The Art Contest has effects on merit distribution. How about merit sources?

Note:
I am going to keep update the thread in next 2 months to see which sort of changes of merit distributions from merit sources (especially when bias-effects from the Art contest cool down - after the vote, after congratulations.

ABSTRACT

The forum users distributed 8659 merits more what they usually did in most of half previous weeks at 4521 (the median of weekly merits).
There is nearly 292% increase in the total merits last week, compares to median of weekly merits (before 2019w46) at 4521.

We all know that theymos sent thousands of merits in his merit sprees.

From my discussion with @fillippone, I ask myself that are there any significant effects from increased reallocation of merits to merit sources?

In my Merit source observations, you can see despite of decrease in total merit sources (from 130 to 90 then increased to 98), the total merits per 30 days rose (from 20895 to 21671).

Quote from: tranthidung on September 02, 2019, 11:22:35 AM

Code:

. list id day month2 year date meritsource smerit, abb(30)

   +-------------------------------------------------------------+
   | id day month2 year date meritsource smerit |
   |-------------------------------------------------------------|
  1. | 1 24 1 2018 24jan2018 35 8125 |
  2. | 2 26 1 2018 26jan2018 49 9025 |
  3. | 3 1 3 2018 01mar2018 57 11975 |
  4. | 4 12 3 2018 12mar2018 77 17650 |
  5. | 5 22 3 2018 22mar2018 80 17800 |
   |-------------------------------------------------------------|
  6. | 6 18 5 2018 18may2018 80 18500 |
  7. | 7 6 7 2018 06jul2018 82 18900 |
  8. | 8 11 7 2018 11jul2018 83 19100 |
  9. | 9 17 9 2018 17sep2018 120 23045 |
10. | 10 17 11 2018 17nov2018 119 22045 |
   |-------------------------------------------------------------|
11. | 11 22 12 2018 22dec2018 120 22055 |
12. | 12 5 1 2019 05jan2019 123 20735 |
13. | 13 6 2 2019 06feb2019 122 20835 |
14. | 14 15 3 2019 15mar2019 130 20605 |
15. | 15 27 6 2019 27jun2019 131 21045 |
   |-------------------------------------------------------------|
16. | 16 30 10 2019 30oct2019 130 20895 |
17. | 17 14 11 2019 14nov2019 90 18821 |
18. | 18 15 11 2019 15nov2019 98 21671 |
   +-------------------------------------------------------------+

I scrap data from total merits theymos sent in the last week (only last week) in the 10^th anniversary Art contest.

He sent 9413 merits in total, from 18/11/2019 to 22/11/2019, forum time; but I stop counting his sent merits after November 22, 2019, 09:37:48 PM (when he locked that thread).

Code:

. list summerits date totalmerits, abb(30)

   +--------------------------------+
   | summerits date totalmerits |
   |--------------------------------|
  1. | 1072 18 9413 |
  2. | 2272 19 9413 |
  3. | 935 20 9413 |
  4. | 2877 21 9413 |
  5. | 2257 22 9413 |
   +--------------------------------+

Those days fit well in the period of the last week, 2019w46, that has 14251 merits distributed in total (see in the 3^rd in column of weeklyrank.

The 2019w46 is from 12/11/ to 18/11/2019,, in GMT time (not forum time). So there is a little lag (hours) between two datasets. However, I don't think there are significant difference (hope that I am not wrong). See

Code:

. list weeklyrank merit week, abb(30)

   +------------------------------+
   | weeklyrank merit week |
   |------------------------------|
  1. | 1 30960 2018w4 |
  2. | 2 19979 2018w5 |
  3. | 3 14251 2019w46 |
  4. | 4 13313 2018w6 |
  5. | 5 11745 2018w7 |
   |------------------------------|
  6. | 6 8833 2018w9 |
  7. | 7 8767 2018w8 |
  8. | 8 7837 2018w38 |
  9. | 9 7317 2018w11 |
10. | 10 7261 2018w10 |
   |------------------------------|
11. | 11 6952 2018w12 |
12. | 12 6744 2018w13 |
13. | 13 6632 2019w2 |
14. | 14 6423 2018w14 |
15. | 15 6130 2019w13 |
   |------------------------------|
16. | 16 5644 2018w37 |
17. | 17 5542 2019w42 |
18. | 18 5494 2018w15 |
19. | 19 5454 2019w19 |
20. | 20 5354 2019w24 |
   |------------------------------|
21. | 21 5317 2019w3 |
22. | 22 5271 2019w15 |
23. | 23 5214 2019w20 |
24. | 24 4975 2019w43 |
25. | 25 4965 2018w18 |
   |------------------------------|
26. | 26 4929 2018w25 |
27. | 27 4913 2019w10 |
28. | 28 4829 2018w42 |
29. | 29 4803 2019w1 |
30. | 30 4766 2018w19 |
   |------------------------------|
31. | 31 4764 2019w18 |
32. | 32 4742 2018w16 |
33. | 33 4735 2019w45 |
34. | 34 4730 2019w44 |
35. | 35 4726 2019w25 |
   |------------------------------|
36. | 36 4688 2019w16 |
37. | 37 4687 2019w23 |
38. | 38 4667 2019w4 |
39. | 39 4638 2019w9 |
40. | 40 4612 2018w17 |
   |------------------------------|
41. | 41 4609 2019w12 |
42. | 42 4580 2019w21 |
43. | 43 4575 2018w47 |
44. | 44 4565 2019w41 |
45. | 45 4538 2018w23 |
   |------------------------------|
46. | 46 4526 2019w14 |
47. | 47 4525 2018w45 |
48. | 48 4521 2019w8 |
49. | 49 4520 2019w38 |
50. | 50 4491 2019w5 |
   |------------------------------|
51. | 51 4465 2018w26 |
52. | 52 4448 2019w17 |
53. | 53 4445 2019w22 |
54. | 54 4395 2018w39 |
55. | 55 4367 2019w26 |
   |------------------------------|
56. | 56 4357 2019w40 |
57. | 57 4353 2018w20 |
58. | 58 4332 2019w6 |
59. | 59 4326 2019w11 |
60. | 60 4318 2019w39 |
   |------------------------------|
61. | 61 4310 2018w40 |
62. | 62 4278 2018w27 |
63. | 63 4277 2019w29 |
64. | 64 4247 2018w28 |
65. | 65 4236 2019w33 |
   |------------------------------|
66. | 66 4225 2019w27 |
67. | 67 4221 2019w7 |
68. | 68 4194 2018w22 |
69. | 69 4176 2019w30 |
70. | 70 4167 2018w29 |
   |------------------------------|
71. | 71 4119 2019w28 |
72. | 72 4043 2019w37 |
73. | 73 4011 2018w32 |
74. | 74 3953 2018w43 |
75. | 75 3864 2018w21 |
   |------------------------------|
76. | 76 3863 2018w31 |
77. | 77 3839 2018w24 |
78. | 78 3816 2018w41 |
79. | 79 3809 2019w36 |
80. | 80 3805 2018w50 |
   |------------------------------|
81. | 81 3805 2018w34 |
82. | 82 3769 2018w51 |
83. | 83 3765 2018w48 |
84. | 84 3747 2018w46 |
85. | 85 3661 2018w30 |
   |------------------------------|
86. | 86 3631 2018w33 |
87. | 87 3622 2019w34 |
88. | 88 3590 2018w36 |
89. | 89 3571 2018w49 |
90. | 90 3549 2019w31 |
   |------------------------------|
91. | 91 3540 2019w35 |
92. | 92 3347 2018w44 |
93. | 93 3338 2018w52 |
94. | 94 3207 2019w32 |
95. | 95 3072 2018w35 |
   +------------------------------+

What does it mean?
Only merits theymos sent on 18/11/2019 belong to the last week.

The median of weekly merits on 2019w45 (before the last week) is 4521.

Quote from: tranthidung on November 18, 2019, 03:14:04 AM

4521 is the median - p50.
< snip >

Code:

. tabstat merit, s(n mean sd p50 p25 p75 min max) format(%9.1f)

variable | N mean sd p50 p25 p75 min max
-------------+--------------------------------------------------------------------------------
merit | 94.0 5290.2 3476.3 4520.5 4043.0 4965.0 3072.0 30960.0
----------------------------------------------------------------------------------------------

To sum up:

In 2019w46

Total distributed merits: 14251
Total merits sent by theymos: 1071
Median of weekly merits before the last week: 4521

A raw caculation from merits (with all possible effects from Art contest - excludes theymos' sent-merits, new merit sources, newly increased sMerits allocation, and the others), we have:

Code:

. di 14251-1071-4521
8659

Unbelievable!

It means that the forum community were extremely active last week in the way they use their merits. They sent 8659 merits more than they usually did in previous weeks (at 4521, the weekly median).