Forecasting Newsletter - Draft

This commit is contained in:
Nuno Sempere 2020-06-29 12:54:41 +02:00
parent e96b47f3b4
commit 98763db632

View File

@ -17,13 +17,13 @@
- Facebook releases a forecasting app ([link to the app](https://www.forecastapp.net/), [press release](https://npe.fb.com/2020/06/23/forecast-a-community-for-crowdsourced-predictions-and-collective-insights/), [TechCrunch take](https://techcrunch.com/2020/06/23/facebook-tests-forecast-an-app-for-making-predictions-about-world-events-like-covid-19/), [hot-takes](https://cointelegraph.com/news/crypto-prediction-markets-face-competition-from-facebook-forecasts)). The release comes before Augur v2 launches, and it is easy to speculate that it might end up being combined with Facebook's stablecoin, Libra.
- The Economist has a new electoral model out ([article](https://www.economist.com/united-states/2020/06/11/meet-our-us-2020-election-forecasting-model), [model](https://projects.economist.com/us-2020-forecast/president)) which gives Trump an 11% chance of winning reelection. Given that Andrew Gelman was involved, I'm hesitant to critizice it, but it seems a tad overconfident.
- The Economist has a new electoral model out ([article](https://www.economist.com/united-states/2020/06/11/meet-our-us-2020-election-forecasting-model), [model](https://projects.economist.com/us-2020-forecast/president)) which gives Trump an 11% chance of winning reelection. Given that Andrew Gelman was involved, I'm hesitant to criticize it, but it seems a tad overconfident.
- [COVID-19 vaccine before US election](https://www.aljazeera.com/ajimpact/wall-street-banking-covid-19-vaccine-election-200619204859320.html). Analysts see White House pushing through vaccine approval to bolster Trump's chances of reelection before voters head to polls. "All the datapoints we've collected make me think we're going to get a vaccine prior to the election," Jared Holz, a health-care strategist with Jefferies, said in a phone interview. The current administration is "incredibly incentivized to approve at least one of these vaccines before Nov. 3."
- ["Israeli Central Bank Forecasting Gets Real During Pandemic"](https://www.nytimes.com/reuters/2020/06/23/world/middleeast/23reuters-health-coronavirus-israel-cenbank.html). Israeli Central Bank is using data to which it has real time access, like credit-card spending, instead of lagging indicators.
- ["Israeli Central Bank Forecasting Gets Real During Pandemic"](https://www.nytimes.com/reuters/2020/06/23/world/middleeast/23reuters-health-coronavirus-israel-cenbank.html). Israeli Central Bank is using data to which it has real-time access, like credit-card spending, instead of lagging indicators.
- [Google](https://www.forbes.com/sites/jeffmcmahon/2020/05/31/thanks-to-renewables-and-machine-learning-google-now-forecasts-the-wind/) produces wind schedules for windfarms. "The result has been a 20 percent increase in revenue for wind farms". See [here](https://www.pv-magazine-australia.com/2020/06/01/solar-forecasting-evolves/) for essentially the same thing on solar forecasting.
- [Google](https://www.forbes.com/sites/jeffmcmahon/2020/05/31/thanks-to-renewables-and-machine-learning-google-now-forecasts-the-wind/) produces wind schedules for wind farms. "The result has been a 20 percent increase in revenue for wind farms". See [here](https://www.pv-magazine-australia.com/2020/06/01/solar-forecasting-evolves/) for essentially the same thing on solar forecasting.
- Survey of macroeconomic researchers predicts economic recovery will take years, reports [538](https://fivethirtyeight.com/features/dont-expect-a-quick-recovery-our-survey-of-economists-says-it-will-likely-take-years/).
@ -37,14 +37,14 @@ Ordered in subjective order of importance:
- Metaculus
- posted [A Preliminary Look at Metaculus and Expert Forecasts](https://www.metaculus.com/news/2020/06/02/LRT/): Metaculus forecasters do better, and the piece is a nice reference point.
- was featured in [Forbes](https://www.forbes.com/sites/erikbirkeneder/2020/06/01/do-crowdsourced-predictions-show-the-wisdom-of-humans/#743b7e106d9d).
- anounced their [Metaculus Summer Academy](https://www.metaculus.com/questions/4566/announcing-a-metaculus-academy-summer-series-for-new-forecasters/): "an introduction to forecasting for those who are relatively new to the activity and are looking for a fresh intellectual pursuit this summer"
- announced their [Metaculus Summer Academy](https://www.metaculus.com/questions/4566/announcing-a-metaculus-academy-summer-series-for-new-forecasters/): "an introduction to forecasting for those who are relatively new to the activity and are looking for a fresh intellectual pursuit this summer"
- [Replication Markets](https://predict.replicationmarkets.com/) might add a new round with social and behavioral science claims related to COVID-19, and a preprint market, which would ask participants to forecast items like publication or citation. Replication Markets is also asking for more participants, with the catchline "If they are knowledgeable and opinionated, Replication Markets is the place to be to make your opinions really count."
- Good Judgement family
- [Good Judgement Open](https://www.gjopen.com/): Superforecasters were able to detect that Russia and the USA would in fact undertake some (albeit limited) form of negotiation, and do so much earlier than the general public, even while posting their reasons in full view. One thread to follow is [this one](https://www.gjopen.com/comments/1039968).
- Good Judgement Analytics continues to provide their [covid dashboard](https://goodjudgment.com/covidrecovery/).
- Good Judgement Analytics continues to provide its [COVID-19 dashboard](https://goodjudgment.com/covidrecovery/).
- [PredictIt](https://www.predictit.org/) & [Election Betting Odds](http://electionbettingodds.com/). I stumbled upon an old 538 piece on fake polls: [Fake Polls are a Real Problem](https://fivethirtyeight.com/features/fake-polls-are-a-real-problem/). Some polls may have been conducted by PredictIt traders in order to mislead or troll other PredictIt traders; all in all, an amusing example of how prediction markets could encourage worse information.
@ -62,16 +62,16 @@ Ordered in subjective order of importance:
- World powers to converge on strategies for presenting COVID-19 information to make forecasters' jobs more interesting:
- [Brazil stops releasing Covid-19 death toll and wipes data from official site](https://www.theguardian.com/world/2020/jun/07/brazil-stops-releasing-covid-19-death-toll-and-wipes-data-from-official-site).
- [Brazil stops releasing COVID-19 death toll and wipes data from official site](https://www.theguardian.com/world/2020/jun/07/brazil-stops-releasing-covid-19-death-toll-and-wipes-data-from-official-site).
- Meanwhile in Russia, [St Petersburg issues 1,552 more death certificates in May than last year, but Covid-19 toll was 171](https://www.theguardian.com/world/2020/jun/04/st-petersburg-death-tally-casts-doubt-on-russian-coronavirus-figures).
- Meanwhile, in Russia, [St Petersburg issues 1,552 more death certificates in May than last year, but Covid-19 toll was 171](https://www.theguardian.com/world/2020/jun/04/st-petersburg-death-tally-casts-doubt-on-russian-coronavirus-figures).
- In the US, [CDC wants states to count probable coronavirus cases and deaths, but most arent doing it](https://www.washingtonpost.com/investigations/cdc-wants-states-to-count-probable-coronavirus-cases-and-deaths-but-most-arent-doing-it/2020/06/07/4aac9a58-9d0a-11ea-b60c-3be060a4f8e1_story.html)
- [India has the fourth-highest number of COVID-19 cases, but the Government denies community transmission](https://www.abc.net.au/news/2020-06-21/india-coronavirus-fourth-highest-covid19-community-transmission/12365738)
- One suspects that this denial is political, because India is otherwise [being](https://www.maritime-executive.com/editorials/advanced-cyclone-forecasting-is-saving-thousands-of-lives) [extremely](https://economictimes.indiatimes.com/news/politics-and-nation/world-meteorological-organization-appreciates-indias-highly-accurate-cyclone-forecasting-system/articleshow/76280763.cms) [competent](https://economictimes.indiatimes.com/news/politics-and-nation/mumbai-to-get-hyperlocal-rain-outlooks-flood-forecasting-launched/articleshow/76343558.cms) in weather forecasting.
- Youyang Gu's model, widely aclaimmed as one of the best coronavirus models for the US, produces 95% confidence intervals which [seem too narrow](https://twitter.com/LinchZhang/status/1270443040860106753) when extended to [Pakistan](https://covid19-projections.com/pakistan).
- Youyang Gu's model, widely acclaimed as one of the best coronavirus models for the US, produces 95% confidence intervals which [seem too narrow](https://twitter.com/LinchZhang/status/1270443040860106753) when extended to [Pakistan](https://covid19-projections.com/pakistan).
## Hard to categorize.
@ -84,7 +84,7 @@ Ordered in subjective order of importance:
- [How to improve space weather forecasting](https://eos.org/research-spotlights/how-to-improve-space-weather-forecasting) (see [here](https://agupubs.onlinelibrary.wiley.com/doi/full/10.1029/2018SW002108#) for the original paper):
> For instance, the National Oceanic and Atmospheric Administrations Deep Space Climate Observatory (DSCOVR) satellite sits at the location in space called L1, where the gravitational pulls of Earth and the Sun cancel out. At this point, which is roughly 1.5 million kilometers from Earth, or barely 1% of the way to the Sun, detectors can provide warnings with only short lead times: about 30 minutes before a storm hits Earth in most cases or as little as 17 minutes in advance of extremely fast solar storms.
- [Coup cast](https://oefresearch.org/activities/coup-cast): A site which estimates the yearly probability of a coup. The color coding is misleading; click on the countries instead.
- [Coup cast](https://oefresearch.org/activities/coup-cast): A site that estimates the yearly probability of a coup. The color coding is misleading; click on the countries instead.
- [Prediction = Compression](https://www.lesswrong.com/posts/hAvGi9YAPZAnnjZNY/prediction-compression-transcript-1). "Whenever you have a prediction algorithm, you can also get a correspondingly good compression algorithm for data you already have, and vice versa."
- Other LessWrong posts which caught my attention were [Betting with Mandatory Post-Mortem](https://www.lesswrong.com/posts/AM5JiWfmbAytmBq82/betting-with-mandatory-post-mortem) and [Radical Probabilism](https://www.lesswrong.com/posts/ZM63n353vh2ag7z4p/radical-probabilism-transcript)
@ -119,7 +119,7 @@ Ordered in subjective order of importance:
> On the other hand, a well-calibrated mechanistic model that accounts for all relevant dynamic factors and external influences could, in principle, have been used to predict the behaviour of the epidemic reliably and precisely. Yet, lack of detailed data on transmission routes and risk factors precluded the parameterisation of such a model and are likely to do so again in future epidemics in resource-poor settings.
- In the selection of quotes above, we gave an example of a forecast which ended up overestimating the incidence, yet might have "served as a call to arms". It's maybe a real life example of a forecast changing the true result, leading to a fixed point problem, like the ones hypothesized in the parable of the [Predict-O-Matic](https://www.lesswrong.com/posts/SwcyMEgLyd4C3Dern/the-parable-of-predict-o-matic).
- In the selection of quotes above, we gave an example of a forecast which ended up overestimating the incidence, yet might have "served as a call to arms". It's maybe a real-life example of a forecast changing the true result, leading to a fixed point problem, like the ones hypothesized in the parable of the [Predict-O-Matic](https://www.lesswrong.com/posts/SwcyMEgLyd4C3Dern/the-parable-of-predict-o-matic).
- It would be a fixed point problem if \[forecast above the alarm threshold\] → epidemic being contained, but \[forecast below the alarm thresold\] → epidemic not being contained.
- Maybe the fix-point solution, i.e., the most self-fulfilling (and thus, accurate) forecast, would have been a forecast on the edge of the alarm threshold, which would have ended up leading to mediocre containment.
- The [troll polls](https://fivethirtyeight.com/features/fake-polls-are-a-real-problem/) created by PredictIt traders are perhaps a more clear cut example of Predict-O-Matic problems.
@ -128,11 +128,11 @@ Ordered in subjective order of importance:
- I also found the following derivation of the logarithmic scoring rule interesting. Consider: If you assign a probability to n events, then the combined probability of these events is p1 x p2 x p3 x ... pn. Taking logarithms, this is log(p1 x p2 x p3 x ... x pn) = Σ log(pn), i.e., the logarithmic scoring rule.
- [Binary Scoring Rules that Incentivize Precision](https://arxiv.org/abs/2002.10669). The results (the closed-form of scoring rules which minimize the a given forecasting error) are interesting, but the journey to get there is kind of a drag, and ultimately the logarithmic scoring rule ends up being pretty decent according to their measure of error.
- [Binary Scoring Rules that Incentivize Precision](https://arxiv.org/abs/2002.10669). The results (the closed-form of scoring rules which minimize a given forecasting error) are interesting, but the journey to get there is kind of a drag, and ultimately the logarithmic scoring rule ends up being pretty decent according to their measure of error.
- Opinion: I'm not sure whether their results are going to be useful for things I'm interested in (like human forecasting tournaments, rather than kaggle data analysis competitions). In practice, what I might do if I wanted to incentivize precision is to ask myself if this is a question where the answer is going to be closer to 50%, or closer to either of 0% or 100%, and then use either the Brier or the logarithmic scoring rules. That is, I don't want to minimize an l-norm of the error over [0,1], I want to minimize an l-norm over the region I think the answer is going to be in, and the paper falls short of addressing that.
- Opinion: I'm not sure whether their results are going to be useful for things I'm interested in (like human forecasting tournaments, rather than Kaggle data analysis competitions). In practice, what I might do if I wanted to incentivize precision is to ask myself if this is a question where the answer is going to be closer to 50%, or closer to either of 0% or 100%, and then use either the Brier or the logarithmic scoring rules. That is, I don't want to minimize an l-norm of the error over [0,1], I want to minimize an l-norm over the region I think the answer is going to be in, and the paper falls short of addressing that.
- [A list of prediction markets](https://docs.google.com/spreadsheets/d/1XB1GHfizNtVYTOAD_uOyBLEyl_EV7hVtDYDXLQwgT7k/edit#gid=0), and their fates, mantained by Jacob Laguerros. Like most startups, most prediction markets fail.
- [A list of prediction markets](https://docs.google.com/spreadsheets/d/1XB1GHfizNtVYTOAD_uOyBLEyl_EV7hVtDYDXLQwgT7k/edit#gid=0), and their fates, maintained by Jacob Laguerros. Like most startups, most prediction markets fail.
Note to the future: All links are added automatically to the Internet Archive. In case of link rot, go [here](https://archive.org/)