2015 Predictions – Outcomes

Below are the probabilities I assigned to different events happening in 2015, along with the outcomes. Summary section at the end.

Politics/World:
1. On Dec 31, 2015, the price of regular gas at my local gas station will be > $3.10: 50% NO
2. On Dec 31, 2015, the price of regular gas at my local gas station will be < $3.50: 85% YES
3. North Korea’s government will not collapse (whether due to revolt/civil war, or war with another country): 90% YES
4. There will not be any terrorist attacks in the US that kill more than 40 people: 85% YES
5. We will end the year with an unemployment rate > 6% in the US: 75% NO
6. At the end of the year AMF will still be a GiveWell recommended charity: 75% YES
7. At the end of the year GiveDirectly will still be a GiveWell recommended charity: 85% YES

Blog:
8. I will post at least 5 blog posts: 95% YES
9. I will post at least 12 blog posts: 75% NO
10. I will post more than 25 blog posts: 10% NO
11. Sam will post at least 5 blog posts: 90% YES
12. Sam will post at least 12 blog posts: 70% NO
13. Sam will post more than 25 blog posts: 15% NO

Personal:
14. I will still be a vegetarian or pescetarian at the end of the year: 90% YES
15. I will be a vegan at the end of the year: 5% NO
16. Sam will be a vegetarian or a pescetarian at the end of the year: 5% NO
17. Sam will be a vegan at the end of the year: 1% NO
18. We will end the year still living in the Cambridge area: 90% YES
19. We will end the year living in a Rationalist house: 10% NO
20. I will still work for my current company: 85% YES
21. Sam will still be involved with the morality lab: 75% YES
22. I will read at least 15 books: 90% YES
23. We will sell the car: 45% YES
24. I will attend Burning Man: 95% YES
25. Sam will attend Burning Man: 90% YES
26. My computer will require repairs or replacement: 15% NO
27. Sam’s computer will require repairs or replacement: 20% YES
28. I will lose my wallet: 10% NO
29. I will run at least 1 marathon: 20% NO
30. Wumpus will still be alive: 90% YES
31. We will have a baby: 1% NO

Summary

Below is a comparison of confidence against accuracy. (30% chance of X is same as 70% chance of not X, so probabilities of less than 50% swapped around for grouping purposes.) If the events you had 80% confidence in happen 80% of the time, you’re perfectly calibrated. If they instead happen 90% of the time, you’re underconfident. And if they only happen 70% of the time, you’re overconfident.

PredictionAccuracyChart

Obviously there are some sample size issues here. I made a lot of additional predictions (more than 600!) throughout the year at predictionbook.com though. Here are the summary statistics from those:

predictionbookchart

From my PredictionBook statistics it looks like I’ve become more accurate over the year. I’ve learned to use extreme estimates (>96% and <4%) more sparingly.

Advertisements
This entry was posted in Eloise and tagged , . Bookmark the permalink.

2 Responses to 2015 Predictions – Outcomes

  1. Nice! You are much better calibrated than I am :). How quickly do you generally get feedback on predictionbook predictions?

    Like

    • sanerthanlasagna says:

      The time scales of my predictions vary a lot, from predictions whose outcomes I’ll know in a few minutes to ones that won’t be resolved for several years. Quick feedback time is probably useful for improving quickly, but I think it’s also valuable to practice predicting things that are far away—far future predictions often feel harder to me because so much can change in the intervening time.

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s