Can the future of the war in Ukraine be predicted? It’s what a community of internet tipsters is trying to do | Technology

The day after Russia invaded Ukraine, one person wrote a message of thanks on a forum: “I just want to say that I moved from kyiv to Lviv on February 13 thanks to this prediction thread and your estimates.[.] I leave Ukraine today. Thank you all”.

The message was left by the user availablegreen in Metaculus, a community dedicated to predicting the future by answering questions like the one that pushed him to leave his city: Will Russia invade Ukraine before 2023? In December, the community prediction said there was a 40% chance. And on February 13, when availablegreen left kyiv, which had 60%. By then, US intelligence was already saying that it was going to happen; many people didn’t believe it, but Metaculus decided to.

Prediction communities are booming. In Metaculus, Polymarket, Good Judgment or Insight questions are asked about everything. On politics: Will Macron win the French elections? Very likely (94%). On the pandemic: Will the WHO add a new variant of covid to the list of “worrying” in 2022? Likely (74%). Or about catastrophes: What risk is there in London of dying next month from a nuclear explosion? Some 24 micro-deaths, 24 options out of a million, according to a group of reputable forecasters of which Nuño Sempere from Madrid is a part.

Sempere, who writes a newsletter On these topics, he explained how the platforms work: “Metaculus is a group of people who think that these questions are important, and that having models of the world that are capable of making predictions is important. Imagine a collaborative community of the type Wikipedia or Reddit, that instead of writing articles or selecting interesting content, generates investigations and a probability that summarizes them”.

These predictions were popularized by three Pennsylvania professors—Philip Tetlock, Barbara Mellers, and Don Moore—who in 2013 won a competition funded by US intelligence. They showed that some people are better at doing this, whom they would later call super-forecasters, and that by adding their forecasts, a team of generalists could match or beat the CIA experts.

But the community that has emerged goes further.

How do they do that?

Forecasters use open sources. They exploit the information that is on the internet, from scientific studies, to press news or public data. As Sam Freedman has said: “Anyone on Twitter can, if they filter the information right, be better informed about the real-time course of the war than Lyndon Johnson is about Vietnam.”

In addition, the recipe is known to predict better. First, you need a certain kind of look: quantitative, probabilistic, parsimonious, willing to change your mind. Second, it helps to use aggregation methods better than a median (for example, it is convenient to extreme: if two people with different information tell you that the probability of rain is 50%, you should bet that it will rain with plus 50% chance). And third, something logical, but often forgotten: you have to have a real will to hit.

How much are they right?

In the community they are not very satisfied with their successes in this war, although it seems to me meritorious to have said in January that the invasion was something probable.

I have also followed another prediction of yours that started out unsuccessful, but moved quickly. I found it useful to organize my coverage: Would kyiv fall under Russian control before April 1? On the second day of the invasion, Metaculus saw the fall as likely (80%), as did most observers who expected a rapid advance by the Russians. But they soon corrected.

On the fifth day the probability had dropped to 67%, on March 6 to 37% and on the 15th, two weeks before the deadline, it was only 10%. Even more interesting: now they believe that the probability that the capital will hold out for another two months, until June 1, is 80-85%. Will they succeed?

A success of Metaculus has been his predictions during the covid pandemic, as he explained John Cambeiro from New York, who I met when he was leading the ranking of the best tipsters, and who now works for the platform. “By December 2, the Metaculus forecasters successfully predicted that omicron would rapidly overtake delta.” Also that it had “intrinsically higher transmissibility, which would erode the protection conferred by vaccines and even be less lethal than delta.”

Of course, the people of Metaculus are far from infallible oracles. They almost never offer strong predictions. But they have been proven correct in one key sense: they are well calibrated. Events to which they assign a probability of 60% occur 60% of the time (more or less), and those with probability 90% occur 90% (more or less).

Why do they do it?

There are platforms where you can predict to earn money or cryptocurrency, but the main reasons, at the moment, seem to be hobby and commitment. If the internet has taught us anything, it is that people can spend a lot of time on their interests.

For Sempere it is not a game: “Sports betting seems absurd to me and I don’t see the point in it. They are structured to generate addiction.” He also predicts in contests and paid markets, but finds advantages and disadvantages: “It allows you to invest more effort. But collaborating is more difficult.”

The key is that both Sempere and Cambeiro believe their work is useful and has a lot of potential. We all have to make quick decisions and under uncertainty. This is obvious for a mayor or a manager, but it is the same for the man who closed his bar in the pandemic or for the young woman who decides not to buy a house for fear of the recession. These platforms, as Sempere says, can produce probabilities about “how long a quarantine is going to last, or about who is going to be the next president.” They won’t decide for you, but they can shed light and inform your decisions. Cambeiro emphasizes that it is already happening: “Many people have made decisions with our forecasts with covid. Many users and I were taking precautions before anyone else.”

Can this be the case availablegreen? I can’t be sure it’s true, although I’ve talked to him and it sure seems like it. A young Belarusian who lived in kyiv and who in February, already worried, remembered having read about these platforms (“I went to see what the prediction markets were saying”). Trusting in them, and in what he saw in New York Times and on Twitter, he made a decision: to leave the city for Lviv, from where he later traveled to Warsaw.

You can follow EL PAÍS TECNOLOGÍA at Facebook and Twitter or sign up here to receive our weekly newsletter.