Predictions for 2021

Every year for the last few years, my local rationalist meetup (in Ottawa, Canada) has done a predictions night at the beginning of the year where we talk about the year to come, estimate and debate the probability of various events, and “score” our previous year’s predictions. Think something like Slate Star Codex publishes every year, only much less rigorous.

Historically we’ve done these events in person, on various scraps of paper, and it’s been pretty hit or miss: sometimes we’ve ended up with slightly different sets of questions, sometimes people can’t find their sheet the next year, etc. This year, because of the pandemic, we did it online in a Google Doc, so we’re theoretically in much better shape (though in hindsight it really should have been a spreadsheet, not a document; I spent way too long organizing it and formatting it after the fact).

For me, I think the majority of the value was in the discussions that arose ad-hoc when people put down divergent predictions. However, the predictions themselves are also pretty interesting in some cases, so with the permission of the group I’m publishing the anonymized predictions of Rational Ottawa for 2021, in five categories: COVID, politics, technology, media, and miscellaneous. A few notes before we get started:

  • We tried to be make clear, testable, and realistic predictions but we weren’t particularly rigorous about this. I’ve already found a few that are ambiguous as I’ve gone through reformatting them.
  • Some predictions ended up in “Miscellaneous” that really belong in other sections, just because of when we thought of them.
  • A few of these predictions are inside jokes but they’re hopefully pretty obvious (e.g. “Finland really doesn’t exist”).
  • I’ve pruned the personal section entirely rather than trying to redact individual questions to varying levels of comfort.
  • I’m not personally concerned about my anonymity in making these predictions, so I’m happy to note that I’m person #1.

COVID

Long-term health effects from those infected with COVID are common by EOY.
1: 10%
2: 35%
3: 15%
4: 65%
5: 15%
6: 50% 

Publicly funded economic aid for people with long-term effects from covid.
1: 5%
2: 7%
3: 5%
4: 10%
5: 3%
6: 15%

50% of US population vaccinated by end of summer.
1: 5%
2: 12%
3: 40%
4: 10%
5: 10%
6: 10%

50% of Canadian population vaccinated by end of summer.
1: 20%
2: 50%
3: 60%
4: 35% 
5: 70%
6: 30% 

50% of World population vaccinated by end of summer.
1: 10%
2: 5%
3: 1%
4: 5%
5: 2%
6: 2%

50% of US population vaccinated by end of year.
1: 65%
2: 50%
3: 60%
4: 55%
5: 25%
6: 65%

50% of Canadian population vaccinated by end of year.
1: 80%
2: 85%
3: 80%
4: 75%
5: 85%
6: 70%

50% of World population vaccinated by end of year.
1: 20%
2: 30%
3: 5%
4: 25%
5: 20%
6: 15%

Substantial subpopulation (>5,000) has ill-effects from vaccine.
1: 40%
2: 70%
3: 20%
4: 50%
5: 35%
6: 70% 

Mask mandate lifted in Ottawa by EOY.
1: 80%
2: 55%
3: 70%
4: 70% 
5: 80%
6: 50%

English (or other) variation overtakes Covid Original in case count by end of spring/summer/year.
1: 1/40/80 %
2: 1/30/55 %
3: 1/25/35%
4: 2/20/55%
5: 3/20/65%
6: 1/10/70%

More than 30k total COVID deaths in Canada by EOY (i.e. more in 2021 than 2020).
1: 60%
2: 35%
3: 30%
4: 60%
5: 45%
6: 50%

uOttawa Fall term live
1: 85%
2: 70%
3: 75%
4: 75%
5: 80%
6: 20%

Biden, McConnell, or Pelosi get COVID.
1: 10%
2: 10%
3: 10%
4: 10%
5: 10%
6: 5%

Someone in Rational Ottawa is diagnosed with covid by EOY.
1: 30%
2: 15%
3: 30%
4: 19%
5: 20%
6: 18%

No significant social distancing in Ottawa by end of spring/summer/year.
1: 5/40/90 %
2: 2/35/65 %
3: 5/50/85%
4: 1/40/85%
5: 10/70/90%
6: 1/40/60%

Senators playing in front of full live audiences by EOY.
1: 95%
2: 80%
3: 80%
4: 85%
5: 85%
6: 60%

Predict a specific date when Rational Ottawa has an unsocially distanced indoor meeting again:
1: September 10th, 2021.
2: November 12th, 2021
3: August 22nd, 2021.
4: September 17th, 2021.
5: September 3rd, 2021.
6: December 3rd, 2021

Most office work in Ottawa no longer done from home by EOY.
1: 75%
2: 65%
3: 80%
4: 49%
5: 75%
6: 65%

The coronavirus crosses the species barrier (as with the Danish mink) and mutates into something that complicates the situation
1: 5%
2: 7%
3: 10%
4: 5%
5: 5%
6: 20% 

Politics

Canadian election called.
1: 15%
2: 15%
3: 30%
4: 35%
5: no clue
6: no clue

Trump still living in US by EOY.
1: 90%
2:  85%
3: 90%
4: 80%
5: 90%
6: 80%

Republicans win both Georgia senate seats.
1: 50%
2: 40%
3: 40%
4: 50%
5: 50%
6: 70%

Democrats win both Georgia senate seats.
1: 20%
2: 40%
3: 45%
4: 25%
5: 35%
6: 30%

Split Georgia senate seats.
1: 30%
2: 20%
3: 15%
4: 25%
5: 15%
6: 45%

Evidence of Mitch McConnell doing cocaine?
1: <1%
2: 1%
3: 1%
4: 2%
5: <1%
6: 5%

Kamala Harris president.
1: 2%
2: 4%
3: 5%
4: 5%
5: 5%
6: 35% 

Ontario election called.
1: 5%
2: 2%
3: 10%
4: 10%
5: no clue
6: no clue

Doug Ford reelected, assuming Ontario election is called.
1: 80%
2: 70%
3: 60%
4: 70%
5: no clue
6: 75%

The Queen yet lives.
1: 95%
2: 85%
3: 85%
4: 85%
5: 90%
6: 90%

New Scottish independence referendum scheduled.
1: 30%
2: 20%
3: 20%
4: 40%
5: 15%
6: 60%

Irish unification referendum scheduled.
1: 5%
2: 10%
3: 2%
4: 5%
5: 30%
6: No clue

NDP remains short on cash by EOY.
1: 70%
2: 50%
3: 60%
4: 60%
5: 60%
6: 50%

Finland really doesn’t exist.
1: -5%
2: 1%
3: 200%
4: 1000%
5: 50%
6: no clue

Technology

VR finally gets significant mainstream uptake.
1: 10%
2: 15%
3: 10%
4: 15%
5: 10%
6: 14%

Quantum computers used for something useful that we couldn’t cheaply do otherwise.
1: <1%
2: 10%
3: 10%
4: 10%
5: 15% 
6: 30%

SpaceX successfully launches 8+ more humans to space.
1: 30%
2: 60%
3: 15%
4: 90%
5: 20%
6: 20%

SN9 starship test RUD
1: 50%
2: 30%
3: 30%
4: 50%
5: 35%
6: 40%

A SpaceX Starship makes it to orbit by EOY
1: 30%
2: 15%
3: 20%
4: 20%
5: 30%
6: 40%

GPT-4 exists
1: 80%
2: 50%
3: 40%
4: 60%
5: 60%
6: 50%

Boston Dynamics starts leasing out atlas as well as spot by EOY
1: 5%
2: 10%
3: 30%
4: 15%
5: 20%
6: 25%

Driverless shipping-trucks approved (on public highways) in some country
1: 30%
2: 50%
3: 20%
4: 75%
5: 50%
6: 60%

Uber owns a driverless fleet of over 200/2000 vehicles in operation
1: 5/1 %
2: 15%/1%
3: 5%/3%
4: 10/2%
5: 5/1%
6: 5/1%

Alpha (AI) used in medical assessment, as a tool
1: <1%
2: 20%
3: 1%
4: 99%
5: 20%
6: 50%

Go AI substantially better than any human player.
1: 90%
2: 95%
3: 99%
4: 99%
5: 99%
6: 90%

Tesla’s stock price higher on 2021/12/31 than on 2021/01/01.
1: 60%
2: 55%
3: 40%
4: 65%
5: 60%
6: 

Bitcoin value higher on 2021/12/21 than 2021/01/01.
1: 60%
2: 50%
3: 30%
4: no idea
5: 40%
6: 30%

OCtranspo electric bus trial goes well.
1: 60%
2: 70%
3: 80%
4: 75%
5: 70%
6: 40%

AI goes wrong somehow.
Everyone: 99.999%

Death by human-out-of-the-loop drone is publicized
1: 20%
2: 10%
3: 10%
4: 15%
5: 15%
6: 25%

Media

SSC substack launched in January as predicted.
1: 90%
2: 80%
3: 70%
4: 80%
5: 50%
6:  ¯\_(ツ)_/¯

NYT publishes an article on Scott/SSC/rationalists.
1: 20%
2: 10%
3: 20%
4: 40%
5: 15%
6: 50%

Dune releases in theatres only
1: 20%
2: 35%
3: 30%
4: 45%
5: 20%
6: 60%

Cinemas see drastic decline and new movies are released for streaming/VOD relative to pre-pandemic
1: 70%
2: 85%
3: 90%
4: 90%
5: 90%
6: 80%

Miscellaneous

[I] will subjectively rate the influence of the far left/far right as lower by EOY
1: 20/20%
2: 35%/30%
3: 40%/90%
4: 50/40%
5: 30/50%
6: 50/40%

At some point during 2021, one of the questions on this list will be rendered obsolete or inapplicable by changing circumstances and will need to be unasked.
1: 30%
2: 50%
3: 90%
4: 90%
5: 15%
6: 100%

Canadian airline closes forever
1: 10%
2: 5%
3: 5%
4: 5%
5: 5%
6: 5%

November/December 2021 is colder on average than in 2020
1: 50%
2: 50/50%
3: 60/80%
4: 85/95%
5: 80/80%
6: 40/40%

Enough snow for cross-country skiing on Christmas 2021
1: 30%
2: 20%
3: 40%
4: 40%
5: 30%
6: 10%

Hugging and kissing is practiced more than before the pandemic
1: 60%
2: 50%
3: 60%
4:…in rational ottawa? All the hugs 99%
5: 30%
6: 75%

Discarded non reusable masks become a waste problem in underdeveloped countries
1: 5%
2: 20%
3: 5%
4: 5%
5: 3%
6: 50%

The Stopped Clock Problem

[Unusually for me, I actually wrote this and published it on Less Wrong first. I’ve never reverse-cross-posted something to my blog before.]

When a low-probability, high-impact event occurs, and the world “got it wrong”, it is tempting to look for the people who did successfully predict it in advance in order to discover their secret, or at least see what else they’ve predicted. Unfortunately, as Wei Dai discovered recently, this tends to backfire.

It may feel a bit counterintuitive, but this is actually fairly predictable: the math backs it up on some reasonable assumptions. First, let’s assume that the topic required unusual levels of clarity of thought not to be sucked into the prevailing (wrong) consensus: say a mere 0.001% of people accomplished this. These people are worth finding, and listening to.

But we must also note that a good chunk of the population are just pessimists. Let’s say, very conservatively, that 0.01% of people predicted the same disaster just because they always predict the most obvious possible disaster. Suddenly the odds are pretty good that anybody you find who successfully predicted the disaster is a crank. The mere fact that they correctly predicted the disaster becomes evidence only of extreme reasoning, but is insufficient to tell whether that reasoning was extremely good, or extremely bad. And on balance, most of the time, it’s extremely bad.

Unfortunately, the problem here is not just that the good predictors are buried in a mountain of random others; it’s that the good predictors are buried in a mountain of extremely poor predictors. The result is that the mean prediction of that group is going to be noticeably worse than the prevailing consensus on most questions, not better.


Obviously the 0.001% and 0.01% numbers above are made up; I spent some time looking for real statistics and couldn’t find anything useful; this article claims roughly 1% of Americans are “preppers”, which might be a good indication, except it provides no source and could equally well just be the lizardman constant. Regardless, my point relies mainly on the second group being an order of magnitude or more larger than the first, which seems (to me) fairly intuitively likely to be true. If anybody has real statistics to prove or disprove this, they would be much appreciated.

What is a “Good” Prediction?

Zvi’s post on Evaluating Predictions in Hindsight is a great walk through some practical, concrete methods of evaluating predictions. This post aims to be a somewhat more theoretical/philosophical take on the related idea of what makes a prediction “good”.

Intuitively, when we ask whether some past prediction was “good” or not, we tend to look at what actually happened. If I predicted that the sun will rise with very high probability, and the sun actually rose, that was a good prediction, right? There is an instrumental sense in which this is true, but also an epistemic sense in which it is not. If the sun was extremely unlikely to rise, then in a sense my prediction was wrong – I just got lucky instead. We can formally divide this distinction as follows:

  • Instrumentally, a prediction was good if believing it guided us to better behaviour. Usually this means it assigned a majority probability to the thing that actually happened regardless of how likely it really was.
  • Epistemically, a prediction was good only if it matched the underlying true probability of the event in question.

But what do we mean by “true probability”? If you believe the universe has fundamental randomness in it then this idea of “true probability” is probably pretty intuitive. There is some probability of an event happening baked into the underlying reality, and like any knowledge, our prediction is good if it matches that underlying reality. If this feels weird because you have a more deterministic bent, then I would remind you that every system seems random from the inside.

For a more concrete example, consider betting on a sports match between two teams. From a theoretical, instrumental perspective there is one optimal bet: 100% on the team that actually wins. But in reality, it is impossible to perfectly predict who will win; either that information literally doesn’t exist, or it exists in a way which we cannot access. So we have to treat reality itself as having a spread: there is some metaphysically real probability that team A will win, and some metaphysically real probability that team B will win. The bet with the best expected outcome is the one that matches those real probabilities.

While this definition of an “epistemically good prediction” is the most theoretically pure, and is a good ideal to strive for, it is usually impractical for actually evaluating predictions (thus Zvi’s post). Even after the fact, we often don’t have a good idea what the underlying “true probability” was. This is important to note, because it’s an easy mistake to make: what actually happened does not tell us the true probability. It’s useful information in that direction, but cannot be conclusive and often isn’t even that significant. It only feels conclusive sometimes because we tend to default to thinking about the world deterministically.


Eliezer has an essay arguing that Probability is in the Mind. While in a literal sense I am contradicting that thesis, I don’t consider my argument here to be incompatible with what he’s written. Probability is in the mind, and that’s what is usually more useful to us. But unless you consider the world to be fully deterministic, probability must also be in the world – it’s just important to distinguish which one you’re talking about.