I’d also like to advertise a challenge for my readers. You can email me with your predictions for a subset of my predictions with your own prediction. I’ll judge your predictions against mine using the logarithmic scoring rule.
Out of curiosity, why logarithmic scoring and not Brier scoring? (I like logarithmic scoring better, but you used Brier in the pseudorandomness contest.)
Would you also take money bets in addition to just virtual scores?
I like logarithmic better in general, but I decided to use Brier for the pseudorandomness contest because I decided I really cared about the difference between a 60% chance (i.e. looks basically random) and a 40% chance (kind of suspect). The log rule is better at rewarding people for being right at the extremes; Brier's rule is better at rewarding people for being right in the middle.
Regarding bets: I'm willing to make bets, but won't have a blanket policy like "I'll take a bet with anyone who disagrees with me by 10% or more", because that opens me up to a ton of adverse selection. (E.g. I wouldn't bet with Zvi on COVID.) So... feel free to message me if you want to bet, but also be aware that the most likely outcome is that it won't result in a bet.
(Also, the better I know you, the more likely I am to be willing to bet with you.)
Just as I did last year, I have some probabilistic predictions for 2021. In January 2022 I will return to grade them, just as in a week or two I’ll grade my 2020 predictions. This year the predictions fall into four categories: U.S. politics (#1-17 below), COVID (#18-39), Miscellaneous (#40-53), and Personal (#54-100).
Note that these predictions are not necessarily well-informed. You should consider my U.S. politics predictions to be more or less reasonable, but it’s totally possible that my COVID predictions are crazy. This is just to say: if you think you have decent COVID information, don’t use my predictions to inform your decisions.
As last year, I’ll be assessing my predictions on calibration (the details are the same as last year, except that [0.9, 1) will be one bucket). I will also be evaluating the predictions on personal pessimism/optimism — whether I’m too optimistic, too pessimistic, or about right in my judgment of what I will get done next year — this time with weights proportional to the importance of each goal (see below).
I’d also like to advertise a challenge for my readers. You can email me with your predictions for a subset of my predictions with your own prediction. I’ll judge your predictions against mine using the logarithmic scoring rule. I have the advantage of having chosen the questions, but you have the advantage of having seen my predictions. In theory this gives you the upper hand: for example, you could put down my probabilities for every event except a few where you think I clearly messed up. Consider this an exercise in second-order knowledge: figuring out how much weight to put on my probabilities despite not knowing my reasoning behind them. Your score will be the sum, over all questions you choose, of your log score minus my log score.1 (So you can guarantee yourself a score of 0 by sending me an empty list.) [Edit: the “choose at least 40 predictions” requirement has been removed.]
Please make your predictions by January 4th at 11:59pm ET and send them to me as a text file where each line has the format “[event #], [probability as a decimal]”. I will throw out any predictions that resolve (or nearly resolve) by January 4th.
Finally, a note: many of the probabilities I give are numbers like 34% (not multiples of 5%). Don’t take this to mean that I somehow precisely calculated the probabilities or that I’m really knowledgeable about the underlying subject; sometimes I just feel like giving two significant digits.
With all that said, here are my predictions:
* For legislation to be considered major, a substantial amount of effort/political capital needs to be spent on it. Major legislation passes on average once every 2-3 years. Examples include the 2009 stimulus bill, Obamacare, and Trump’s 2017 tax law.
18. I receive my first dose of a COVID vaccine by the end of March: 12%
19. …the end of April: 34%
20. …the end of May: 60%
21. …the end of June: 74%
22. …the end of July: 82%
23. …the end of August: 87%
24. …the end of 2021: 97%
25. At least 50% of people living in the U.S. receive at least one COVID vaccine dose by the end of 2021: 75%
26. At least 60%: 58%
27. At least 70%: 43%
28. At least 80%: 20%
29. At least 90%: 4%
30. Per official statistics, at least 100 thousand Americans die of COVID in 2021: 82%
31. …at least 200 thousand: 64%
32. …at least 500 thousand: 25%
33. …at least 1 million: 8%
34. I or one of the seven people I share 25% of my genes with tests positive for COVID: 30%
35. I test positive for COVID: 5%
36. I go to my office at Columbia at least once by the end of May: 32%
37. EC is held at least partially in Budapest: 36%
38. Canada/USA Mathcamp is held in person (I have no inside information on this): 25%
39. SPARC is held in person (I have no inside information on this): 38%
40. China is involved in an international (counting Taiwan and Hong Kong) conflict that has 1,000 casualties: 9%
41. A normalization of relationships between Israel and at least one majority-Muslim country is initiated in 2021 during the Biden administration: 48%
42. Putin is the president of Russia at the end of 2021: 88%
43. Benjamin Netanyahu is the Prime Minister of Israel at the end of 2021: 57%
44. Scott Alexander starts publishing again: 85%
45. Taylor Swift releases her tenth studio album: 65% (45b, won’t be graded — Taylor Swift releases her eleventh studio album: 13%)
46. “Foklore” wins a Grammy for Album of the Year: 61%
47. The third book in the Kingkiller Chronicle has a publication date set by the end of 2021 (the date doesn’t have to be in 2021): 13%
48. Roger Federer wins a grand slam tournament in 2021: 26%
49. Someone besides Djokovic, Nadal, and Federer wins a men’s singles grand slam tournament in 2021: 59%
50. Serena Williams wins a grand slam tournament in 2021: 32%
51. All women’s singles grand slam tournaments in 2021 are won by different people: 68%
52. P vs. NP is widely considered resolved by the end of 2021: 1%
53. A (non-trivial) update on GPT-3 is released: 62%
54. I summarize for my blog, or review for a journal, at least 20 papers: 75%
55. …at least 30 papers: 65%
56. …at least 40 papers: 48%
57. …at least 50 papers: 20%
58. I attend EC (counts if I go to at least five talks): 74%
59. The paper I’m writing on aggregating predictions is accepted to a conference or journal: 73%
60 I write and submit a paper on prediction aggregation and online learning (this is a different one from the one in #59): 77%
61. …and that paper is accepted: 47%
62. I resolve the “preventing arbitrage from collusion” scoring rules problem: 40%
63. My scoring rules paper from a while ago finally gets accepted somewhere: 55%
64. I publish, or begin writing with the intention to publish, a paper following up directly on “No-Regret and Incentive Compatible Online Learning”: 45%
65. I publish a computer science paper in a conference held in 2021 or a journal edition issued in 2021: 85%
66. I publish a paper or note on Zipf’s law: 28%
67. I write 10 or more blog posts in 2021: 92%
68. I write 20 or more blog posts in 2021: 78%
69. I write 30 or more blog posts in 2021: 50%
70. I write 50 or more blog posts in 2021: 9%
71. The total number of views of my blog in 2021 is at least 5,000: 95%
72. The total number of views of my blog in 2021 is at least 10,000: 83%
73. The total number of views of my blog in 2021 is at least 20,000: 64%
74. The total number of views of my blog in 2021 is at least 50,000: 27%
75. The total number of views of my blog in 2021 is at least 100,000: 11%
76. (Intentionally vague to avoid spoliers) I write a blog post about big aliens: 42%
77. I publish a blog post on setting the right price: 36%
78. I publish a blog post about Pi: 33%
79. I publish a blog post about slowly converging series: 40%
80. I publish a blog post on Zipf’s law: 80%
81. I publish a blog post on Bayesian injustice: 25%
82. I vote in the Democratic primary of the New York mayoral election: 93%
83. I rank Andrew Yang first in the Democratic primary of the New York mayoral election: 51%
84. I stick to my virtue points system, or some variation, through the end of 2021: 70%
85. I’m a SPARC staff member in 2021: 31%
86. I’m a Mathcamp mentor in 2021: 55%
87. I (co-)run some OBNYC (NYC rationalist) meetup in 2021: 47%
88. I consider myself a vegetarian at the end of 2021: 29%
89. I consider myself a vegan at the end of 2021: 4%
90. I make a donation of at least $500 to a third world poverty charity in 2021: 66%
91. I make a donation of at least $500 to existential risk/long-term in 2021: 79%
92. I make a donation of at least $500 to animal welfare in 2021: 23%
93. I have a tentative plan to take a gap year (or I take a gap year): 24%
94. I play squash on at least 10 days in 2021: 65%
95. I play squash on at least 20 days in 2021: 44%
96. I visit a country that is not Hungary: 23%
97. I publish a non-academic piece of writing in some publication in 2021: 33%
98. I read a book in 2021: 65%
99. I read at least two books in 2021: 44%
100. I read at least three books in 2021: 30%
1. That is, for each question, if I assign probability p to the outcome that ends up happening and you assign probability q, your score will be ln(q) - ln(p). For instance, if you say that Ossoff has a 20% chance of winning and he ends up losing, your score will be ln(0.8) - ln(0.55) (since I assigned a 0.55 chance to Ossoff losing).