Bopo v1.0 , Release

Daniel Shawul · Post by **Daniel Shawul** » Mon Oct 01, 2012 3:22 am

The best rating system has just been released! Download quick before link expires https://dl.dropbox.com/u/55295461/Bopo.zip
It calculates ratings like never before. The algorithms are kept secret for now but will do so if there is enough pressure to do so. Due to popular request scale has been set to 1 by default. Lets face it people want to see ratings they like. It is stupid to scale. In the future I will add features that will blow every rating tool on its site. Some of planned things include some markov chain monte-carlo sampler for calculating better error margins.
The Bopo team.
Yes I have a tester that does my PRing but decided not to mention his name.
Enjoy and send me results to bopo194@yahoo.com.

ZirconiumX · Post by **ZirconiumX** » Mon Oct 01, 2012 7:43 am

It isn't April 1st yet.

Matthew:out

Graham Banks · Post by **Graham Banks** » Mon Oct 01, 2012 9:05 am

Daniel Shawul wrote:The best rating system has just been released! Download quick before link expires https://dl.dropbox.com/u/55295461/Bopo.zip
It calculates ratings like never before. The algorithms are kept secret for now but will do so if there is enough pressure to do so. Due to popular request scale has been set to 1 by default. Lets face it people want to see ratings they like. It is stupid to scale. In the future I will add features that will blow every rating tool on its site. Some of planned things include some markov chain monte-carlo sampler for calculating better error margins.
The Bopo team.
Yes I have a tester that does my PRing but decided not to mention his name.
Enjoy and send me results to bopo194@yahoo.com.

I think you're being a bit harsh Daniel.

Daniel Shawul · Post by **Daniel Shawul** » Mon Oct 01, 2012 2:21 pm

The email that I posted is wrong. It is bopo194@ymail.com . We apologize for those of you who sent your comments at the @yahoo.com address.
The Bopo team

Vinvin · Post by **Vinvin** » Mon Oct 01, 2012 6:29 pm

Daniel Shawul wrote:The best rating system has just been released! Download quick before link expires https://dl.dropbox.com/u/55295461/Bopo.zip
It calculates ratings like never before. The algorithms are kept secret for now but will do so if there is enough pressure to do so. Due to popular request scale has been set to 1 by default. Lets face it people want to see ratings they like. It is stupid to scale. In the future I will add features that will blow every rating tool on its site. Some of planned things include some markov chain monte-carlo sampler for calculating better error margins.
The Bopo team.
Yes I have a tester that does my PRing but decided not to mention his name.
Enjoy and send me results to bopo194@yahoo.com.

What are the commands ?
I try "read" but il allocates 2GB and crash ...

Code: Select all

g:\ratinglist\2012septembre>bopo
$: read cwarout.pgn
Games 66296

This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.

Daniel Shawul · Post by **Daniel Shawul** » Mon Oct 01, 2012 7:51 pm

Vinvin wrote:
Daniel Shawul wrote:The best rating system has just been released! Download quick before link expires https://dl.dropbox.com/u/55295461/Bopo.zip
It calculates ratings like never before. The algorithms are kept secret for now but will do so if there is enough pressure to do so. Due to popular request scale has been set to 1 by default. Lets face it people want to see ratings they like. It is stupid to scale. In the future I will add features that will blow every rating tool on its site. Some of planned things include some markov chain monte-carlo sampler for calculating better error margins.
The Bopo team.
Yes I have a tester that does my PRing but decided not to mention his name.
Enjoy and send me results to bopo194@yahoo.com.
What are the commands ?
I try "read" but il allocates 2GB and crash ...
Code: Select all
g:\ratinglist\2012septembre>bopo
$: read cwarout.pgn
Games 66296

This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.

Bug reports are welocome at bopo194@ymail.com. We have run CCRL 40/40 and blitz which have more than 1 million games in them with no problem, so it should handle cwarout.pgn that has much less games. More help can be found from the same place you found about 'read' for now. That is all we have so far but we will add documentation later. Our goal is to become the first commercial rating system.
The bopo team

Ajedrecista · Post by **Ajedrecista** » Mon Oct 01, 2012 9:18 pm

Hello Daniel:

Thanks for your effort. I have tried results.pgn file (the one that is included in Ordo 0.6 download, with 161200 games) and everything seems fine here. You are right when you say that a little more documentation is needed. Here is a typo:

Code: Select all

cg         -- compute ratings uing conjugate gradient method

It has a very simple correction: using instead of uing.

You may also provide info about plotstat and cross commands, and also about elodrawslope and likelihood:

Code: Select all

$: likelihood
-356681.152410
$: elodrawslope
1.0000
$: cross
Section 1:
83. 7.639132e-006
0. 85.051195
1. 34.926979
2. 70.403447
3. 24.526243
4. 69.778514
5. 27.552308
6. 88.511461
7. 31.216250
8. 41.695428
9. 32.470741
10. 31.985416
11. 59.471258
12. 61.700888
13. 39.790684
14. 51.487159
15. 40.279251
16. 81.726082
17. 119.833791
18. 68.952446
19. 133.219026
20. 89.383539
21. 114.974896
22. 128.558585
23. 60.442447
24. 120.882308
25. 99.582630
26. 59.725154
27. 136.270146
28. 61.629273
29. 104.661825
30. 93.769824
31. 50.496135
32. 175.882258
33. 94.370114
34. 80.301800
35. 72.037954
36. 146.628085
37. 86.504900
38. 75.343896
39. 39.379298
40. 115.803002
41. 84.854385
42. 49.270708
43. 106.877147
44. 132.888183
45. 145.969086
46. 163.541405
47. 161.214054
48. 28.173614
49. 81.406466
50. 52.188872
51. 76.777776
52. 81.340844
53. 168.427666
54. 100.854154
55. 84.623348
56. 96.249085
57. 97.686792
58. 161.305347
59. 198.088155
60. 82.525507
61. 56.193211
62. 55.348610
63. 94.046418
64. 79.494712
65. 54.138663
66. 67.625973
67. 46.324621
68. 70.788386
69. 68.504988
70. 63.342182
71. 49.199403
72. 70.981871
73. 52.940141
74. 67.948243
75. 77.852459
76. 56.602591
77. 63.064419
78. 56.960036
79. 59.676637
80. 66.687320
81. 86.654566
82. 65.474000
83. 70.084280
84. 96.499249
85. 67.600414
86. 78.613618
87. 81.325365
88. 95.486220
89. 100.168470
90. 84.510805
91. 86.957853
92. 104.465273
93. 108.517415
94. 111.487512
Time 0.09 sec
Weighted mean square error: 0.000833
Pearson's chi squared: 404.580457
LogLikelihood: -153158.197187
Section 2:
83. 8.077728e-006
0. 86.856386
1. 35.117386
2. 70.777198
3. 24.435229
4. 70.266694
5. 27.371120
6. 84.077339
7. 31.933105
8. 42.136084
9. 32.340499
10. 32.099611
11. 60.481809
12. 63.136320
13. 40.044470
14. 52.441834
15. 40.575004
16. 83.012635
17. 117.745765
18. 68.680939
19. 133.526938
20. 87.588187
21. 115.158862
22. 127.214737
23. 59.767691
24. 119.340542
25. 101.233630
26. 62.552793
27. 137.025813
28. 63.139254
29. 102.722276
30. 93.611707
31. 50.728748
32. 177.720126
33. 96.523646
34. 79.989410
35. 72.276300
36. 145.630153
37. 87.762866
38. 73.920898
39. 39.229885
40. 114.048958
41. 84.757177
42. 49.549770
43. 108.895704
44. 135.127520
45. 142.226192
46. 160.040582
47. 163.645644
48. 28.203163
49. 82.324742
50. 52.177007
51. 75.249370
52. 82.523930
53. 169.259246
54. 99.973383
55. 85.105301
56. 98.779309
57. 99.411247
58. 160.874834
59. 199.304980
60. 80.699538
61. 55.653648
62. 54.695237
63. 96.558292
64. 81.022375
65. 53.620880
66. 66.390488
67. 46.372941
68. 71.590761
69. 69.617492
70. 64.388849
71. 50.154145
72. 74.531332
73. 52.869699
74. 68.098953
75. 79.976587
76. 56.751446
77. 63.091991
78. 56.965875
79. 59.099115
80. 65.933556
81. 84.764291
82. 66.107384
83. 68.320257
84. 96.430077
85. 68.383339
86. 80.753971
87. 77.532145
88. 91.886225
89. 102.939839
90. 85.309570
91. 85.565292
92. 105.739706
93. 106.192370
94. 110.801720
Time 0.06 sec
Weighted mean square error: 0.000634
Pearson's chi squared: 339.090810
LogLikelihood: -153359.248888
$: plotstat
 -573       7       9     100       5       7
 -543       5      11     100       6       8
 -538       7       8     200       6       9
 -533       2       4     200       6       9
 -528       6      10     100       6       9
 -523       9      15     100       6       9
 -518      10      10     100       6       9
 -513       4       9     100       7      10
 -498       8      12     200       7      11
 -493      13      18     100       7      11
 -488       9      17     100       8      11
 -478       8      13     400       8      12
 -473       5       7     100       8      12
 -468       9      12     500       8      12
 -463       9      14     300       9      12
 -458       8      14     400       9      13
 -453       9      14     600       9      13
 -448       9      14     300       9      13
 -443       9      15     400      10      14
 -438      10      14     700      10      14
 -433       9      14     400      10      14
 -428      11      17     500      10      15
 -423       9      15     500      11      15
 -418      10      19     300      11      15
 -413      13      20     400      11      16
 -408      12      17     500      11      16
 -403      13      20     200      12      17
 -398      13      20     200      12      17
 -393      12      18     600      12      17
 -388      13      24     300      13      18
 -383      14      20     700      13      18
 -378      14      20     600      13      18
 -373      12      20    1100      13      19
 -368      13      21     600      14      19
 -363      15      23     500      14      20
 -358      14      23     700      14      20
 -353      15      22     900      15      20
 -348      14      21     700      15      21
 -343      15      22     700      15      21
 -338      15      24     500      16      22
 -333      17      26     700      16      22
 -328      15      24     800      17      23
 -323      18      24     600      17      23
 -318      18      24     700      17      23
 -313      18      27     700      18      24
 -308      19      26    1000      18      24
 -303      20      26    1000      18      25
 -298      19      28    1400      19      25
 -293      19      28    1300      19      26
 -288      18      25    1000      20      26
 -283      19      26    1600      20      27
 -278      20      27    1200      20      27
 -273      23      30    1200      21      27
 -268      21      31    1000      21      28
 -263      21      28    1100      22      28
 -258      21      30    1300      22      29
 -253      21      30    1200      23      29
 -248      23      30    2000      23      30
 -243      24      30    1200      24      30
 -238      23      33    2600      24      31
 -233      24      31    1400      24      31
 -228      24      32    1700      25      32
 -223      25      31    1300      25      32
 -218      24      34    1900      26      32
 -213      25      36    1500      26      33
 -208      27      34    1800      27      33
 -203      26      34    2300      27      34
 -198      28      35    1800      28      34
 -193      27      34    2200      28      35
 -188      28      36    2200      29      35
 -183      28      38    2300      29      36
 -178      29      36    2000      30      36
 -173      29      35    2000      30      36
 -168      31      38    2300      31      37
 -163      31      36    1700      31      37
 -158      32      37    2200      32      38
 -153      32      38    2400      32      38
 -148      32      36    2400      33      38
 -143      33      37    2400      34      39
 -138      33      37    3200      34      39
 -133      35      38    2300      35      39
 -128      34      37    2300      35      40
 -123      34      38    2500      36      40
 -118      36      38    2000      36      40
 -113      37      38    2100      37      41
 -108      38      41    2400      37      41
 -103      36      38    2600      38      41
  -98      38      37    1900      38      41
  -93      39      40    3300      39      42
  -88      38      36    2500      40      42
  -83      41      39    3500      40      42
  -78      40      42    2300      41      42
  -73      40      41    1900      41      43
  -68      42      39    2700      42      43
  -63      41      42    3200      42      43
  -58      42      38    3100      43      43
  -53      44      40    3900      44      43
  -48      43      42    2800      44      43
  -43      45      41    3400      45      44
  -38      45      42    2400      45      44
  -33      45      42    2800      46      44
  -28      45      40    4000      46      44
  -23      46      40    3700      47      44
  -18      48      44    2800      48      44
  -13      47      45    3600      48      44
   -8      49      42    4100      49      44
   -3      49      42    3500      49      44
    2      50      43    6000      50      44
    7      50      42    3500      50      44
   12      50      42    4100      51      44
   17      52      45    3600      52      44
   22      52      44    2800      52      44
   27      53      40    3700      53      44
   32      54      40    4000      53      44
   37      54      42    2800      54      44
   42      54      42    2400      54      44
   47      54      41    3400      55      43
   52      56      42    2800      56      43
   57      55      40    3900      56      43
   62      57      38    3100      57      43
   67      58      42    3200      57      43
   72      57      39    2700      58      43
   77      59      41    1900      58      42
   82      59      42    2300      59      42
   87      58      39    3500      60      42
   92      61      36    2500      60      42
   97      60      40    3300      61      41
  102      61      37    1900      61      41
  107      63      38    2600      62      41
  112      61      41    2400      62      41
  117      62      38    2100      63      40
  122      63      38    2000      63      40
  127      65      38    2500      64      40
  132      65      37    2300      65      39
  137      64      38    2300      65      39
  142      66      37    3200      66      39
  147      66      37    2400      66      38
  152      67      36    2400      67      38
  157      67      38    2400      67      37
  162      68      37    2200      68      37
  167      68      36    1700      68      37
  172      68      38    2300      69      36
  177      70      35    2000      69      36
  182      70      36    2000      70      35
  187      71      38    2300      70      35
  192      71      36    2200      71      35
  197      72      34    2200      71      34
  202      71      35    1800      72      34
  207      73      34    2300      72      33
  212      72      34    1800      73      33
  217      74      36    1500      73      32
  222      75      34    1900      74      32
  227      74      31    1300      74      31
  232      76      32    1700      75      31
  237      75      31    1400      75      31
  242      76      33    2600      76      30
  247      75      30    1200      76      30
  252      76      30    2000      76      29
  257      78      30    1200      77      29
  262      78      30    1300      77      28
  267      78      28    1100      78      28
  272      78      31    1000      78      27
  277      76      30    1200      79      27
  282      79      27    1200      79      26
  287      80      26    1600      79      26
  292      81      25    1000      80      25
  297      80      28    1300      80      25
  302      80      28    1400      81      25
  307      80      26    1000      81      24
  312      80      26    1000      81      24
  317      81      27     700      82      23
  322      81      24     700      82      23
  327      81      24     600      83      22
  332      84      24     800      83      22
  337      82      26     700      83      22
  342      84      24     500      84      21
  347      84      22     700      84      21
  352      85      21     700      84      20
  357      84      22     900      85      20
  362      85      23     700      85      19
  367      84      23     500      85      19
  372      86      21     600      86      19
  377      87      20    1100      86      18
  382      85      20     600      86      18
  387      85      20     700      87      17
  392      86      24     300      87      17
  397      87      18     600      87      17
  402      86      20     200      87      16
  407      86      20     200      88      16
  412      87      17     500      88      16
  417      86      20     400      88      15
  422      89      19     300      89      15
  427      90      15     500      89      15
  432      88      17     500      89      14
  437      91      14     400      89      14
  442      89      14     700      90      14
  447      91      15     400      90      13
  452      91      14     300      90      13
  457      90      14     600      90      13
  462      91      14     400      91      12
  467      90      14     300      91      12
  472      90      12     500      91      12
  477      94       7     100      91      11
  482      91      13     400      91      11
  492      90      17     100      92      11
  497      87      18     100      92      10
  502      91      12     200      92      10
  517      95       9     100      93       9
  522      90      10     100      93       9
  527      90      15     100      93       9
  532      94      10     100      93       9
  537      97       4     200      93       8
  542      92       8     200      94       8
  547      94      11     100      94       8
  577      92       9     100      95       7
Weighted mean square error: 0.000523
Pearson's chi squared: 503.972452
LogLikelihood: -306281.336534

If I use plotstat just after read, I get the following:

Code: Select all

$: read results.pgn
Games 161200
$: plotstat
    2      50      35   322400     50      27
Weighted mean square error: 0.006519
Pearson's chi squared: 7625.650440
LogLikelihood: -358687.505449

Which is somewhat different to the other run of plotstat. I only understand 322400 = 2*161200 in the second run.

I see a strange result with cg command:

Code: Select all

$: read results.pgn
Games 161200
$: cg
Iteration     0: -356927.674677 1.612000e+005 0.150000 {32.800000 1.000000 97.30
0000 1.000000}
Iteration    10: -343825.949547 3.809635e+002 0.150000 {-24.101644 -50.945642 13
7.561488 50.862331}
Iteration    20: -1.#IND00 1.#QNAN0e+000 0.150000 {-1.#IND00 -1.#IND00 -1.#IND00
 -1.#IND00}
Iteration    30: -1.#IND00 1.#QNAN0e+000 0.150000 {-1.#IND00 -1.#IND00 -1.#IND00
 -1.#IND00}
Iteration    40: -1.#IND00 1.#QNAN0e+000 0.150000 {-1.#IND00 -1.#IND00 -1.#IND00
 -1.#IND00}
Iteration    50: -1.#IND00 1.#QNAN0e+000 0.150000 {-1.#IND00 -1.#IND00 -1.#IND00
 -1.#IND00}
Iteration    60: -1.#IND00 1.#QNAN0e+000 0.150000 {-1.#IND00 -1.#IND00 -1.#IND00
 -1.#IND00}
Iteration    70: -1.#IND00 1.#QNAN0e+000 0.150000 {-1.#IND00 -1.#IND00 -1.#IND00
 -1.#IND00}
Iteration    80: -1.#IND00 1.#QNAN0e+000 0.150000 {-1.#IND00 -1.#IND00 -1.#IND00
 -1.#IND00}
Iteration    90: -1.#IND00 1.#QNAN0e+000 0.150000 {-1.#IND00 -1.#IND00 -1.#IND00
 -1.#IND00}
Iteration   100: -1.#IND00 1.#QNAN0e+000 0.150000 {-1.#IND00 -1.#IND00 -1.#IND00
 -1.#IND00}

I can only stop it closing Bopo. In iteration 0, I understand the number of games = 1.612e+5, alpha = 0.15, elohome = 32.8, scale = 1 and elodraw = 97.3; I also saw that the default prior (just after read command) is 2.

------------------------

Daniel Shawul wrote:Bug reports are welocome at bopo194@ymail.com. We have run CCRL 40/40 and blitz which have more than 1 million games in them with no problem, so it should handle cwarout.pgn that has much less games. More help can be found from the same place you found about 'read' for now. That is all we have so far but we will add documentation later. Our goal is to become the first commercial rating system.
The bopo team

If you are not kidding, I foresee few sales because BayesElo is great and also free! Just my humble opinion.

Good luck with Bopo development! Sorry for this very long post.

Regards from Spain.

Ajedrecista.

Daniel Shawul · Post by **Daniel Shawul** » Mon Oct 01, 2012 10:19 pm

Hi Jesuz:
Thanks for your interest. Are you reverse engineering our product ?

We appreciate efforts to help us develope the tool more. Note that this release came at a very unexpected time for us, so we have bugs everywhere. The Cg method used to work before but I was experimenting in using a linear variation for approximating elodraw and elohome, rather than assuming constant values. Unfortunately that fails to converge most of the time and I forgot to remove it. IIRC I reported about that some time ago.
I am in a bit of a rush now but I will look to your post in detail later and give you feed back.
You got the first t-shirt from the the bopo team.

Ajedrecista · Post by **Ajedrecista** » Mon Oct 01, 2012 10:28 pm

Hello again:

Daniel Shawul wrote:Hi Jesuz:
Thanks for your interest. Are you reverse engineering our product ? We appreciate efforts to help us develope the tool more. Note that this release came at a very unexpected time for us, so we have bugs everywhere. The Cg method used to work before but I was experimenting in using a linear variation for approximating elodraw and elohome, rather than assuming constant values. Unfortunately that fails to converge most of the time and I forgot to remove it. IIRC I reported about that some time ago.
I am in a bit of a rush now but I will look to your post in detail later and give you feed back.
You got the first t-shirt from the the bopo team.

Not reverse engineering at all! I simply changed the file extension from .exe to .txt:

Code: Select all

ren *.exe *.txt

Just writing this line in the command prompt, in the folder where Bopo.exe was. I simply took a look on readable things, that are too few by the way. Then, I re-changed the file extension using the same trick:

Code: Select all

ren *.txt *.exe

Each change takes less than a second, so it is little effort. Good luck with Bopo! Thanks in advance for your future feedback.

Regards from Spain.

Ajedrecista.

Vinvin · Post by **Vinvin** » Tue Oct 02, 2012 2:41 am

Daniel Shawul wrote:
Vinvin wrote:
Daniel Shawul wrote:The best rating system has just been released! Download quick before link expires https://dl.dropbox.com/u/55295461/Bopo.zip
It calculates ratings like never before. The algorithms are kept secret for now but will do so if there is enough pressure to do so. Due to popular request scale has been set to 1 by default. Lets face it people want to see ratings they like. It is stupid to scale. In the future I will add features that will blow every rating tool on its site. Some of planned things include some markov chain monte-carlo sampler for calculating better error margins.
The Bopo team.
Yes I have a tester that does my PRing but decided not to mention his name.
Enjoy and send me results to bopo194@yahoo.com.
What are the commands ?
I try "read" but il allocates 2GB and crash ...
Code: Select all
g:\ratinglist\2012septembre>bopo
$: read cwarout.pgn
Games 66296

This application has requested the Runtime to terminate it in an unusual way.
Please contact the application's support team for more information.
Bug reports are welocome at bopo194@ymail.com. We have run CCRL 40/40 and blitz which have more than 1 million games in them with no problem, so it should handle cwarout.pgn that has much less games. More help can be found from the same place you found about 'read' for now. That is all we have so far but we will add documentation later. Our goal is to become the first commercial rating system.
The bopo team

I tried with 2 other files, one crashed and the other is OK.
There should be a bug somewhere ...
There's the file with about 1.3M games who crashed bopo at the "read" command : http://home.scarlet.be/vincentlejeune/shrunkout.7z

Bopo v1.0 , Release

Bopo v1.0 , Release

Re: Bopo v1.0 , Release

Re: Bopo v1.0 , Release

Re: Bopo v1.0 , Release

Re: Bopo v1.0 , Release

Re: Bopo v1.0 , Release

Re: Bopo v1.0, release.

Re: Bopo v1.0, release.

Re: Bopo v1.0, release.

Re: Bopo v1.0 , Release