Socratic Form Microscopy

Grading Zach's 2019 Predictions

by Tessa Alexanian in Falsifiable

Let’s look at how accurately Zach predicted 2019. As usual (2017 results, 2018 results), the predictions are broken down by region.

Canada

  1. Liberals win the next Canadian election – 60%
  2. Majority government in the next Canadian election – 60% 
  3. NDP loses seats in the next Canadian election – 80%
  4. Conservatives gain seats in the next Canadian election – 70%
  5. Jagmeet Singh is not in parliament at the end of 2019 – 51%
  6. No court finds the assisted dying bill isn’t in line with Carter v Canada in 2019 – 80% 
  7. No court rules on carbon tax constitutionality in 2019 – 90%
  8. No terrorist attack in Canada that kills > 10 Canadians in 2019 – 90%
  9. Opioid poisoning deaths decline from 2018 levels in Canada – 80%
  10. Construction on the Trans Mountain pipeline does not begin again in 2019 – 90%
  11. Canadian unemployment rate at or below 6% at the end of 2019 – 60%
  12. Health Canada still doesn’t allow Soylent to enter the country by the end of 2019 – 70%

America

  1. Less than 100km of concrete wall (or steel slat wall) on the border with Mexico will be constructed – 80%
  2. No registry of Muslims created – 90%
  3. No department of the Federal Government is eliminated – 90%
  4. Congress and Trump do not sign into law an extension of DACA – 60%
  5. Mueller’s investigation finishes in 2019 – 70%
  6. Impeachment proceedings aimed at Trump are not started in 2019 – 60%
  7. More than a dozen credible (having held past elected office representing at least 100,000 people) candidates declared for the Democratic primary by the end of 2019 – 80%
  8. Less than two dozen credible candidates declared for the democratic primary by the end of 2019 – 70%
  9. Trump is still president at the end of 2019 – 90%
  10. No terrorist attack in America that kills > 10 Americans – 70%
  11. No terrorist attack in America that kills > 100 Americans – 90%
  12. Opioid poisoning deaths decline from 2018 levels in America – 70%
  13. NGDP growth below 5% for 2019 – 60%
  14. US unemployment rate at or below 4% at the end of 2019 – 60%
  15. Some form of North American free trade deal still in effect at the end of 2019 – 80%
  16. At least two metros beyond Phoenix see open, commercial self-driving car services by the end of 2019 – 51%
  17. More than one month of cumulative government shut downs in 2019 – 51%

South America

  1. FARC peace deal remains in place on January 1, 2020 – 60%
  2. Inflation in Venezuela is above 10,000% for the year of 2019 (as measured by DolarToday) – 80%
  3. Venezuelan GDP continues to contract in 2019 – 90%
  4. United Socialist party retains control of the Venezuelan presidency in 2019 – 80%
  5. Protests (and the official response to those protests) result in more than 100 fatalities in Venezuela in 2019 – 51%
  6. Major Venezuelan opposition groups do not enter any sort of power sharing agreement with the Venezuelan regime in 2019 – 90%
  7. No successful coup in Venezuela in 2019 – 90%
  8. Venezuelan oil production below 750,000 bpd by the end of 2019 – 60%
  9. Brazilian Real down against the US dollar at the end of 2019 – 51%
  10. Argentina meets its 17% inflation target by ±2% – 70%
  11. Argentina meets its IMF target of primary budget balance by the start of 2020 – 51% 

Middle East

  1. No Israeli politician is indicted by the ICC over settlement activity in 2019 – 90%
  2. Netanyahu is no longer prime minister of Israel at the end of 2019 – 51%
  3. No Palestinian led Intifada in Israel that results in the deaths of >1000 combined attackers, security forces, and civilians (this is a conflict characterized by suicide bombing and police responses) – 90%
  4. No Israeli led operation in the West Bank or Gaza that results in the deaths of >1000 combined soldiers, civilians, and militants (this is a conflict characterized by rocket fire and military strikes) – 80%
  5. Fatah and Hamas do not meaningfully reconcile in 2019 (e.g. Fatah still doesn’t control Gaza by January 1, 2020) – 70%
  6. No significant resurgence in ISIL in 2019 (e.g. it does not gain territory over the next year) – 90%
  7. Fewer casualties in the Syrian Civil War in 2019 than in 2018 – 70%
  8. No power sharing agreement or durable ceasefire (typified by the three months following the agreement each having less than 500 fatalities) in Syria in 2019 – 80%
  9. Bashar Al Assad is still President of Syria on January 1, 2020 – 90%
  10. Protests in Iran do not result in more than 100 fatalities by the end of 2019 – 70%
  11. Hassan Rouhani is still President of Iran on January 1, 2020 – 90%
  12. No new international sanctions against Iran (does not include adding new organizations or individuals to old categories and requires coordinated participation of at least two countries) – 80%
  13. America does not rejoin the JCPA – 90%
  14. No attack on the Iranian nuclear program by Israel – 90%
  15. Iran does not withdraw from the deal limiting its nuclear program – 80%
  16. Conditional on Iran remaining in the nuclear deal, inspectors find no evidence of violations after the deal began – 90%
  17. Yemen Civil War continues – 90%
  18. The Hodeidah ceasefire in Yemen is maintained for 2019 – 51%
  19. America withdraws support for Saudi Arabia in the Yemeni civil war – 51%
  20. Mohammed bin Salman either remains as crown prince of Saudi Arabia, or becomes king (i.e. no coup or succession shake-up) – 80%
  21. No resolution or lifting of embargo in the Qatar crisis – 80%
  22. OPEC production cuts continue through to the end of 2019 – 70%

Africa

  1. South Sudan peace deal holds – 60%
  2. Libya still has two rival governments on January 1, 2020 – 60%
  3. No protests, riots, or rebellion in Egypt that kills >100 people in a one week period – 80%
  4. No protests, riots, or rebellion in Tunisian kills >50 people in a one week period – 90%
  5. No terrorist attack in Tunisia kills >20 people – 80%
  6. ANC wins South African elections – 80%
  7. Botswana’s Transparency International rating closest to January 1, 2020 shows improvement under president Masisi – 60%

Asia

  1. Inflation rate in Japan remains below 2% in 2019 – 90%
  2. Japanese constitutional reform (removing pacifism) does not occur in 2019 – 80%
  3. At least two more countries ratify the CPTPP in 2019 – 51%
  4. All signatories have not ratified the CPTPP by the end of 2019 – 90%
  5. North Korean détente with Trump last through 2019 (typified by no new sanctions or missile tests) – 51%
  6. Xi Jinping visits North Korea in 2019 – 80%
  7. US-China trade war ends – 51%
  8. Modi wins elections in India – 70%
  9. No return to democracy in Thailand in 2019 – 90%
  10. Legal deadlock on same-sex marriage in Taiwan not resolved this year – 51%

Europe

  1. No resolution to the crisis in Ukraine – 80%
  2. Ukrainian elections are held – 70%
  3. Poroshenko no longer president of Ukraine by end of 2019 – 60%
  4. Russian GDP growth is less than 3% – 80%
  5. No gain of greater than 20% in the value of the ruble vs. the dollar – 90%
  6. Fall of more than 10% in the value of the ruble vs. the dollar – 51%
  7. Sanctions against Russia are not significantly rolled back (e.g. sanctions remain in place against Rosneft, Novate, Gazprombank and Vnesheconombank by all members of the G7 remain in place at the end of 2019) – 90%
  8. Teresa May remains prime minister of the United Kingdom – 70%
  9. Brexit does not occur on or by March 29, 2019 – 60%
  10. No snap election/vote of no-confidence in the UK in 2019 – 60%
  11. No resolution to the Julian Assange situation – 60%
  12. Poland’s EU voting rights aren’t suspended – 70%
  13. Euroskeptics make gains in 2019 EU elections – 80%
  14. The average of the last 5 opinion polls of Macron collected on Wikipedia have his net favourability below -30% – 80%
  15. Italy keeps its budget within the agreed upon EU boundaries (triggering no fines or other disciplinary action) – 80%
  16. DeepMind releases an algorithm that can beat top humans in StarCraft II – 51%

Notes

  • Zach was confidently wrong that Canadian courts wouldn’t rule on assisted dying (Canada #6) and carbon taxes (Canada #7). This may have been an over-correction from their incorrect 2018 prediction that the courts would rule on assisted dying.
  • America #7 is false with the provided definition of credible; most news outlets identified 13 major democratic candidates at the start of 2020, but Andrew Yang and Tom Steyer have never held elected office.
  • The Lyft/Aptive pilot in Las Vegas, NV clearly counts as an “open, commercial self-driving car service” (America #16) but is a free-to-ride service funded by a subsidized city contract (the  Drive.ai pilot in Arlington, TX) “commercial”? Is a service available only to pre-screened, though regular, users (the Uber pilot in Pittsburgh, PA) “open”? I decided the the TX and PA services could round to a metro area.
  • I went back and forth on whether protests resulted in more than 100 fatalities in Venezuela (South America #5). Wikipedia says 107+, but its sources for that number are all from before the OHCHR report of 66 deaths from January to May. In July, Human Rights Watch reported thousands of extrajudicial killings, but those don’t seem linked directly to protests, and I don’t know if they count as “official”. I decided that there were probably more than 100 fatalities.
  • The South Sudan peace deal (Africa #1) hasn’t been explicitly abandoned, but it has been violated. I decided it still “holds”, in that both sides are still trying to keep the peace.
  • Botswana’s Transparency International rating (Africa #7) didn’t get worse under Masisi, it’s just still 61.
  • I think Thailand’s elections don’t qualify as a return to democracy (Asia #9) given that the military was able to appoint the senate and the senate voted in the leader of the 2014 coup as prime minister.
  • Although Julian Assange is no longer hiding in the Ecuadorian embassy, there are ongoing extradition hearings, so I wouldn’t call the situation (Europe #11) resolved.
  • I decided the intent of “beat top humans” in Europe #16 is fulfilled by AlphaStar playing at the Grandmaster level with all three StarCraft II races, though it being better than 99.8% of Battle.net players still means it has no high-profile defeats of individual top humans.

Calibration

The whole point of having predictions with binned probabilities (here, 51%, 60%, 70%, 80%, 90%) is that you can then check your accuracy; an individual prediction can be right or wrong, but when you pool several predictions you were equally sure of, you can tell if you were right or wrong to be that sure. Here’s how Zach did:

  • Of predictions at a 51% confidence level, I marked 5 right and 10 wrong (33%).
  • Of predictions at a 60% confidence level, I marked 11 right and 5 wrong (69%).
  • Of predictions at a 70% confidence level, I marked 11 right and 4 wrong (73%).
  • Of predictions at an 80% confidence level, I marked 21 right and 3 wrong (87%).
  • Of predictions at a 90% confidence level, I marked 23 and 2 wrong (92%).

Represented graphically (along with previous years) this looks like:

Line graph comparing the percentage of correct predictions to perfect calibration. An animation contrasts 2019 scores with 2017 and 2018 scores.

The worsened calibration for 2019 is likely a reflection of the fact that Zach didn’t grade these predictions. My best interpretation of the predictions (and of the available evidence) is almost certainly less accurate than Zach’s would have been.

A Final Note

You may wonder why I graded these predictions. I am sorry to say that Zach could not; they passed away suddenly last August. (You can read the obituary here, if you’d like.) I wish, with all the sharp vehemence of recent grief, that I could still eavesdrop as Zach honed their beliefs about the world. I hope my years of doing so will inspire me to cultivate truer (and better-calibrated) beliefs on my own.


Tags: predictions