| View previous topic :: View next topic |
| Author |
Message |
Jesús Muñoz
Joined: 13 Jul 2011
Posts: 707
Location: Madrid, Spain.
|
 Posted: Sun Jan 15, 2012 12:27 pm Post subject: Doubt: perft estimate averaging N MonteCarlo samples. |
  |
|
Hello:
I have been thinking about an idea that will be surely clumsy: using 'perftmc' command of GNU 5.07.173b (which gives different values in each try), could I do a good estimate averaging results? Here is a typical output:
| Code: |
m=6.188431e+019 sd=5.191844e+015 ci(99%)=[6.187093e+019,6.189768e+019] n=2001154
971 sdn=2.322533e+020 t=37391.12s |
It is an example of MonteCarlo Perft(14) for the initial position of standard chess. I thought about using NebiyuChess 1.42, which automatically interrupts MonteCarlo Perft at 5e+8 nodes more less... but always gives the same results in different tries, so I discarded it. This is the reason why I thought about GNU.
My question is the following: if I run N MonteCarlo samples of the same position and I stop all of them at the same point (with the same number of nodes ± a very little amount), can I average the results in this way?
| Code: |
m_1, m_2, ..., m_N
(Average m) = <m> = (1/N)·(m_1 + m_2 + ... + m_N)
sd_1, sd_2, ... sd_N
(Average standard deviation) = <sd> = sqrt{(1/N)·[(sd_1)² + (sd_2)² + ... + (sd_N)²]} |
Or, taking into account that different samples have different number of nodes:
| Code: |
<m> = [(n_1)·(m_1) + (n_2)·(m_2) + ... + (n_N)·(m_N)]/(n_1 + n_2 + ... n_N)
<sd> = sqrt{[(sdn_1)² + (sdn_2)² + ... + (sdn_N)²]/(n_1 + n_2 + ... n_N)} |
Where sdn_i = (sd_i)·sqrt(n_i) and GNU prints these numbers. I think that n_i ~ 1e+8 or 5e+8 nodes in each sample is accurate enough, but comments are welcome.
Are my assumptions correct or I have failure concepts? In the case that I am right (very unlikely), the estimate would be good/accurate or is it a waste of time because of the size of the standard deviation (or other issues)? Thanks in advance.
Regards from Spain.
Ajedrecita.
_________________
Six Fortran 95 tools.
Chess will never be solved. |
|
| Back to top |
|
 |
Daniel Shawul
Joined: 14 Mar 2006
Posts: 2268
Location: Ethiopia
|
| Posted: Sun Jan 15, 2012 1:31 pm Post subject: Re: Doubt: perft estimate averaging N MonteCarlo samples. |
|
|
Hello Jesus
| Quote: |
I thought about using NebiyuChess 1.42, which automatically interrupts MonteCarlo Perft at 5e+8 nodes more less... but always gives the same results in different tries, so I discarded it. This is the reason why I thought about GNU.
|
You can set different random number seed before calling perft.
| Code: |
seed 0 => sets the seed to time(NULL) and prints the number used so that it can be used to debug that specific game (sequence of random numbers)
seed N => sets seed to that number N for N != 0 eg. 457632
|
I found reproduciblity very important in many situations, which is why i decided to make not setting to time a default.
regards,
Daniel
_________________
https://sites.google.com/site/dshawul/
https://github.com/dshawul |
|
| Back to top |
|
|
Jesús Muñoz
Joined: 13 Jul 2011
Posts: 707
Location: Madrid, Spain.
|
| Posted: Mon Jan 16, 2012 8:35 pm Post subject: Averages with 27 MonteCarlo samples. |
|
|
Hello:
| Daniel Shawul wrote: |
Hello Jesus
| Quote: |
I thought about using NebiyuChess 1.42, which automatically interrupts MonteCarlo Perft at 5e+8 nodes more less... but always gives the same results in different tries, so I discarded it. This is the reason why I thought about GNU.
|
You can set different random number seed before calling perft.
| Code: |
seed 0 => sets the seed to time(NULL) and prints the number used so that it can be used to debug that specific game (sequence of random numbers)
seed N => sets seed to that number N for N != 0 eg. 457632
|
I found reproduciblity very important in many situations, which is why i decided to make not setting to time a default.
regards,
Daniel |
Thanks for your tip: it works, as expected. I had to search 'random number seed' in Google for little clarification, but you explained it quite well.
I finally used GNU instead of Nebiyu for avoiding the change of seed each time (I suppose I would set seed 1, seed 2, ..., seed N) because is easier for me. But using GNU should not be an offence for you, because I stop each MonteCarlo sample of GNU with Nebiyu criterion: as soon as number of nodes are greater than 5e+8.
I average the data taking into account the number of nodes of each sample, that is the second way I explained in the initial post of this topic; as nobody has commented about the validity of my idea, I will use it without changes.
I have ran 27 samples until now. Here are the results (hoping no typos in the whole post):
| Code: |
perftmc 14 (GNU 5.07.173b w32):
1) m=6.187678e+019 sd=9.750798e+015 ci(99%)=[6.185166e+019,6.190189e+019] n=501279094 sdn=2.183132e+020 t=9565.06s
2) m=6.189193e+019 sd=1.051857e+016 ci(99%)=[6.186484e+019,6.191903e+019] n=501280285 sdn=2.355033e+020 t=2637.27s
3) m=6.188427e+019 sd=1.006031e+016 ci(99%)=[6.185836e+019,6.191019e+019] n=501279965 sdn=2.252432e+020 t=2639.08s
4) m=6.187349e+019 sd=1.007542e+016 ci(99%)=[6.184753e+019,6.189944e+019] n=501279854 sdn=2.255814e+020 t=2635.56s
5) m=6.187708e+019 sd=1.032063e+016 ci(99%)=[6.185050e+019,6.190367e+019] n=501279004 sdn=2.310713e+020 t=1803.31s
6) m=6.188005e+019 sd=1.020789e+016 ci(99%)=[6.185375e+019,6.190634e+019] n=501279218 sdn=2.285471e+020 t=1801.80s
7) m=6.189288e+019 sd=1.071983e+016 ci(99%)=[6.186527e+019,6.192050e+019] n=501279423 sdn=2.400092e+020 t=1801.86s
8) m=6.190143e+019 sd=1.049982e+016 ci(99%)=[6.187438e+019,6.192847e+019] n=501278163 sdn=2.350831e+020 t=1796.52s
9) m=6.187719e+019 sd=1.072126e+016 ci(99%)=[6.184957e+019,6.190481e+019] n=501278534 sdn=2.400410e+020 t=1795.84s
10) m=6.188709e+019 sd=1.088681e+016 ci(99%)=[6.185904e+019,6.191513e+019] n=501280532 sdn=2.437481e+020 t=1796.27s
11) m=6.189915e+019 sd=9.883658e+015 ci(99%)=[6.187369e+019,6.192461e+019] n=501279126 sdn=2.212878e+020 t=1795.14s
12) m=6.187896e+019 sd=1.079972e+016 ci(99%)=[6.185114e+019,6.190678e+019] n=501281179 sdn=2.417983e+020 t=1795.41s
13) m=6.188828e+019 sd=1.063891e+016 ci(99%)=[6.186088e+019,6.191569e+019] n=501280070 sdn=2.381975e+020 t=1798.03s
14) m=6.188041e+019 sd=1.117936e+016 ci(99%)=[6.185161e+019,6.190921e+019] n=501276617 sdn=2.502971e+020 t=1790.55s
15) m=6.189834e+019 sd=1.103298e+016 ci(99%)=[6.186992e+019,6.192676e+019] n=501280169 sdn=2.470205e+020 t=1790.69s
16) m=6.187616e+019 sd=9.864250e+015 ci(99%)=[6.185075e+019,6.190157e+019] n=501279165 sdn=2.208533e+020 t=1792.44s
17) m=6.187847e+019 sd=1.005758e+016 ci(99%)=[6.185256e+019,6.190437e+019] n=501278458 sdn=2.251817e+020 t=1804.33s
18) m=6.189063e+019 sd=1.052264e+016 ci(99%)=[6.186352e+019,6.191774e+019] n=501278921 sdn=2.355941e+020 t=1801.36s
19) m=6.189348e+019 sd=9.871633e+015 ci(99%)=[6.186805e+019,6.191891e+019] n=501277767 sdn=2.210183e+020 t=1788.69s
20) m=6.187280e+019 sd=1.078500e+016 ci(99%)=[6.184502e+019,6.190058e+019] n=501280242 sdn=2.414684e+020 t=1789.23s
21) m=6.186935e+019 sd=1.068577e+016 ci(99%)=[6.184182e+019,6.189688e+019] n=501279639 sdn=2.392466e+020 t=1790.42s
22) m=6.188677e+019 sd=1.043255e+016 ci(99%)=[6.185990e+019,6.191365e+019] n=501280046 sdn=2.335774e+020 t=1779.53s
23) m=6.189836e+019 sd=1.067554e+016 ci(99%)=[6.187086e+019,6.192586e+019] n=501277372 sdn=2.390171e+020 t=1779.41s
24) m=6.187303e+019 sd=1.050259e+016 ci(99%)=[6.184598e+019,6.190009e+019] n=501280197 sdn=2.351455e+020 t=1778.72s
25) m=6.187547e+019 sd=1.011699e+016 ci(99%)=[6.184941e+019,6.190154e+019] n=501278694 sdn=2.265118e+020 t=1803.11s
26) m=6.188085e+019 sd=1.080187e+016 ci(99%)=[6.185303e+019,6.190868e+019] n=501278742 sdn=2.418458e+020 t=1802.89s
27) m=6.188734e+019 sd=9.995885e+015 ci(99%)=[6.186159e+019,6.191309e+019] n=501279495 sdn=2.238006e+020 t=1803.56s |
The first sample was ran with an Intel Pentium D930 almost a month ago; the others are from today, running three simultaneous samples each time in an Intel i5-760.
I have averaged the data with Excel 2003, so you can expect some roundings. Here are the averages I get for these 27 samples:
| Code: |
<m> ~ 61,884,075,552,228,100,000
<sd> ~ 10,437,106,444,028,200
<m>/<sd> ~ 5929.237 |
I add <m>/<sd> for a quick look of the quality of the standard deviation compared with the mean. I hope that all this data will be somewhat helpful in the future.
If I feel energized, I will continue growing the number of samples (at this moment: N = 27) and updating the data periodically, in the way that Steven J. Edwards upgrades his Perft(13) run (a very huge task).
Regards from Spain.
Ajedrecista.
_________________
Six Fortran 95 tools.
Chess will never be solved. |
|
| Back to top |
|
|
Jesús Muñoz
Joined: 13 Jul 2011
Posts: 707
Location: Madrid, Spain.
|
| Posted: Fri Jan 20, 2012 3:27 pm Post subject: Perft(14) estimate after averaging 54 MC perft samples. |
|
|
Hello:
I have been running other 27 MonteCarlo Perft(14) samples and this is what I get:
| Code: |
perftmc 14 (GNU 5.07.173b w32):
28) m=6.188011e+019 sd=9.946225e+015 ci(99%)=[6.185449e+019,6.190573e+019] n=501279154 sdn=2.226887e+020 t=1794.44s
29) m=6.187823e+019 sd=1.083952e+016 ci(99%)=[6.185030e+019,6.190615e+019] n=501280058 sdn=2.426891e+020 t=1793.56s
30) m=6.190087e+019 sd=1.062835e+016 ci(99%)=[6.187350e+019,6.192825e+019] n=501279242 sdn=2.379609e+020 t=1794.03s
31) m=6.186841e+019 sd=1.001161e+016 ci(99%)=[6.184262e+019,6.189420e+019] n=501282355 sdn=2.241533e+020 t=1803.33s
32) m=6.186260e+019 sd=1.058766e+016 ci(99%)=[6.183533e+019,6.188988e+019] n=501278370 sdn=2.370498e+020 t=1802.84s
33) m=6.187918e+019 sd=1.047857e+016 ci(99%)=[6.185219e+019,6.190617e+019] n=501278678 sdn=2.346074e+020 t=1808.39s
34) m=6.187713e+019 sd=1.089137e+016 ci(99%)=[6.184907e+019,6.190518e+019] n=501279776 sdn=2.438498e+020 t=1794.05s
35) m=6.188263e+019 sd=1.121174e+016 ci(99%)=[6.185375e+019,6.191151e+019] n=501280644 sdn=2.510231e+020 t=1794.03s
36) m=6.187222e+019 sd=1.063405e+016 ci(99%)=[6.184483e+019,6.189962e+019] n=501279318 sdn=2.380886e+020 t=1797.12s
37) m=6.187374e+019 sd=1.057575e+016 ci(99%)=[6.184650e+019,6.190099e+019] n=501279868 sdn=2.367833e+020 t=1794.12s
38) m=6.186556e+019 sd=9.809724e+015 ci(99%)=[6.184029e+019,6.189083e+019] n=501280175 sdn=2.196327e+020 t=1796.73s
39) m=6.187162e+019 sd=1.029827e+016 ci(99%)=[6.184510e+019,6.189815e+019] n=501277766 sdn=2.305705e+020 t=1793.00s
40) m=6.188765e+019 sd=1.113190e+016 ci(99%)=[6.185898e+019,6.191633e+019] n=501281045 sdn=2.492355e+020 t=1813.12s
41) m=6.187547e+019 sd=1.037987e+016 ci(99%)=[6.184874e+019,6.190221e+019] n=501280995 sdn=2.323981e+020 t=1814.27s
42) m=6.189318e+019 sd=1.031386e+016 ci(99%)=[6.186661e+019,6.191975e+019] n=501279435 sdn=2.309197e+020 t=1829.69s
43) m=6.188305e+019 sd=1.050592e+016 ci(99%)=[6.185598e+019,6.191011e+019] n=501279973 sdn=2.352199e+020 t=1787.39s
44) m=6.185973e+019 sd=9.954686e+015 ci(99%)=[6.183408e+019,6.188537e+019] n=501281580 sdn=2.228786e+020 t=1789.70s
45) m=6.185511e+019 sd=1.036460e+016 ci(99%)=[6.182841e+019,6.188181e+019] n=501277568 sdn=2.320555e+020 t=1789.03s
46) m=6.188931e+019 sd=1.069488e+016 ci(99%)=[6.186176e+019,6.191686e+019] n=501280008 sdn=2.394508e+020 t=1801.38s
47) m=6.188533e+019 sd=1.025258e+016 ci(99%)=[6.185892e+019,6.191174e+019] n=501279859 sdn=2.295478e+020 t=1796.28s
48) m=6.188584e+019 sd=1.024875e+016 ci(99%)=[6.185944e+019,6.191225e+019] n=501280494 sdn=2.294622e+020 t=1796.48s
49) m=6.187672e+019 sd=1.142710e+016 ci(99%)=[6.184729e+019,6.190616e+019] n=501278206 sdn=2.558442e+020 t=1806.58s
50) m=6.187417e+019 sd=1.029176e+016 ci(99%)=[6.184766e+019,6.190068e+019] n=501280296 sdn=2.304251e+020 t=1802.50s
51) m=6.187121e+019 sd=1.047623e+016 ci(99%)=[6.184422e+019,6.189819e+019] n=501280150 sdn=2.345552e+020 t=1810.09s
52) m=6.188088e+019 sd=1.113788e+016 ci(99%)=[6.185219e+019,6.190957e+019] n=501279478 sdn=2.493689e+020 t=1804.67s
53) m=6.189411e+019 sd=1.098293e+016 ci(99%)=[6.186581e+019,6.192240e+019] n=501278721 sdn=2.458996e+020 t=1803.09s
54) m=6.188640e+019 sd=1.065530e+016 ci(99%)=[6.185896e+019,6.191385e+019] n=501280051 sdn=2.385646e+020 t=1804.02s |
| Code: |
Averages after N = 54 MonteCarlo perft samples:
<m> ~ 61,881,120,367,616,500,000
<sd> ~ 10,495,306,430,406,300
<m>/<sd> ~ 5896.076
<n> ~ 501,279,504.33 |
I have included <n> (the average number of nodes per sample) which should be around 5.0128e+8 nodes. I am not an expert, but these results do not seem unreasonable.
I will be very busy for running more samples until mid February, so I postpone this Perft(14) estimate experiment for a while. I hope to exceed 100 samples, trying to get more accurate results.
Regards from Spain.
Ajedrecista.
_________________
Six Fortran 95 tools.
Chess will never be solved. |
|
| Back to top |
|
|
Jesús Muñoz
Joined: 13 Jul 2011
Posts: 707
Location: Madrid, Spain.
|
| Posted: Sun Jan 22, 2012 4:09 pm Post subject: Some explanations. |
|
|
Hello:
In the output of GNU appears sdn, which is by definition (sd)·sqrt(n) for each sample. I said in my first post that I would use (sdn_i)² for averaging the standard deviation, but actually I am using (n_i)·(sd_i)² for two reasons:
a) More accurate results: I assume that both m_i and sd_i are rounded, but also sdn_i is rounded; so I suppose that the output of (n_i)·(sd_i)² is a little more exact than the output of (sdn_i)².
b) Save time: for each sample, I do three 'copy & paste' (m_i, sd_i and n_i) from a notepad to Excel instead of four (m_i, sd_i, n_i and sdn_i). I save N 'copy & paste' in N samples.
So, the method of averaging the data is now fully explained. I have just realized that Peter Österlund did a similar thing with Perft(13) averaging lots of samples, but he averaged around 180,000 samples! His relative error with the true Perft(13) value was ~ 0.000449%! I have done 54 until the moment, the 0.03%, and with much larger standard deviations (that is why I compare <m>/<sd>, for viewing the quality of the average standard deviation against the average mean). Something is something...
I see that GNU gives a confidence interval with 99% confidence. I knew that this level of confidence is reached between m ± (2.57)sd and m ± (2.58)sd in a normal distribution; so, m ± k·sd has 99% confidence, where k ~ 2.575829303 according to Derive 6. I can give now a minimum value and a maximum value for Perft(14) estimate:
| Code: |
Averages after N = 54 MonteCarlo perft samples:
<m> ~ 61,881,120,367,616,500,000
<sd> ~ 10,495,306,430,406,300
(Minimum value with 99% confidence) ~ <m> - (2.575829303)<sd> ~ 61,854,086,249,769,100,000
(Maximum value with 99% confidence) ~ <m> + (2.575829303)<sd> ~ 61,908,154,485,463,900,000
<m>/<sd> ~ 5,896.076
<n> ~ 501,279,504.33 |
I remember that Excel also rounds its results. In view of the evolution of my tiny experiment, I expect that <sd> will be around 1.05e+16 (with the criterion of stop GNU at more less 5.0128e+8 nodes).
I have done a little search of Perft(14) estimates of people of this forum (except Mr. Labelle), and here are the estimates that I consider more accurate:
| Quote: |
61,803,489,628,662,504,195 by Joshua Haglund.
6.187e+19 by François Labelle.
61,886,459,822,115,294,738 by myself.
6.188925e+19 by H.G.Muller.
6.19009592e+19 by Reinhard Scharnagl. |
Each person uses his own estimate method: this is the reason why I give the links. The more different estimates (Joshua's and Reinhard's) differ in less than 0.16%!
Regards from Spain.
Ajedrecista.
_________________
Six Fortran 95 tools.
Chess will never be solved. |
|
| Back to top |
|
|
Jesús Muñoz
Joined: 13 Jul 2011
Posts: 707
Location: Madrid, Spain.
|
| Posted: Thu Jan 26, 2012 4:44 pm Post subject: Perft(14) estimate after averaging 96 MC perft samples. |
|
|
Hello:
It was totally unexpected, but I managed to run 42 more MonteCarlo samples of Perft(14) this week. I copy the results:
| Code: |
perftmc 14 (GNU 5.07.173b w32):
55) m=6.189126e+019 sd=1.025756e+016 ci(99%)=[6.186484e+019,6.191768e+019] n=501279326 sdn=2.296592e+020 t=1795.62s
56) m=6.189066e+019 sd=1.028560e+016 ci(99%)=[6.186416e+019,6.191716e+019] n=501280839 sdn=2.302874e+020 t=1799.20s
57) m=6.188512e+019 sd=1.023282e+016 ci(99%)=[6.185876e+019,6.191148e+019] n=501278469 sdn=2.291052e+020 t=1798.42s
58) m=6.187755e+019 sd=1.016845e+016 ci(99%)=[6.185136e+019,6.190375e+019] n=501281149 sdn=2.276645e+020 t=1798.14s
59) m=6.188874e+019 sd=1.071921e+016 ci(99%)=[6.186113e+019,6.191635e+019] n=501280297 sdn=2.399956e+020 t=1799.11s
60) m=6.188642e+019 sd=1.041763e+016 ci(99%)=[6.185959e+019,6.191326e+019] n=501279319 sdn=2.332430e+020 t=1804.00s
61) m=6.188827e+019 sd=1.107349e+016 ci(99%)=[6.185975e+019,6.191680e+019] n=501279766 sdn=2.479274e+020 t=1807.48s
62) m=6.187425e+019 sd=1.091419e+016 ci(99%)=[6.184613e+019,6.190236e+019] n=501278351 sdn=2.443606e+020 t=1808.16s
63) m=6.187935e+019 sd=1.041703e+016 ci(99%)=[6.185252e+019,6.190619e+019] n=501280432 sdn=2.332298e+020 t=1811.73s
64) m=6.188126e+019 sd=1.025338e+016 ci(99%)=[6.185484e+019,6.190767e+019] n=501279973 sdn=2.295658e+020 t=1817.97s
65) m=6.188444e+019 sd=1.081156e+016 ci(99%)=[6.185659e+019,6.191229e+019] n=501279716 sdn=2.420631e+020 t=1804.44s
66) m=6.187651e+019 sd=1.120478e+016 ci(99%)=[6.184764e+019,6.190537e+019] n=501278857 sdn=2.508666e+020 t=1805.70s
67) m=6.186686e+019 sd=9.913670e+015 ci(99%)=[6.184132e+019,6.189239e+019] n=501279446 sdn=2.219599e+020 t=1815.36s
68) m=6.190257e+019 sd=1.033446e+016 ci(99%)=[6.187595e+019,6.192919e+019] n=501278726 sdn=2.313809e+020 t=1806.02s
69) m=6.188596e+019 sd=1.104028e+016 ci(99%)=[6.185752e+019,6.191440e+019] n=501278991 sdn=2.471837e+020 t=1805.50s
70) m=6.190975e+019 sd=1.006684e+016 ci(99%)=[6.188382e+019,6.193568e+019] n=501279280 sdn=2.253892e+020 t=1806.19s
71) m=6.189794e+019 sd=1.075091e+016 ci(99%)=[6.187024e+019,6.192563e+019] n=501279168 sdn=2.407050e+020 t=1816.84s
72) m=6.187742e+019 sd=1.021332e+016 ci(99%)=[6.185111e+019,6.190372e+019] n=501279298 sdn=2.286688e+020 t=1806.98s
73) m=6.189354e+019 sd=1.071593e+016 ci(99%)=[6.186593e+019,6.192114e+019] n=501280695 sdn=2.399222e+020 t=1804.38s
74) m=6.189176e+019 sd=1.107339e+016 ci(99%)=[6.186323e+019,6.192028e+019] n=501280582 sdn=2.479254e+020 t=1804.09s
75) m=6.189556e+019 sd=1.080762e+016 ci(99%)=[6.186772e+019,6.192340e+019] n=501279602 sdn=2.419747e+020 t=1809.20s
76) m=6.186469e+019 sd=1.031877e+016 ci(99%)=[6.183811e+019,6.189127e+019] n=501279784 sdn=2.310298e+020 t=1804.23s
77) m=6.189422e+019 sd=1.006358e+016 ci(99%)=[6.186830e+019,6.192014e+019] n=501280566 sdn=2.253164e+020 t=1803.45s
78) m=6.188256e+019 sd=1.076320e+016 ci(99%)=[6.185483e+019,6.191028e+019] n=501277490 sdn=2.409797e+020 t=1801.77s
79) m=6.187859e+019 sd=1.060298e+016 ci(99%)=[6.185128e+019,6.190590e+019] n=501281395 sdn=2.373935e+020 t=1797.48s
80) m=6.187284e+019 sd=1.076798e+016 ci(99%)=[6.184511e+019,6.190058e+019] n=501280409 sdn=2.410875e+020 t=1793.48s
81) m=6.187718e+019 sd=1.072083e+016 ci(99%)=[6.184956e+019,6.190480e+019] n=501280975 sdn=2.400319e+020 t=1793.64s
82) m=6.188221e+019 sd=1.118553e+016 ci(99%)=[6.185339e+019,6.191102e+019] n=501278295 sdn=2.504355e+020 t=1803.11s
83) m=6.188124e+019 sd=1.075494e+016 ci(99%)=[6.185353e+019,6.190894e+019] n=501280270 sdn=2.407955e+020 t=1803.45s
84) m=6.186572e+019 sd=1.005592e+016 ci(99%)=[6.183982e+019,6.189162e+019] n=501279188 sdn=2.251447e+020 t=1811.27s
85) m=6.188960e+019 sd=1.038603e+016 ci(99%)=[6.186284e+019,6.191635e+019] n=501280925 sdn=2.325359e+020 t=1792.14s
86) m=6.188382e+019 sd=1.053394e+016 ci(99%)=[6.185668e+019,6.191095e+019] n=501280516 sdn=2.358475e+020 t=1788.02s
87) m=6.186771e+019 sd=9.579151e+015 ci(99%)=[6.184304e+019,6.189239e+019] n=501278125 sdn=2.144699e+020 t=1786.83s
88) m=6.187624e+019 sd=1.029300e+016 ci(99%)=[6.184972e+019,6.190275e+019] n=501280503 sdn=2.304530e+020 t=1780.06s
89) m=6.188133e+019 sd=9.987345e+015 ci(99%)=[6.185561e+019,6.190706e+019] n=501281412 sdn=2.236098e+020 t=1779.09s
90) m=6.188588e+019 sd=1.124572e+016 ci(99%)=[6.185691e+019,6.191485e+019] n=501280892 sdn=2.517839e+020 t=1780.97s
91) m=6.188649e+019 sd=1.087649e+016 ci(99%)=[6.185848e+019,6.191451e+019] n=501279350 sdn=2.435167e+020 t=1818.27s
92) m=6.188796e+019 sd=1.003571e+016 ci(99%)=[6.186211e+019,6.191381e+019] n=501279124 sdn=2.246922e+020 t=1811.00s
93) m=6.188901e+019 sd=1.107146e+016 ci(99%)=[6.186049e+019,6.191753e+019] n=501278422 sdn=2.478817e+020 t=1810.81s
94) m=6.188876e+019 sd=9.339720e+015 ci(99%)=[6.186470e+019,6.191282e+019] n=501281744 sdn=2.091100e+020 t=1794.28s
95) m=6.189020e+019 sd=1.054152e+016 ci(99%)=[6.186305e+019,6.191736e+019] n=501276286 sdn=2.360162e+020 t=1795.55s
96) m=6.188467e+019 sd=1.074019e+016 ci(99%)=[6.185700e+019,6.191233e+019] n=501280472 sdn=2.404652e+020 t=1797.77s |
Averaging the accumulated data with Excel (hoping no typos):
| Code: |
Averages after N = 96 MonteCarlo perft samples:
<m> ~ 61,882,464,686,162,800,000
<sd> ~ 10,506,880,305,211,500
(Minimum value with 99% confidence) ~ <m> - (2.575829303)<sd> ~ 61,855,400,755,989,600,000
(Maximum value with 99% confidence) ~ <m> + (2.575829303)<sd> ~ 61,909,528,616,336,100,000
<m>/<sd> ~ 5,889.709
<n> ~ 501,279,600.61 |
Three simultaneous samples take around half an hour in an Intel i5-760 (at 2.8 GHz). So, a rough time estimate for this experiment is N/6 hours. At this moment, 96/6 = 16 hours is a good estimate of the effective spent time (spread over several days).
I have updated my data (IIRC) for N = 60, 75, 78 and 90 (without posting the intermediate results in TalkChess) and the average mean value <m> ranged from ~ 6.188211e+19 to ~ 6.188279e+19 more less... I could be lucky (very likely) or a slight sort of convergence is appearing around <m>, probably indicating that the first four numbers of Perft(14) are 6188. But I know that 96 samples is not a big amount of data, and the average standard deviation <sd> is more less 0.017% of <m> (quite large, but it is my criterion stopping around 500 millions of nodes per sample). <sd> increased up to ~ 1.0514e+16 (for N = 75) and now has decreased a little. The amount <m>/<sd> almost has not varied from N = 54 to N = 96.
I will not be able of resume this experiment until mid February. However, I do not know the number of samples that I will run until I finish the experiment (I guess that N will not be very large). I hope that all these averages will be helpful in the future.
Regards from Spain.
Ajedrecista.
_________________
Six Fortran 95 tools.
Chess will never be solved. |
|
| Back to top |
|
|
Jesús Muñoz
Joined: 13 Jul 2011
Posts: 707
Location: Madrid, Spain.
|
| Posted: Fri Feb 17, 2012 8:00 pm Post subject: Perft(14) estimate after averaging 120 MC perft samples. |
|
|
Hello:
As I promised some weeks ago, I resumed this experiment in mid February; I have ran other 24 MonteCarlo samples:
| Code: |
perftmc 14 (GNU 5.07.173b w32):
97) m=6.189968e+019 sd=9.968961e+015 ci(99%)=[6.187400e+019,6.192536e+019] n=501276978 sdn=2.231972e+020 t=1817.47s
98) m=6.187422e+019 sd=1.047597e+016 ci(99%)=[6.184724e+019,6.190121e+019] n=501280763 sdn=2.345497e+020 t=1805.58s
99) m=6.187648e+019 sd=1.095249e+016 ci(99%)=[6.184826e+019,6.190469e+019] n=501280884 sdn=2.452187e+020 t=1807.66s
100) m=6.188386e+019 sd=1.081047e+016 ci(99%)=[6.185602e+019,6.191171e+019] n=501278574 sdn=2.420383e+020 t=1804.91s
101) m=6.188416e+019 sd=1.081914e+016 ci(99%)=[6.185629e+019,6.191203e+019] n=501278937 sdn=2.422324e+020 t=1805.38s
102) m=6.188435e+019 sd=1.033910e+016 ci(99%)=[6.185771e+019,6.191098e+019] n=501279884 sdn=2.314851e+020 t=1805.92s
103) m=6.188559e+019 sd=1.117139e+016 ci(99%)=[6.185682e+019,6.191437e+019] n=501279281 sdn=2.501193e+020 t=1808.84s
104) m=6.189134e+019 sd=1.133385e+016 ci(99%)=[6.186215e+019,6.192054e+019] n=501279579 sdn=2.537567e+020 t=1815.66s
105) m=6.187967e+019 sd=1.013839e+016 ci(99%)=[6.185356e+019,6.190579e+019] n=501278569 sdn=2.269909e+020 t=1804.41s
106) m=6.187396e+019 sd=9.922992e+015 ci(99%)=[6.184839e+019,6.189952e+019] n=501279194 sdn=2.221685e+020 t=1818.95s
107) m=6.187536e+019 sd=1.101518e+016 ci(99%)=[6.184699e+019,6.190374e+019] n=501279098 sdn=2.466218e+020 t=1804.69s
108) m=6.187363e+019 sd=1.109060e+016 ci(99%)=[6.184506e+019,6.190220e+019] n=501280477 sdn=2.483108e+020 t=1806.41s
109) m=6.186728e+019 sd=1.111241e+016 ci(99%)=[6.183865e+019,6.189590e+019] n=501280536 sdn=2.487991e+020 t=1808.86s
110) m=6.187670e+019 sd=1.072955e+016 ci(99%)=[6.184907e+019,6.190434e+019] n=501278981 sdn=2.402266e+020 t=1811.28s
111) m=6.186937e+019 sd=1.113884e+016 ci(99%)=[6.184068e+019,6.189806e+019] n=501278139 sdn=2.493902e+020 t=1808.75s
112) m=6.188844e+019 sd=1.005461e+016 ci(99%)=[6.186254e+019,6.191434e+019] n=501276761 sdn=2.251148e+020 t=1837.12s
113) m=6.189127e+019 sd=1.050394e+016 ci(99%)=[6.186421e+019,6.191833e+019] n=501280428 sdn=2.351758e+020 t=1829.86s
114) m=6.188861e+019 sd=1.064436e+016 ci(99%)=[6.186119e+019,6.191603e+019] n=501279647 sdn=2.383195e+020 t=1837.08s
115) m=6.188280e+019 sd=1.089486e+016 ci(99%)=[6.185473e+019,6.191086e+019] n=501277456 sdn=2.439274e+020 t=1787.16s
116) m=6.188394e+019 sd=1.041013e+016 ci(99%)=[6.185712e+019,6.191075e+019] n=501279502 sdn=2.330752e+020 t=1786.92s
117) m=6.188412e+019 sd=1.011535e+016 ci(99%)=[6.185806e+019,6.191018e+019] n=501278484 sdn=2.264750e+020 t=1787.61s
118) m=6.187346e+019 sd=1.024464e+016 ci(99%)=[6.184707e+019,6.189985e+019] n=501278990 sdn=2.293699e+020 t=1788.97s
119) m=6.187020e+019 sd=1.016764e+016 ci(99%)=[6.184401e+019,6.189639e+019] n=501279750 sdn=2.276461e+020 t=1788.77s
120) m=6.188485e+019 sd=1.097793e+016 ci(99%)=[6.185657e+019,6.191313e+019] n=501278229 sdn=2.457876e+020 t=1788.67s |
It is not easy for me find the requested time for running those samples, so the number of them will grow, but (very) slowly. I do not know the number of samples I will run until I finish this experiment.
Averaging the accumulated data with Excel:
| Code: |
Averages after N = 120 MonteCarlo perft samples:
<m> ~ 61,882,166,247,766,000,000
<sd> ~ 10,532,596,445,177,000
(Minimum value with 99% confidence) ~ <m> - (2.575829303)<sd> ~ 61,855,036,077,205,900,000
(Maximum value with 99% confidence) ~ <m> + (2.575829303)<sd> ~ 61,909,296,418,326,200,000
<m>/<sd> ~ 5,875.300
<n> ~ 501,279,506.50 |
Results seem credible IMHO.
Regards from Spain.
Ajedrecista.
_________________
Six Fortran 95 tools.
Chess will never be solved. |
|
| Back to top |
|
|
Jesús Muñoz
Joined: 13 Jul 2011
Posts: 707
Location: Madrid, Spain.
|
| Posted: Fri Mar 02, 2012 5:26 pm Post subject: Perft(14) estimate after averaging 174 MC perft samples. |
|
|
Hello:
I have ran 54 more MonteCarlo Perft(14) samples during the last two weeks (39 of them in this current week). Here are the results:
| Code: |
perftmc 14 (GNU 5.07.173b w32):
121) m=6.189284e+019 sd=1.008047e+016 ci(99%)=[6.186688e+019,6.191881e+019] n=501278987 sdn=2.256944e+020 t=1794.00s
122) m=6.187890e+019 sd=1.123685e+016 ci(99%)=[6.184995e+019,6.190785e+019] n=501279826 sdn=2.515849e+020 t=1797.20s
123) m=6.188323e+019 sd=1.055339e+016 ci(99%)=[6.185604e+019,6.191041e+019] n=501279158 sdn=2.362827e+020 t=1796.28s
124) m=6.188984e+019 sd=1.047161e+016 ci(99%)=[6.186286e+019,6.191681e+019] n=501278865 sdn=2.344515e+020 t=1809.95s
125) m=6.188092e+019 sd=1.007806e+016 ci(99%)=[6.185496e+019,6.190688e+019] n=501277680 sdn=2.256399e+020 t=1810.48s
126) m=6.190602e+019 sd=1.012834e+016 ci(99%)=[6.187993e+019,6.193211e+019] n=501279540 sdn=2.267661e+020 t=1809.20s
127) m=6.189542e+019 sd=9.590962e+015 ci(99%)=[6.187071e+019,6.192012e+019] n=501279403 sdn=2.147346e+020 t=1793.72s
128) m=6.189455e+019 sd=1.003952e+016 ci(99%)=[6.186869e+019,6.192042e+019] n=501279233 sdn=2.247775e+020 t=1797.50s
129) m=6.187941e+019 sd=1.052887e+016 ci(99%)=[6.185229e+019,6.190653e+019] n=501278958 sdn=2.357335e+020 t=1794.56s
130) m=6.188222e+019 sd=9.940390e+015 ci(99%)=[6.185662e+019,6.190783e+019] n=501280319 sdn=2.225583e+020 t=1789.98s
131) m=6.188558e+019 sd=1.040922e+016 ci(99%)=[6.185876e+019,6.191239e+019] n=501279962 sdn=2.330550e+020 t=1788.97s
132) m=6.187640e+019 sd=1.007706e+016 ci(99%)=[6.185045e+019,6.190236e+019] n=501278639 sdn=2.256179e+020 t=1789.83s
133) m=6.188719e+019 sd=1.103660e+016 ci(99%)=[6.185876e+019,6.191562e+019] n=501280461 sdn=2.471017e+020 t=1789.38s
134) m=6.187670e+019 sd=8.907128e+015 ci(99%)=[6.185375e+019,6.189964e+019] n=501279361 sdn=1.994241e+020 t=1788.58s
135) m=6.190021e+019 sd=1.031477e+016 ci(99%)=[6.187364e+019,6.192678e+019] n=501280856 sdn=2.309405e+020 t=1789.97s
136) m=6.186606e+019 sd=1.070175e+016 ci(99%)=[6.183850e+019,6.189363e+019] n=501279708 sdn=2.396046e+020 t=1779.16s
137) m=6.187397e+019 sd=1.159761e+016 ci(99%)=[6.184410e+019,6.190385e+019] n=501279745 sdn=2.596622e+020 t=1782.75s
138) m=6.187630e+019 sd=9.482699e+015 ci(99%)=[6.185187e+019,6.190072e+019] n=501280498 sdn=2.123109e+020 t=1779.52s
139) m=6.189303e+019 sd=1.103679e+016 ci(99%)=[6.186460e+019,6.192146e+019] n=501280212 sdn=2.471058e+020 t=1796.73s
140) m=6.188546e+019 sd=1.137474e+016 ci(99%)=[6.185616e+019,6.191476e+019] n=501279468 sdn=2.546721e+020 t=1799.14s
141) m=6.187604e+019 sd=1.004837e+016 ci(99%)=[6.185016e+019,6.190193e+019] n=501278151 sdn=2.249755e+020 t=1800.30s
142) m=6.187485e+019 sd=1.011928e+016 ci(99%)=[6.184878e+019,6.190092e+019] n=501279472 sdn=2.265633e+020 t=1798.12s
143) m=6.188353e+019 sd=9.887368e+015 ci(99%)=[6.185806e+019,6.190900e+019] n=501277751 sdn=2.213706e+020 t=1795.67s
144) m=6.188995e+019 sd=9.882406e+015 ci(99%)=[6.186449e+019,6.191541e+019] n=501279360 sdn=2.212599e+020 t=1798.36s
145) m=6.188272e+019 sd=9.780943e+015 ci(99%)=[6.185752e+019,6.190791e+019] n=501282312 sdn=2.189888e+020 t=1800.72s
146) m=6.187672e+019 sd=1.084942e+016 ci(99%)=[6.184877e+019,6.190467e+019] n=501277682 sdn=2.429101e+020 t=1798.41s
147) m=6.188110e+019 sd=1.053987e+016 ci(99%)=[6.185394e+019,6.190825e+019] n=501277514 sdn=2.359796e+020 t=1797.59s
148) m=6.187031e+019 sd=1.014276e+016 ci(99%)=[6.184418e+019,6.189644e+019] n=501281110 sdn=2.270894e+020 t=1798.31s
149) m=6.187199e+019 sd=1.117055e+016 ci(99%)=[6.184321e+019,6.190076e+019] n=501278954 sdn=2.501003e+020 t=1797.16s
150) m=6.188504e+019 sd=1.115479e+016 ci(99%)=[6.185631e+019,6.191378e+019] n=501279919 sdn=2.497477e+020 t=1802.39s
151) m=6.186887e+019 sd=1.062790e+016 ci(99%)=[6.184149e+019,6.189625e+019] n=501279154 sdn=2.379509e+020 t=1794.09s
152) m=6.188174e+019 sd=9.944726e+015 ci(99%)=[6.185612e+019,6.190736e+019] n=501279662 sdn=2.226552e+020 t=1790.56s
153) m=6.188554e+019 sd=1.041242e+016 ci(99%)=[6.185872e+019,6.191237e+019] n=501279522 sdn=2.331265e+020 t=1794.12s
154) m=6.190050e+019 sd=9.932023e+015 ci(99%)=[6.187491e+019,6.192608e+019] n=501278635 sdn=2.223706e+020 t=1798.81s
155) m=6.189155e+019 sd=1.149098e+016 ci(99%)=[6.186195e+019,6.192115e+019] n=501279808 sdn=2.572747e+020 t=1793.88s
156) m=6.189017e+019 sd=1.005289e+016 ci(99%)=[6.186427e+019,6.191606e+019] n=501278928 sdn=2.250768e+020 t=1794.17s
157) m=6.187382e+019 sd=1.033040e+016 ci(99%)=[6.184721e+019,6.190044e+019] n=501279086 sdn=2.312901e+020 t=1800.12s
158) m=6.189066e+019 sd=1.103952e+016 ci(99%)=[6.186223e+019,6.191910e+019] n=501280079 sdn=2.471670e+020 t=1804.23s
159) m=6.189075e+019 sd=1.044283e+016 ci(99%)=[6.186385e+019,6.191765e+019] n=501280564 sdn=2.338075e+020 t=1801.02s
160) m=6.187843e+019 sd=1.125054e+016 ci(99%)=[6.184944e+019,6.190741e+019] n=501279519 sdn=2.518915e+020 t=1819.77s
161) m=6.187819e+019 sd=1.075871e+016 ci(99%)=[6.185048e+019,6.190591e+019] n=501280569 sdn=2.408799e+020 t=1819.33s
162) m=6.188285e+019 sd=1.081494e+016 ci(99%)=[6.185499e+019,6.191071e+019] n=501279010 sdn=2.421386e+020 t=1813.84s
163) m=6.188064e+019 sd=1.086191e+016 ci(99%)=[6.185266e+019,6.190862e+019] n=501279776 sdn=2.431903e+020 t=1794.53s
164) m=6.186738e+019 sd=1.015982e+016 ci(99%)=[6.184121e+019,6.189355e+019] n=501279864 sdn=2.274711e+020 t=1798.00s
165) m=6.189776e+019 sd=1.036952e+016 ci(99%)=[6.187105e+019,6.192447e+019] n=501280786 sdn=2.321662e+020 t=1805.20s
166) m=6.187368e+019 sd=1.017968e+016 ci(99%)=[6.184746e+019,6.189991e+019] n=501279713 sdn=2.279156e+020 t=1806.56s
167) m=6.187860e+019 sd=1.142511e+016 ci(99%)=[6.184916e+019,6.190803e+019] n=501280526 sdn=2.558001e+020 t=1816.75s
168) m=6.189199e+019 sd=1.053784e+016 ci(99%)=[6.186485e+019,6.191914e+019] n=501280494 sdn=2.359347e+020 t=1804.31s
169) m=6.188049e+019 sd=1.104201e+016 ci(99%)=[6.185204e+019,6.190893e+019] n=501280569 sdn=2.472228e+020 t=1801.94s
170) m=6.188333e+019 sd=1.055069e+016 ci(99%)=[6.185615e+019,6.191050e+019] n=501280229 sdn=2.362224e+020 t=1804.25s
171) m=6.188742e+019 sd=1.044487e+016 ci(99%)=[6.186051e+019,6.191433e+019] n=501278692 sdn=2.338529e+020 t=1802.48s
172) m=6.187370e+019 sd=1.010134e+016 ci(99%)=[6.184768e+019,6.189973e+019] n=501279299 sdn=2.261615e+020 t=1803.36s
173) m=6.188538e+019 sd=1.209636e+016 ci(99%)=[6.185422e+019,6.191654e+019] n=501278304 sdn=2.708284e+020 t=1800.78s
174) m=6.189349e+019 sd=1.093528e+016 ci(99%)=[6.186533e+019,6.192166e+019] n=501279202 sdn=2.448329e+020 t=1804.97s |
Averaging the accumulated data with Excel:
| Code: |
Averages after N = 174 MonteCarlo perft samples:
<m> ~ 61,882,548,159,772,100,000
<sd> ~ 10,527,431,703,577,400
(Minimum value with 99% confidence) ~ <m> - (2.575829303)<sd> ~ 61,855,431,292,704,700,000
(Maximum value with 99% confidence) ~ <m> + (2.575829303)<sd> ~ 61,909,665,026,839,500,000
<m>/<sd> ~ 5,878.219
<n> ~ 501,279,516.52 |
The average (weighted by the number of nodes of each sample) of the mean values (<m>) still remains between 6.1882e+19 and 6.1883e+19.
Regards from Spain.
Ajedrecista.
_________________
Six Fortran 95 tools.
Chess will never be solved. |
|
| Back to top |
|
|
Jesús Muñoz
Joined: 13 Jul 2011
Posts: 707
Location: Madrid, Spain.
|
| Posted: Fri Mar 09, 2012 3:50 pm Post subject: Perft(14) estimate after averaging 222 MC perft samples. |
|
|
Hello:
I managed to run 48 MonteCarlo Perft(14) samples this week:
| Code: |
perftmc 14 (GNU 5.07.173b w32):
175) m=6.188097e+019 sd=1.088038e+016 ci(99%)=[6.185294e+019,6.190900e+019] n=501280604 sdn=2.436040e+020 t=1798.61s
176) m=6.189682e+019 sd=1.140451e+016 ci(99%)=[6.186745e+019,6.192620e+019] n=501278241 sdn=2.553384e+020 t=1804.20s
177) m=6.188218e+019 sd=1.073110e+016 ci(99%)=[6.185454e+019,6.190983e+019] n=501279423 sdn=2.402616e+020 t=1799.52s
178) m=6.189257e+019 sd=1.079571e+016 ci(99%)=[6.186476e+019,6.192038e+019] n=501278349 sdn=2.417079e+020 t=1792.98s
179) m=6.186884e+019 sd=1.019336e+016 ci(99%)=[6.184259e+019,6.189510e+019] n=501279827 sdn=2.282221e+020 t=1796.72s
180) m=6.187771e+019 sd=9.683364e+015 ci(99%)=[6.185277e+019,6.190266e+019] n=501279426 sdn=2.168035e+020 t=1794.55s
181) m=6.189102e+019 sd=1.074193e+016 ci(99%)=[6.186335e+019,6.191869e+019] n=501279002 sdn=2.405039e+020 t=1808.94s
182) m=6.188231e+019 sd=1.175920e+016 ci(99%)=[6.185202e+019,6.191260e+019] n=501280331 sdn=2.632802e+020 t=1811.45s
183) m=6.187141e+019 sd=1.053201e+016 ci(99%)=[6.184428e+019,6.189854e+019] n=501279393 sdn=2.358041e+020 t=1803.80s
184) m=6.189440e+019 sd=1.080515e+016 ci(99%)=[6.186657e+019,6.192224e+019] n=501281193 sdn=2.419199e+020 t=1787.52s
185) m=6.189122e+019 sd=1.054490e+016 ci(99%)=[6.186406e+019,6.191839e+019] n=501278700 sdn=2.360924e+020 t=1789.36s
186) m=6.188205e+019 sd=1.076295e+016 ci(99%)=[6.185433e+019,6.190978e+019] n=501281325 sdn=2.409750e+020 t=1787.95s
187) m=6.190043e+019 sd=1.084941e+016 ci(99%)=[6.187248e+019,6.192838e+019] n=501279658 sdn=2.429105e+020 t=1818.27s
188) m=6.187634e+019 sd=9.958161e+015 ci(99%)=[6.185069e+019,6.190199e+019] n=501280099 sdn=2.229561e+020 t=1809.98s
189) m=6.188988e+019 sd=1.046943e+016 ci(99%)=[6.186291e+019,6.191685e+019] n=501280212 sdn=2.344030e+020 t=1808.45s
190) m=6.188849e+019 sd=1.113586e+016 ci(99%)=[6.185981e+019,6.191718e+019] n=501279431 sdn=2.493237e+020 t=1808.48s
191) m=6.187719e+019 sd=1.096988e+016 ci(99%)=[6.184893e+019,6.190545e+019] n=501278973 sdn=2.456076e+020 t=1819.70s
192) m=6.187995e+019 sd=9.954302e+015 ci(99%)=[6.185431e+019,6.190560e+019] n=501279573 sdn=2.228696e+020 t=1809.52s
193) m=6.186921e+019 sd=9.928545e+015 ci(99%)=[6.184364e+019,6.189479e+019] n=501278751 sdn=2.222927e+020 t=1818.11s
194) m=6.187739e+019 sd=1.092481e+016 ci(99%)=[6.184925e+019,6.190553e+019] n=501279477 sdn=2.445985e+020 t=1810.14s
195) m=6.190203e+019 sd=1.064405e+016 ci(99%)=[6.187461e+019,6.192944e+019] n=501279422 sdn=2.383126e+020 t=1814.94s
196) m=6.188280e+019 sd=1.097746e+016 ci(99%)=[6.185452e+019,6.191108e+019] n=501280748 sdn=2.457775e+020 t=1819.36s
197) m=6.186709e+019 sd=1.052132e+016 ci(99%)=[6.183999e+019,6.189419e+019] n=501277695 sdn=2.355643e+020 t=1817.78s
198) m=6.189555e+019 sd=9.757581e+015 ci(99%)=[6.187041e+019,6.192069e+019] n=501281256 sdn=2.184655e+020 t=1882.45s
199) m=6.188959e+019 sd=1.064004e+016 ci(99%)=[6.186218e+019,6.191700e+019] n=501279547 sdn=2.382228e+020 t=1793.67s
200) m=6.186608e+019 sd=1.074596e+016 ci(99%)=[6.183840e+019,6.189376e+019] n=501279334 sdn=2.405943e+020 t=1793.77s
201) m=6.188662e+019 sd=1.071486e+016 ci(99%)=[6.185901e+019,6.191422e+019] n=501279648 sdn=2.398980e+020 t=1793.83s
202) m=6.187063e+019 sd=1.161566e+016 ci(99%)=[6.184071e+019,6.190056e+019] n=501279135 sdn=2.600660e+020 t=1793.48s
203) m=6.188481e+019 sd=1.041675e+016 ci(99%)=[6.185797e+019,6.191164e+019] n=501280223 sdn=2.332237e+020 t=1791.39s
204) m=6.188920e+019 sd=1.035624e+016 ci(99%)=[6.186252e+019,6.191588e+019] n=501279765 sdn=2.318687e+020 t=1791.20s
205) m=6.188009e+019 sd=9.863987e+015 ci(99%)=[6.185468e+019,6.190550e+019] n=501279443 sdn=2.208475e+020 t=1811.45s
206) m=6.188296e+019 sd=1.117885e+016 ci(99%)=[6.185417e+019,6.191176e+019] n=501279363 sdn=2.502862e+020 t=1808.45s
207) m=6.186660e+019 sd=1.151582e+016 ci(99%)=[6.183694e+019,6.189627e+019] n=501280529 sdn=2.578311e+020 t=1808.88s
208) m=6.188055e+019 sd=1.052447e+016 ci(99%)=[6.185344e+019,6.190767e+019] n=501279637 sdn=2.356352e+020 t=1804.09s
209) m=6.188877e+019 sd=1.021727e+016 ci(99%)=[6.186246e+019,6.191509e+019] n=501282237 sdn=2.287579e+020 t=1802.30s
210) m=6.188829e+019 sd=1.007519e+016 ci(99%)=[6.186234e+019,6.191425e+019] n=501279869 sdn=2.255762e+020 t=1804.25s
211) m=6.188321e+019 sd=9.848306e+015 ci(99%)=[6.185784e+019,6.190858e+019] n=501280406 sdn=2.204966e+020 t=1806.23s
212) m=6.188375e+019 sd=1.071925e+016 ci(99%)=[6.185613e+019,6.191136e+019] n=501279255 sdn=2.399961e+020 t=1805.22s
213) m=6.189752e+019 sd=1.053322e+016 ci(99%)=[6.187039e+019,6.192466e+019] n=501279910 sdn=2.358313e+020 t=1815.97s
214) m=6.187089e+019 sd=1.079322e+016 ci(99%)=[6.184309e+019,6.189870e+019] n=501279030 sdn=2.416522e+020 t=1784.67s
215) m=6.187456e+019 sd=9.834222e+015 ci(99%)=[6.184923e+019,6.189989e+019] n=501279505 sdn=2.201811e+020 t=1788.14s
216) m=6.188619e+019 sd=1.160983e+016 ci(99%)=[6.185629e+019,6.191610e+019] n=501280097 sdn=2.599358e+020 t=1784.86s
217) m=6.188617e+019 sd=1.053039e+016 ci(99%)=[6.185904e+019,6.191330e+019] n=501280732 sdn=2.357681e+020 t=1795.52s
218) m=6.189534e+019 sd=1.122056e+016 ci(99%)=[6.186643e+019,6.192424e+019] n=501280484 sdn=2.512205e+020 t=1789.41s
219) m=6.188727e+019 sd=1.029030e+016 ci(99%)=[6.186076e+019,6.191378e+019] n=501279376 sdn=2.303923e+020 t=1792.22s
220) m=6.187117e+019 sd=1.125649e+016 ci(99%)=[6.184217e+019,6.190017e+019] n=501279734 sdn=2.520247e+020 t=1809.67s
221) m=6.189189e+019 sd=1.032011e+016 ci(99%)=[6.186531e+019,6.191848e+019] n=501278415 sdn=2.310594e+020 t=1806.98s
222) m=6.188647e+019 sd=9.984144e+015 ci(99%)=[6.186075e+019,6.191219e+019] n=501279959 sdn=2.235378e+020 t=1812.36s |
Averaging the accumulated data with Excel:
| Code: |
Averages after N = 222 MonteCarlo perft samples:
<m> ~ 61,882,747,071,807,300,000
<sd> ~ 10,550,127,986,642,200
(Minimum value with 99% confidence) ~ <m> - (2.575829303)<sd> ~ 61,855,571,742,988,900,000
(Maximum value with 99% confidence) ~ <m> + (2.575829303)<sd> ~ 61,909,922,400,625,700,000
<m>/<sd> ~ 5,865.592
<n> ~ 501,279,561.33 |
I still get 6.1882e+19 < (<m>) < 6.1883e+19; OTOH, the average standard deviation is raising slowly because I obtain some samples with high standard deviations, and those ones raise <sd> (the smaller standard deviations can not compensate the higher ones).
Regards from Spain.
Ajedrecista.
_________________
Six Fortran 95 tools.
Chess will never be solved. |
|
| Back to top |
|
|
Jesús Muñoz
Joined: 13 Jul 2011
Posts: 707
Location: Madrid, Spain.
|
| Posted: Fri Mar 16, 2012 4:57 pm Post subject: Perft(14) estimate after averaging 270 MC perft samples. |
|
|
Hello:
48 MonteCarlo Perft(14) samples were ran this week:
| Code: |
perftmc 14 (GNU 5.07.173b w32):
223) m=6.189385e+019 sd=1.084110e+016 ci(99%)=[6.186592e+019,6.192178e+019] n=501279642 sdn=2.427243e+020 t=1789.30s
224) m=6.187004e+019 sd=1.073432e+016 ci(99%)=[6.184239e+019,6.189769e+019] n=501281449 sdn=2.403341e+020 t=1792.52s
225) m=6.188099e+019 sd=1.015489e+016 ci(99%)=[6.185483e+019,6.190715e+019] n=501280191 sdn=2.273608e+020 t=1790.08s
226) m=6.185762e+019 sd=1.039561e+016 ci(99%)=[6.183085e+019,6.188440e+019] n=501280644 sdn=2.327503e+020 t=1803.44s
227) m=6.185942e+019 sd=1.058966e+016 ci(99%)=[6.183214e+019,6.188670e+019] n=501279804 sdn=2.370948e+020 t=1817.53s
228) m=6.186120e+019 sd=1.062726e+016 ci(99%)=[6.183382e+019,6.188857e+019] n=501279393 sdn=2.379367e+020 t=1805.62s
229) m=6.189011e+019 sd=9.831376e+015 ci(99%)=[6.186479e+019,6.191544e+019] n=501279881 sdn=2.201174e+020 t=1809.88s
230) m=6.188422e+019 sd=9.893324e+015 ci(99%)=[6.185873e+019,6.190970e+019] n=501279402 sdn=2.215043e+020 t=1804.72s
231) m=6.188332e+019 sd=1.016738e+016 ci(99%)=[6.185713e+019,6.190951e+019] n=501279672 sdn=2.276402e+020 t=1808.75s
232) m=6.188688e+019 sd=9.999232e+015 ci(99%)=[6.186112e+019,6.191264e+019] n=501278999 sdn=2.238754e+020 t=1803.78s
233) m=6.189257e+019 sd=1.023457e+016 ci(99%)=[6.186620e+019,6.191893e+019] n=501281044 sdn=2.291449e+020 t=1815.62s
234) m=6.188391e+019 sd=1.134061e+016 ci(99%)=[6.185470e+019,6.191313e+019] n=501279113 sdn=2.539079e+020 t=1804.88s
235) m=6.188103e+019 sd=1.046489e+016 ci(99%)=[6.185407e+019,6.190799e+019] n=501278920 sdn=2.343011e+020 t=1806.86s
236) m=6.188196e+019 sd=9.812251e+015 ci(99%)=[6.185668e+019,6.190724e+019] n=501279864 sdn=2.196892e+020 t=1803.61s
237) m=6.187069e+019 sd=1.072360e+016 ci(99%)=[6.184307e+019,6.189831e+019] n=501280988 sdn=2.400939e+020 t=1815.00s
238) m=6.187368e+019 sd=1.140424e+016 ci(99%)=[6.184430e+019,6.190306e+019] n=501279699 sdn=2.553328e+020 t=1803.89s
239) m=6.188927e+019 sd=1.040781e+016 ci(99%)=[6.186246e+019,6.191608e+019] n=501279894 sdn=2.330233e+020 t=1803.70s
240) m=6.186939e+019 sd=1.027641e+016 ci(99%)=[6.184292e+019,6.189586e+019] n=501280089 sdn=2.300815e+020 t=1803.33s
241) m=6.187985e+019 sd=1.172311e+016 ci(99%)=[6.184965e+019,6.191005e+019] n=501281215 sdn=2.624724e+020 t=1799.59s
242) m=6.187433e+019 sd=1.007750e+016 ci(99%)=[6.184837e+019,6.190029e+019] n=501280722 sdn=2.256282e+020 t=1800.05s
243) m=6.187623e+019 sd=1.080829e+016 ci(99%)=[6.184839e+019,6.190407e+019] n=501279819 sdn=2.419898e+020 t=1800.81s
244) m=6.190023e+019 sd=1.095258e+016 ci(99%)=[6.187201e+019,6.192844e+019] n=501281380 sdn=2.452207e+020 t=1795.69s
245) m=6.189770e+019 sd=1.046610e+016 ci(99%)=[6.187074e+019,6.192466e+019] n=501279274 sdn=2.343283e+020 t=1797.91s
246) m=6.186297e+019 sd=1.071092e+016 ci(99%)=[6.183538e+019,6.189057e+019] n=501277786 sdn=2.398093e+020 t=1794.33s
247) m=6.186454e+019 sd=1.017963e+016 ci(99%)=[6.183832e+019,6.189076e+019] n=501279566 sdn=2.279145e+020 t=1792.36s
248) m=6.188582e+019 sd=1.017347e+016 ci(99%)=[6.185961e+019,6.191202e+019] n=501279996 sdn=2.277768e+020 t=1793.11s
249) m=6.188076e+019 sd=1.124135e+016 ci(99%)=[6.185180e+019,6.190972e+019] n=501280860 sdn=2.516860e+020 t=1790.23s
250) m=6.188355e+019 sd=1.048577e+016 ci(99%)=[6.185654e+019,6.191056e+019] n=501277895 sdn=2.347685e+020 t=1804.52s
251) m=6.188824e+019 sd=1.029093e+016 ci(99%)=[6.186173e+019,6.191475e+019] n=501279094 sdn=2.304063e+020 t=1803.67s
252) m=6.187363e+019 sd=1.036214e+016 ci(99%)=[6.184694e+019,6.190033e+019] n=501281214 sdn=2.320011e+020 t=1806.84s
253) m=6.188831e+019 sd=1.068899e+016 ci(99%)=[6.186077e+019,6.191584e+019] n=501279391 sdn=2.393187e+020 t=1796.05s
254) m=6.189269e+019 sd=1.088491e+016 ci(99%)=[6.186465e+019,6.192073e+019] n=501278684 sdn=2.437051e+020 t=1802.28s
255) m=6.190091e+019 sd=1.077227e+016 ci(99%)=[6.187316e+019,6.192866e+019] n=501279952 sdn=2.411834e+020 t=1799.67s
256) m=6.187218e+019 sd=1.046309e+016 ci(99%)=[6.184523e+019,6.189914e+019] n=501279390 sdn=2.342609e+020 t=1818.70s
257) m=6.190159e+019 sd=1.040234e+016 ci(99%)=[6.187479e+019,6.192838e+019] n=501280509 sdn=2.329011e+020 t=1828.83s
258) m=6.188909e+019 sd=9.918563e+015 ci(99%)=[6.186354e+019,6.191464e+019] n=501279475 sdn=2.220694e+020 t=1828.77s
259) m=6.188671e+019 sd=1.080453e+016 ci(99%)=[6.185888e+019,6.191454e+019] n=501280379 sdn=2.419058e+020 t=1811.81s
260) m=6.187343e+019 sd=1.001726e+016 ci(99%)=[6.184762e+019,6.189923e+019] n=501280529 sdn=2.242794e+020 t=1809.36s
261) m=6.187563e+019 sd=1.028033e+016 ci(99%)=[6.184915e+019,6.190211e+019] n=501279734 sdn=2.301692e+020 t=1811.84s
262) m=6.188912e+019 sd=1.055022e+016 ci(99%)=[6.186194e+019,6.191629e+019] n=501278057 sdn=2.362113e+020 t=1806.97s
263) m=6.191272e+019 sd=1.040150e+016 ci(99%)=[6.188592e+019,6.193951e+019] n=501278618 sdn=2.328818e+020 t=1804.91s
264) m=6.188305e+019 sd=1.015731e+016 ci(99%)=[6.185689e+019,6.190922e+019] n=501279604 sdn=2.274148e+020 t=1809.97s
265) m=6.187664e+019 sd=1.071063e+016 ci(99%)=[6.184905e+019,6.190423e+019] n=501278320 sdn=2.398029e+020 t=1791.80s
266) m=6.187710e+019 sd=1.030565e+016 ci(99%)=[6.185056e+019,6.190365e+019] n=501280355 sdn=2.307363e+020 t=1790.69s
267) m=6.188167e+019 sd=1.189517e+016 ci(99%)=[6.185103e+019,6.191231e+019] n=501278054 sdn=2.663237e+020 t=1792.86s
268) m=6.188379e+019 sd=1.079089e+016 ci(99%)=[6.185599e+019,6.191159e+019] n=501278845 sdn=2.416001e+020 t=1818.69s
269) m=6.188676e+019 sd=1.132908e+016 ci(99%)=[6.185758e+019,6.191594e+019] n=501281072 sdn=2.536502e+020 t=1818.06s
270) m=6.188484e+019 sd=1.064403e+016 ci(99%)=[6.185742e+019,6.191226e+019] n=501279075 sdn=2.383119e+020 t=1822.56s |
Averaging the accumulated data with Excel:
| Code: |
Averages after N = 270 MonteCarlo perft samples:
<m> ~ 61,882,607,703,086,900,000
<sd> ~ 10,551,871,275,442,700
(Minimum value with 99% confidence) ~ <m> - (2.575829303)<sd> ~ 61,855,427,883,854,100,000
(Maximum value with 99% confidence) ~ <m> + (2.575829303)<sd> ~ 61,909,787,522,319,600,000
<m>/<sd> ~ 5,864.610
<n> ~ 501,279,593.21 |
The average results have not varied so much from last week.
Regards from Spain.
Ajedrecista.
_________________
Six Fortran 95 tools.
Chess will never be solved. |
|
| Back to top |
|
|
|
|
You cannot post new topics in this forum
You cannot reply to topics in this forum
You cannot edit your posts in this forum
You cannot delete your posts in this forum
You cannot vote in polls in this forum
|
|
FAQ
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
