Superforecasters’ Toolbox: Fermi-ization in Forecasting

Written by

Superforecasters’ Toolbox: Fermi-ization in Forecasting

Although usually a very private person, Superforecaster Peter Stamp agreed to be interviewed by a major Polish daily, Rzeczpospolita, on Good Judgment’s request. The reporter started the interview with a pop quiz. He asked Peter to estimate the number of tram cars that serve the city of Warsaw, Poland’s capital. Without using the internet, or having ever been to Warsaw, in under three minutes Peter came up with a remarkably accurate answer (only 10% away from the actual number, according to the reporter, Marek Wierciszewski). All he needed to know for his calculations were the typical size of a Warsaw tram and the relative importance of this means of transportation.

The method Peter used was Fermi-ization, and it is one of the key techniques Superforecasters employ to tackle complex questions even with minimal information.

What Is Fermi-ization?

In his day, physicist Enrico Fermi (1901-1954) was known not only for his groundbreaking contributions to nuclear physics. He was also able to come up with surprisingly accurate estimates using scarce information. The technique he used was elegant in its simplicity: He would break down grand, seemingly intractable questions into smaller sub-questions or components that could be analyzed or researched. He would then make educated guesses about each component until he arrived at his final estimate.

Many science and engineering faculties today teach this method, including through assignments like “estimate the number of square inches of pizza the students will eat during one semester.” Instead of blurting out a random number, students are expected to break the question down into smaller bits and engage with each one to produce a thoughtful answer (in this example, the estimate would depend on such factors as the number of students, the number of pizzas a student would eat per week, and the size of an average pizza).

Fermi-ization is a valuable tool in a Superforecaster’s toolbox. Since the days of the original Good Judgment Project and continuing in Good Judgment Inc’s work today, Superforecasters have proved the usefulness of this technique in producing accurate forecasts on seemingly impossible questions—from the scale of bird-flu epidemics, oil prices, and interest rates to election outcomes, regional conflict, and vaccinations during the Covid-19 pandemic.

Uses of Fermi-ization in Forecasting

In their seminal book Superforecasting, Philip Tetlock and Dan Gardner list Fermi-ization as the second of the Ten Commandments for Aspiring Superforecasters. This placement is not a coincidence. In the world of Superforecasters—experts known for their consistently accurate forecasts—Fermi-ization is a fundamental tool, enabling them to arrive at accurate predictions even in response to questions that initially seem impossible to quantify.

“Channel the playful but disciplined spirit of Enrico Fermi,” Tetlock and Gardner write. “Decompose the problem into its knowable and unknowable parts. Flush ignorance into the open. Expose and examine your assumptions. Dare to be wrong by making your best guesses. Better to discover errors quickly than to hide them behind vague verbiage.”

Depending on the question, this process can take just a few minutes, as it did when Peter worked out an estimated number of Warsaw’s tram cars, or it could be methodical, slow, and painstaking. But it is an invaluable road map whether accuracy is the goal.

Fermi-ization in forecasting has multiple uses:

    • It helps the forecaster to avoid the classic cognitive trap of relying on quick-and-easy—and often incorrect!—answers where more thought is called for.
    • It forces the forecaster to sort the relevant components from the irrelevant ones.
    • It enables the forecaster to separate the elements of the question that are knowable from those that are unknowable.
    • It makes the forecasters examine their assumptions more carefully and pushes them toward making educated—rather than blind—guesses.
    • It informs both the outside and the inside view in approaching the question.


Three Steps in Fermi-ization

Fermi-ization becomes easier and increasingly effective with practice. Keep these three steps in mind as you give it a try.

    1. Unpack the question by asking, “What would it take for the answer to be yes? What would it take for it to be no?” or “What information would allow me to answer the question?”
    2. Give each scenario your best estimate.
    3. Dare to be wrong.


Not the Only Tool

Of course, Fermi-ization is not the only tool in a Superforecaster’s toolbox. Mitigation of cognitive biases, ability to recognize and minimize noise, being actively open-minded, and keeping scores are all crucial components of the Superforecasting process. You can learn these techniques during one of our Superforecasting Workshops, or you can pose your own questions for Superforecasters to engage with through a subscription to FutureFirst™.

Superforecasting® the Fed: A Look Back over the Last 6 Months

Written by

Superforecasting® the Fed: A Look Back over the Last 6 Months

 

The Federal Reserve’s target range for the federal funds rate is the single most important driver in financial markets. Anticipating inflection points in the Fed’s policy has immense value, and Good Judgment’s Superforecasters have been beating the futures markets this year, signaling the Fed would continue to hike until the June pause while markets and experts alike flipflopped on their calls.

  • Ahead of the Fed’s March meeting, when Silicon Valley Bank went under, the futures markets priced out a hike and began flirting with a possibility of a cut as early as summer. Leading market observers like Goldman Sachs said the Fed would pause, and Nomura said they would start to cut at that meeting.

  • Then the futures markets priced in a pause for the May meeting. Experts like Pimco’s former chief economist Paul McCulley also prematurely predicted that the Fed would go on hold. As the date of the meeting approached, the futures markets—as well as most market participants—came to share the Superforecasters’ view that another hike was in the cards, but lagged behind the Superforecasters by nearly a month.

  • In the weeks heading into the June meeting, the futures were oscillating between a pause, a hike, and possibly even a cut. A stream of stronger economic data led experts such as Mohamed El-Erian to forecast that the Fed would continue to raise rates for at least another meeting and perhaps longer. Not the Superforecasters. They have been saying since 2 April 2023 that the Fed would most likely hit pause—a view that, once again, eventually became the consensus.

When comparing the forecasts of two groups—Good Judgment’s Superforecasters and the futures markets using the CME FedWatch Tool—for the last four Federal Reserve meetings, the Superforecasters assigned higher probabilities to the correct outcome. They were 66% more accurate than the futures (as measured by Brier scores) and had lower noise in their forecasts (as measured by standard deviation).

See our new whitepaper for details. We also provide subscribers with a full summary of all our active Fed forecasts, which is updated before and after each meeting (available on request). 

Good Judgment’s Superforecasters have been providing a clear signal on the Fed’s policy well before the futures and many market participants. Subscribers to FutureFirst™ have 24/7 access to evolving forecasts by the Superforecasters on questions that matter, including Fed policy through the rest of the year and beyond, along with a rich cross-section of other questions crowd-sourced directly from users, including questions on Ukraine, China, and the upcoming US elections.

“The Most Important Election in 2023”: Superforecasting the Vote in Turkey

Written by

“The Most Important Election in 2023”: Superforecasting the Vote in Turkey

 

The Superforecasters’ forecast on the outcome of the 2023 presidential election in Turkey as published in The Economist.

When The Economist’s “The World Ahead” issue was being prepared for publication in October 2022, Good Judgment’s professional Superforecasters were assigning a 71% probability to President Recep Tayyip Erdogan’s victory in the 2023 Turkish presidential election. Since then, we witnessed Erdogan’s increasingly unorthodox fiscal policy that resulted in high inflation and a devastating earthquake that killed more than 50,000 people. These developments led many in the West to start betting on the candidate from Turkey’s united opposition, Kemal Kilicdaroglu. Superforecasters, however, stayed the course. With a one-day dip to 50-50 just ahead of the first round, they otherwise remained consistently at 60–70% for Erdogan throughout the seven months of this question period, and they are currently at 98%.

For Foreign Policy Expert and Superforecaster Dr. Balkan Devlen, this question was close to home. Born in Izmir, Turkey, Dr. Devlen felt many in the media were getting carried away by the narrative of an underdog win. In this interview, he discusses the key drivers behind his forecast, factors that Western commentators failed to take sufficiently into account, and his assessment of what lies in store for Turkey—and the region—after the election.

GJ: The Superforecasters started at 71% for Erdogan. They are now at 98%. How has your forecast evolved over this period? What were the key drivers?

BD: I was consistently around 80–85% throughout the question period. In that sense, my forecast didn’t evolve much, primarily because the key drivers of my forecast didn’t really change.

Number one is the makeup of the electorate in Turkey, which consistently suggests that about 40% of the voters support the more conservative nationalist coalition led by Erdogan. And that didn’t change much over the period, despite difficulties when it comes to economics or other issues.

Number two is the fact that Erdogan has been in power over 20 years, and during that period managed to install his own cadres across the state and bureaucracy, as well as to gain control over almost all of the media. This created an information environment in which the opposition had a hard time reaching voters who were not already anti-Erdogan.

Lastly, Erdogan does not have the luxury of losing. The extreme polarization within Turkish society and Erdogan’s increasing authoritarianism, especially in the last 10–12 years, means that retiring after the election is just not an option for him, and his family and cronies have also been implicated throughout his rule in corruption, oppression of free speech and freedom of media, etc. Taking those three drivers together, I did not see the wherewithal that the opposition can come together and push out a win.

GJ: Several polls and many commentators in the West were predicting that the opposition would win. Did you find their arguments convincing?

BD: I did not find the polling or the Western commentary particularly convincing, primarily because they tend to base their arguments as if the elections were taking place without the context that I’ve just discussed. I also think that some were engaging in motivated reasoning and wishful thinking. One argument is that the impact of the earthquake in Turkey could have shifted the balance, but that again is a misunderstanding of the fundamental dynamics in the region.

GJ: What would have made you change your forecast?

BD: Perhaps two things. One, if there had been another candidate, either Istanbul Mayor Ekrem Imamoglu or Ankara Mayor Mansur Yavas, the opposition would have had a much better chance.

Two, if I were to see any high-level defections from the AKP or the close circle of Erdogan prior to the election, that would have suggested that, at least, there is a fracture within the ruling elite, that they were considering their post-election fates, and therefore they were breaking rank and jumping ship. We didn’t see any of that, so I didn’t see any reason to change my forecast.

GJ: Superforecasters as a group are now at 98% probability for Erdogan’s victory. Is there a chance they’re wrong?

BD: Of course, there is a chance that the Superforecasters could be wrong, and I am at 98% myself. There are always black swan events, despite the fact that there are only a few days left before the second round. But I don’t see any particular dynamic today that would suggest that the fundamentals prior to the first round were altered in any meaningful way.

Incidentally, I was expecting that this would go to the second round, partly because of the third-party candidate, Sinan Ogan, who now declared his support for Erdogan. That base will probably break 2-1, at least, for Erdogan. But the implication is that if Ogan weren’t in the running, Erdogan probably would have won in the first round.

GJ: What’s in store for Turkey now and for the region?

BD: Well, that’s a big question and I don’t think that’s enough space to go into a detailed examination. But I can tell you that the region and Turkey will probably need to accept the fact that Erdogan will be in power as long as he’s alive. And actors in Europe, in the Middle East, in the Caucasus, and elsewhere would need to adjust to that particular fact.

For Turkey, the results would not lead to a more democratic system. There are those in the parliament now, as part of Erdogan’s coalition, who are hardcore Islamists, and who are calling for much more Islamist policies, for example.

The election results will also lead to recriminations among the opposition coalition, as the only thing that really unites them is their opposition to Erdogan. Therefore, one can expect turmoil within the opposition parties in the post-election era.

Erdogan will probably consolidate his power. That might provide some stability in terms of regional policies now that the need to play for domestic audiences has decreased. Therefore, we are likely to see a more predictable foreign policy attitude from Turkey. I do not necessarily see, though, much change in the direction in which Erdogan wants to take the country, both domestically and internationally.

I don’t see Sweden’s membership being approved by the Turkish parliament before the end of summer as it’s just not a priority, although timing is very hard to predict in these cases.

But like I said, a proper answer to this question requires much more detail and a much longer exposition than can be provided in a short interview. One thing is quite clear, though, and that is Erdogan will be in power for the foreseeable future.