Kerrigan replies with all possible violence.
>

> I usually agree with Bob, although I wouldn't write the replies quite the

> way he does. Some people may have the tendency to applaud Tom, though.
I
> often write replies that are essentially "shut up, Tom", although I
> generally have the sense to discard them before sending.
>
> It's interesting the way we have alliances in here. I usually lump you
> and Thorsten together,

Yes, it has been noticed :)

>
> although I do realize that he has a lot more snakes
> and lizards and lightning bolts coming out of his ears than you do.

I usually lump you and the ICCA together, although I do realise that
basically you are well meaning, try to uncover the truth, and use logical
thought and arguments, as opposed to human emotionality, leading almost
invariably to unsound conclusions :)

Chris Whittington


mclane

unread,
Jan 26, 1997, 3:00:00 AM1/26/97
to

kerr...@merlin.pn.org (Tom C. Kerrigan) wrote:

>Robert Hyatt (hy...@cis.uab.edu) wrote:

>Cheers,
>Tom
Come on folks. I will play some games versus stobor and crafty !
Calm down both.

Stobor will kill crafty , but no reason to fight that tough here. Let
the programs fight.


mclane

unread,
Jan 26, 1997, 3:00:00 AM1/26/97
to

kerr...@merlin.pn.org (Tom C. Kerrigan) wrote:

>Chris Whittington (chr...@demon.co.uk) wrote:

>> Weren't Hyatt and Kerrigan the two guys who got themselves into the 'my
>> dick's bigger than your dick' argument last time.
>> And I jest not, it really was about phallus size :)

>Yes, after a rather drawn-out argument here,

>Hyatt send me a message on
>ICC telling me that he's some huge-ass bodybuilder Karate teacher guy, and
>proceeded to censor me on every Internet forum...

What can a karate guy do against a aikido-guy ?? Nothing.
He can kick, he can fight with whatever, he only gets tired against
the defense of the aikido. So Tom, no problem. No karate can ever hurt
an aikido guy.

Robert Hyatt

unread,
Jan 27, 1997, 3:00:00 AM1/27/97
to

Tom C. Kerrigan (kerr...@merlin.pn.org) wrote:
: Chris Whittington (chr...@demon.co.uk) wrote:

: > Weren't Hyatt and Kerrigan the two guys who got themselves into the 'my
: > dick's bigger than your dick' argument last time.
: > And I jest not, it really was about phallus size :)

: Yes, after a rather drawn-out argument here, Hyatt send me a message on
: ICC telling me that he's some huge-ass bodybuilder Karate teacher guy, and

: proceeded to censor me on every Internet forum... I think my response was
: nearly as immature... :)

: Cheers,
: Tom

And I'm going to do it again. Post your drivel. I hate to let outlandish
statements stand, but in your case, I'll make an exception. Hopefully before you
either get everyone else here pissed, or they discover that a good bit of what
you have to say is "less than scientific," you'll figure out how to conduct
yourself and live in cyberspace.

For the record, the "karate nonsense quote" originated in the following way:

You first intimated that I had a poor classroom demeanor, and that it was likely
that a student would take issue with the way I address the class and "deck me"
(quoted words yours.) I responded that (a) I'd gotten good reviews in teaching
for over 20 years, (b) students in this part of the country are brought up a
little better at home and don't think about "decking" instructors; and (c) but,
if the inclination was to arise, I might not be the best candidate to deck. I'm
not a bodybuilding nut, but I do lift weights, and I did teach karate for a long
time. Long enough to not be concerned by loudmouthed punks, anyway.... And every
bit of the above explanation was sent to you via private email... you were the one
that saw fit to post it, which is pretty typical behavior for you. Ah well, I've
learned yet another valuable lesson...

the last word is yours, use it wisely...

Robert Hyatt

unread,
Jan 27, 1997, 3:00:00 AM1/27/97
to

Chris Whittington (chr...@demon.co.uk) wrote:

: --
: http://www.demon.co.uk/oxford-soft

: brucemo <bru...@nwlink.com> wrote in article <32E92A...@nwlink.com>...

: > Chris Whittington wrote:
: >
: > > Weren't Hyatt and Kerrigan the two guys who got themselves into the 'my
: > > dick's bigger than your dick' argument last time.
: > >
: > > And I jest not, it really was about phallus size :)
: >

: > Yes, they were involved, but it was a one-sided argument from Kerrigan, I

: > believe.
: >
: > The problems that I've seen have tended to start with Kerrigan making a
: > very authoritative-sounding statement that makes Bob furious.

: This wasn't how it started last time.

: Kerrigan (who is 17 years old) said something Bob didn't like, and Bob (who
: is a 47 year old professor) made a very disparaging remark, absolutely
: guaranteed to set a 17 year-old (with more testosterone than he knows what
: to do with) into a rage.

: Bob knew it, it was calculated. Kerrigan fell into the trap.

: Someone can go and get the posts back from Deja Vu. I thought Bob was being
: highly manipulative, just to get Kerrigan to make a prat of himself.

In that case, you are simply mistaken. I don't have enough hours in a day to
worry about how to manipulate people. You want my opinion, you get it,
forthright, clear and in unmistakable terms. As I've said many times, my
opinion *is* just an opinion.

If you think I started the original fracus, re-read everything from Deja-News.
It started with someone wanting a generic interface for everything so that it
would be possible to plug in a different eval, or a different search, or a
different anything and try new ideas. I responded that this was a classic
software engineering idea, although most commercial programmers wouldn't be
interested because they aren't interested in releasing the source. Tom's
response was that the idea was stupid. I pointed him to a couple of classic
Software Engineering texts to see why the idea (code reuse by name) isn't
stupid, and it went downhill from there.

This last started by the "stobor won *every game* except for one draw, which
was simply at least a "tad" optimistic from what I know about both programs.
The details he posted further supported that something was wrong, and anyone
that watches Crafty play on ICC could tell from the games posted that something
was grossly wrong book-wise. Lang used to play c6 quite a bit and Genius was
quite good at the opening. There are c6 lines that Crafty likes, but there
are also many that it hates. Adding all this up simply screamed (to me, at
least) that this wasn't "crafty" as I know it. Nothing more, nothing less.
If you noticed, after you and I played our match, I didn't rant and rave about
that, although I could certainly expect that CSTal was as good as it could be at
that point in time. But if you sent me a copy to test, and I made my own book
and so forth, I would *not* post those results. I might send them to you,
assuming you wanted them, but that's it.

I'll play anyone, anytime. If Crafty loses, it loses. However, I'd at least
like to be sure it was "crafty" that lost... That was my "point." Not to
manipulate, not to inflame, but to "correct."

: > Bob replies

brucemo

unread,
Jan 27, 1997, 3:00:00 AM1/27/97
to

Chris Whittington wrote:

> brucemo <bru...@nwlink.com> wrote in article <32E92A...@nwlink.com>...

> Kerrigan (who is 17 years old) said something Bob didn't like, and Bob (who


> is a 47 year old professor) made a very disparaging remark, absolutely
> guaranteed to set a 17 year-old (with more testosterone than he knows what
> to do with) into a rage.
>
> Bob knew it, it was calculated. Kerrigan fell into the trap.

This may be approximately as accurate as my depiction. I don't think Bob is
out to get Kerrigan, but I have been surprised by the degree to which he takes
him seriously -- his initial responses to Kerrigan posts have been more violent
than you'd expect between any two random people on here.

> I usually lump you and the ICCA together, although I do realise that
> basically you are well meaning, try to uncover the truth, and use logical
> thought and arguments, as opposed to human emotionality, leading almost
> invariably to unsound conclusions :)

I can see how you could arrive at this, no problem.

To be absolutely clear, I am a member of the ICCA, but I am not (nor have I
ever been) an officer of the ICCA, nor am I otherwise officially or
unofficially attached to them.

I figure you know this, but perhaps some others don't.

bruce

brucemo

unread,
Jan 27, 1997, 3:00:00 AM1/27/97
to

Tom C. Kerrigan wrote:
>
> brucemo (bru...@nwlink.com) wrote:
>
> > The problems that I've seen have tended to start with Kerrigan making a
> > very authoritative-sounding statement that makes Bob furious. Bob replies

> > very vehemently and Kerrigan replies with all possible violence.
>
> Yes, you have criticized me for not giving people a graceful way out of an
> argument, and I should have taken this to heart. Bummer. Maybe I'll do
> better next time.

Actually I think you arrived at this conclusion on your own.

bruce

Don Fong

unread,
Jan 27, 1997, 3:00:00 AM1/27/97
to

In article <5chgr4$i...@juniper.cis.uab.edu>,

Robert Hyatt <hy...@cis.uab.edu> wrote:
>Tom C. Kerrigan (kerr...@merlin.pn.org) wrote:
>: Yes, after a rather drawn-out argument here, Hyatt send me a message on
>: ICC telling me that he's some huge-ass bodybuilder Karate teacher guy, and
>: proceeded to censor me on every Internet forum... I think my response was
>: nearly as immature... :)

to Tom: i think your first mistake was posting to rgcc about
some silly exchange that occurred on ICC. that could be considered
a provocation.

[...]


>For the record, the "karate nonsense quote" originated in the following way:

to Bob: as i recall, you'd said several times that in your classroom,
a student who talked like Tom would be ``out on his ear'' in a flash.
that could be considered a provocation.

to both Tom and Bob: you both tend to make boastful but imprecise
statements that beg to be misinterpreted. and you both escalate by
choosing the most inflammatory interpretation of what the other said.
that is my analysis.

--
don fong ``i still want the peace dividend'' http://got.net/~dfong/

Tom C. Kerrigan

unread,
Jan 27, 1997, 3:00:00 AM1/27/97
to

chessman (ches...@voicenet.com) wrote:
> >A *lot* of people have asked me for these, so I'll post what I have. A few
> >things to note:
> who are these *people* ....a lot ?

People who you probably don't know and who I don't care to list. I'd be
glad to send you a few examples via e-mail, but I doubt they want their
names all over this.

> >* Stobor is always white (it plays black starting tonight...)
> >* Stobor always wins, except for game 4, which was a draw
> ~~~~~~~~~~~~~~~~~~
> inflammatory - stated as fact and truly has its roots the my * is
> bigger than yours genre

Just as inflammatory as "Stobor is always white." I was obviously
referring to the games. Notice nobody flamed me because Crafty is capable
of playing white, too.

> >* books are generated from world champions' PGN archives on the ftp site
> at the public crafty site..games are from the pitt.ftp site ..which at
> the time and probably still does contain thousands of *trash* games

I won't argue with this. I'm sure whatever is at Bob's site is much
better, but getting access to Bob's site is fairly inconvienant for me. I
will point out that both programs generated opening books from the same
PGN wad. I will also point out that after getting out of the book, neither
program thought it was down by more than half a pawn, and this has held
for 15 games so far. If the programs were getting out of the book a piece
down, then the games wouldn't be worth anything to anybody and I certainly
wouldn't have posted any of them.

> the <grin> is accurate ...your post misleading and, quite rightfully,
> is recognized as such....more appropriately posted in

It wasn't intended to be. I hope this clears things up.

Cheers,
Tom

Chris Whittington

unread,
Jan 27, 1997, 3:00:00 AM1/27/97
to

--
http://www.demon.co.uk/oxford-soft

Tom C. Kerrigan <kerr...@merlin.pn.org> wrote in article
<5ce7ok$q...@merlin.pn.org>...

No, no. Don't stop.

Kerrigan-Hyatt flame wars are great entertainment.

We all just love watching you two making total dicks of yourselves.

Chris Whittington

>
> Cheers,
> Tom
>

Peter Kappler

unread,
Jan 28, 1997, 3:00:00 AM1/28/97
to

On 26 Jan 1997 01:16:20 +0100, kerr...@merlin.pn.org (Tom C.
Kerrigan) wrote:

>Robert Hyatt (hy...@cis.uab.edu) wrote:
>
>> I simply want a claim that doesn't sound like the current claim Vincent is making,
>> yet that's exactly how it came across to me.
>
>Oh, fuck the damn claim.
>
>*I AM NOT MAKING A CLAIM.*
>
>You have invented your own claim and attributed it to me.
>
>People who read this thread ask for games. I post games. What is so bloody
>hard to understand about this?
>
>If you want me to tell the world that I have no idea how Stobor would do
>against an 'official' Crafty, on ICC or in a tournament:
>
>I have no idea how Stobor would do against an 'official' Crafty, on ICC or
>in a tournament.
>
>Probably much worse than 4.5/5, is my guess.
>

Yes, much worse, at least at blitz time controls.

A quick search of the ICC database produced 24 games played between
Prisma and Stobor on 1/4/97.

According to the ICC 'finger notes':
Prisma = Crafty running on a P6/200, 64 MB RAM
Ratbert = Stobor running on a P6/200, 64 MB RAM

So we have a contest between Stobor and Crafty, with each program
running on the same hardware.

In 21 blitz games (played at time controls of 5 0, and 3 0) Crafty
scored 62% (+10, -6, =5). In 3 bullet games (time control = 1 1)
Crafty won 2 and lost 1.


Ratbert hasn't played any ICC games since Jan 4, so I suppose it's
possible that you made significant improvements between Jan 4, and the
Jan 18 ( the date this thread began).


However, it seems that these data are more meaningful (larger sample
size, equal hardware, and a public playing environment) than the
results you've been posting.

Peter

Tom C. Kerrigan

unread,
Jan 28, 1997, 3:00:00 AM1/28/97
to

Peter Kappler (pe...@bitsource.com) wrote:

> Ratbert hasn't played any ICC games since Jan 4, so I suppose it's
> possible that you made significant improvements between Jan 4, and the
> Jan 18 ( the date this thread began).

Or between August 20, 1996 and January 18, 1997. In other words, a third
of the life of the program.

Perhaps it would do you good to read the other posts to this thread.

BTW, I would be surprised if Crafty (with book learning) didn't bust
something after that many games.

Cheers,
Tom

Harald Faber

unread,
Jan 30, 1997, 3:00:00 AM1/30/97
to

quoting a mail from mclane # prima.ruhr.de

Hello mclane,


m> From: mcl...@prima.ruhr.de (mclane)
m> Subject: Re: Stobor vs. Crafty
m> Organization: Prima e.V. Dortmund


m> What can a karate guy do against a aikido-guy ?? Nothing.
m> He can kick, he can fight with whatever, he only gets tired against
m> the defense of the aikido. So Tom, no problem. No karate can ever hurt
m> an aikido guy.

:-)))
You pretend to know much but you have no idea of karate, aikido, jiu-jitsu
or whatswhoever. :-)

I tell you the faster one wins.
It is much harder for an aikido-fighter to react because he learns too
many techniques to defend so he has problems to automatize (right word?)
his defence.

However this is far beyond on-topic threads... :-)

Harald
--

0 new messages