RNG vs Actuals: A challenge

[Bayes]

Hey bombus, did you post all 100 predictions yet? I can only see 90.

[bombus]

wLL wLL LLw Lww nolinks wLL Lww wwL LLL wLL nolinks wLw


1: rng
2: rng
3: rng ?
4: actuals
5: rng
6: actuals?
7: rng?
8: actuals
9: rng
10: actuals
11: rng
12: actuals?
13: rng?
14:actuals
15: rng
16: rng
17: actuals?
18: rng?
19: actuals?
20: actuals
21: actuals?
22: rng
23: actuals
24: actuals?
25: actuals
26: rng
27: rng
28: actuals
29: rng
30: actuals
31: rng
32: actuals?
33: actuals
34: rng
35: actuals?
36:rng
37: actuals
38:actuals?
39: rng
40: actuals
41: actuals
42: actuals
43: rng
44: actuals
45: actuals
46: rng
47: rng
48: actuals
49: rng
50: actuals
51: rng
52: actuals
53: rng
54: rng
55: rng
56: rng
57: rng
58: rng
59: rng
60: rng
61: rng
62: actuals
63: actuals
64: actuals
65: actuals
66: actuals
67: actuals
68: rng
69: actuals
70: actuals
71: rng
72: actuals
73: actuals
74: rng
75: rng
76: actuals
77: rng
78: rng
79: actuals
80: actuals
81: actuals
82: actuals
83: actuals
84: actuals
85: actuals
86: rng
87: actuals
88: rng
89: rng
90: rng
91: actuals
92: rng
93: actuals
94: actuals
95: rng
96: actuals
97: actuals
98: rng
99: rng
100: rng


Ok, that's it Bayes!

Let's have the results whenever you're ready.

Cheers.

[Bayes]

sample #1 - RNG
sample #2 - Actuals
sample #3 - Actuals
sample #4 - Actuals
sample #5 - Actuals
sample #6 - RNG
sample #7 - Actuals
sample #8 - RNG
sample #9 - RNG
sample #10 - RNG
sample #11 - RNG
sample #12 - Actuals
sample #13 - RNG
sample #14 - Actuals
sample #15 - RNG
sample #16 - RNG
sample #17 - RNG
sample #18 - Actuals
sample #19 - RNG
sample #20 - Actuals
sample #21 - Actuals
sample #22 - RNG
sample #23 - Actuals
sample #24 - RNG
sample #25 - RNG
sample #26 - Actuals
sample #27 - Actuals
sample #28 - Actuals
sample #29 - Actuals
sample #30 - RNG
sample #31 - RNG
sample #32 - Actuals
sample #33 - Actuals
sample #34 - RNG
sample #35 - RNG
sample #36 - RNG
sample #37 - Actuals
sample #38 - Actuals
sample #39 - Actuals
sample #40 - Actuals
sample #41 - Actuals
sample #42 - Actuals
sample #43 - Actuals
sample #44 - RNG
sample #45 - RNG
sample #46 - Actuals
sample #47 - RNG
sample #48 - Actuals
sample #49 - Actuals
sample #50 - RNG
sample #51 - Actuals
sample #52 - Actuals
sample #53 - RNG
sample #54 - Actuals
sample #55 - Actuals
sample #56 - RNG
sample #57 - RNG
sample #58 - Actuals
sample #59 - Actuals
sample #60 - Actuals
sample #61 - RNG
sample #62 - Actuals
sample #63 - RNG
sample #64 - RNG
sample #65 - Actuals
sample #66 - RNG
sample #67 - Actuals
sample #68 - Actuals
sample #69 - RNG
sample #70 - Actuals
sample #71 - Actuals
sample #72 - Actuals
sample #73 - Actuals
sample #74 - RNG
sample #75 - RNG
sample #76 - RNG
sample #77 - Actuals
sample #78 - RNG
sample #79 - RNG
sample #80 - Actuals
sample #81 - RNG
sample #82 - RNG
sample #83 - Actuals
sample #84 - RNG
sample #85 - Actuals
sample #86 - Actuals
sample #87 - RNG
sample #88 - RNG
sample #89 - Actuals
sample #90 - RNG
sample #91 - Actuals
sample #92 - RNG
sample #93 - RNG
sample #94 - RNG
sample #95 - RNG
sample #96 - RNG
sample #97 - Actuals
sample #98 - RNG
sample #99 - Actuals
sample #100 - RNG



?      ?

[maestro]

sample-4/wrong/    L
sample-5/actuals/  W
sample-6/rng/      W
sample-7/wrong/    L
sample-8/rng/      W
sample-9/wrong/    L
sample-10/rng/     W
sample-11/rng/     W
sample-12/actual/  W
sample-13/wrongl/  L
sample-14/wrong/   L
sample-15/wrong/   L
sample-16/rng/     W
sample-17/wrong/   L
sample-18/actuals/ W
sample-19/wrong/   L
sample-20/actuals/ W
sample-21/wrong/   L
sample-22/rng/     W
sample-23/wrong/   L
sample-24/rng/     W
sample-25/wrong/   L
sample-26/wrong/   L
sample-27/actuals/ W
sample-28/actuals/ W
sample-29/actuals/ W
sample-30/rng/     W
sample-31/wrong/   L
sample-32/actuals/ W
sample-33/wrong/   L
sample-34/wrong/   L
sample-35/rng/     W
sample-36/wrong/   L
sample-37/rng/     L
sample-38/actuals/ W
sample-39/rng/     L
sample-40/actuals/ W
sample-41/actuals/ W
sample-42/actuals/ W
sample-43/rng/     L
sample-44/rng/     W
sample-45/actuals/ L
sample-46/actuals/ W
sample-47/rng/     W
sample-48/rng/     L
sample-49/actuals/ W
sample-50/actuals/ L
sample-51/actuals/ W
sample-52/actuals/ W
sample-53/actuals/ L
sample-54/actuals/ W
sample-55/actuals/ W
sample-56/rng/     W
sample-57/rng/     W
sample-58/rng/     L
sample-59/rng/     L
sample-60/actuals/ W
sample-61/rng/     W
sample-62/actuals/ W
sample-63/actuals/ L
sample-64/actuals/ L
sample-65/rng/     L
sample-66/rng/     W
sample-67/rng/     L
sample-68/actuals/ W
sample-69/actuals/ L
sample-70/rng/     L
sample-71/rng/     L
sample-72/rng/     L
sample-73/rng/     L
sample-74/actuals/ L
sample-75/rng/     W
sample-76/rng/     W
sample-77/actuals/ W
sample-78/rng/     W
sample-79/actuals/ L
sample-80/rng/     L
sample-81/actuals/ L
sample-82/rng/     W
sample-83/actuals/ W
sample-84/rng/     W
sample-85/rng/     L
sample-86/actuals/ W
sample-87/rng/     W
sample-88/actuals/ L
sample-89/actuals/ W
sample-90/actuals/ L
sample-91/actuals/ W
sample-92/rng/     W
sample-93/rng/     W
sample-94/rng/     W
sample-95/actuals/ L
sample-96/actuals/ L
sample-97/actuals/ L
sample-98/rng/     W
sample-99/actuals/ W
sample-100/rng/    W


thank you very much Bayes for a good topic...

[reddwarf]

Hi Bayes,

I'm dying of curiosity: how did our heros fare? Do you already have a significant endresult???

reddwarf

[bombus]

 
I had my crack at picking the difference between rng & actuals. The result was bang on average with 51 correct & 49 incorrect.

Using the characteristics I chose for the exercise was not an accurate measure for spotting the difference, and in hindsight we probably could have stopped after 10 or so samples.

Thanks for the topic Bayes; I have now got reams of fertile statistics to play around with. Who knows, maybe my next super-duper system will sprout from it.

Cheers.

[Bayes]

Thanks guys for participating.  

I'm not surprised at the outcome, but as bombus points out, it hasn't shown that it can't be done, only that the methods used weren't effective at discriminating between the two groups. However, any RNG worth its salt has to pass a battery of statistical tests, so personally I find it hard to believe that any simple method could be successful. True, human intuition can sometimes spot anomalies which a formal method might miss, but intuition is also sometimes TOO good at spotting patterns which aren't significant, plus you have to take into account the tendency to bias (if you're suspicious or paranoid, you tend to read too much into information).

BTW, you might be interested to know that I got the actuals from here (go to the download section).

[bombus]

Bump.

[know when to quit]

Just ran across this, thanks for posting Bayes. I know this took quite a bit of work. I had a free morning and only did the first 50. It was fun. I never realized there was such a difference between the two till now.

[know when to quit]

-