efosong

About:
Show more

-0.072812

Relative Brier Score

0

Forecasts

0

Upvotes
Forecasting Activity
Forecasting Calendar
 

Past Week Past Month Past Year This Season All Time
Forecasts 0 8 126 26 377
Comments 0 1 19 6 93
Questions Forecasted 0 4 25 10 57
Upvotes on Comments By This User 0 5 54 16 102
 Definitions
New Badge
efosong
earned a new badge:

Active Forecaster

New Prediction
Confirmed previous forecast
Files
New Prediction
efosong
made their 9th forecast (view all):
Probability
Answer
98% (+16%)
Less than $90
2% (-15%)
More than or equal to $90 but less than $100
0% (-1%)
More than or equal to $100 but less than $110
0% (0%)
More than or equal to $110 but less than $120
0% (0%)
More than or equal to $120

Model update. With the time remaining, exceeding $90 tomorrow is very unlikely.

Files
New Prediction
efosong
made their 8th forecast (view all):
Probability
Answer
82% (+3%)
Less than $90
17% (-2%)
More than or equal to $90 but less than $100
1% (-1%)
More than or equal to $100 but less than $110
0% (0%)
More than or equal to $110 but less than $120
0% (0%)
More than or equal to $120

Updating based on latest price

Files
New Badge
efosong
earned a new badge:

Upvotes Received

New Prediction
efosong
made their 7th forecast (view all):
Probability
Answer
79% (+6%)
Less than $90
19% (-6%)
More than or equal to $90 but less than $100
2% (0%)
More than or equal to $100 but less than $110
0% (0%)
More than or equal to $110 but less than $120
0% (0%)
More than or equal to $120
Update latest data.
Files
New Prediction
efosong
made their 6th forecast (view all):
Probability
Answer
73% (+19%)
Less than $90
25% (-19%)
More than or equal to $90 but less than $100
2% (0%)
More than or equal to $100 but less than $110
0% (0%)
More than or equal to $110 but less than $120
0% (0%)
More than or equal to $120

Model-based update in light of new data.

Files
New Prediction
efosong
made their 1st forecast (view all):
Probability
Answer
Forecast Window
10%
Yes
Mar 19, 2024 to Mar 19, 2025
90%
No
Mar 19, 2024 to Mar 19, 2025

IMO the crowd is too low.

  • Base rate of death within the next year for an 85 year old man (in the UK) is 10%.
  • There have been rumours about his health before (but he is almost 85...) so I wouldn't update too much based on that.
  • He surely has better healthcare than the average UK 85 year old, but I'd think even the best healthcare wouldn't more than half the probability. He probably also has slightly less chance of dying each year for genetic reasons.
  • There's a small chance (maybe 2%?) that he does not die but becomes incapacitated enough to force resignation.
  • I think risk of exile or other ousting is around the 1-2% mark, so in total I'd go for 10%.

Checking crowd activity after writing the above rationale, seems like @404_NOT_FOUND has come to a pretty similar conclusion with similar estimates for similar possibilities, but with a more detailed rationale which I broadly agree with. You might want to read that rationale for more.

Files
jrl
made a comment:
One other scenario is that Iran is in the process of an organized transition to the next generation of leadership. I have no insight into how long that will take, but Khamenei has stacked the body that does that deliberation with his own people, and appears to want one of his sons to be the next leader. i.e. there's at least some chance of a peaceful and intentional transition of leadership in the next year.
Files
New Prediction
efosong
made their 5th forecast (view all):
Probability
Answer
54% (-30%)
Less than $90
44% (+30%)
More than or equal to $90 but less than $100
2% (0%)
More than or equal to $100 but less than $110
0% (0%)
More than or equal to $110 but less than $120
0% (0%)
More than or equal to $120

Recent large movements in the market. Somewhat unexpected, but it can happen. Sticking with my model, and mostly ignoring the intraday clarification (side forecast: 100% chance I will be mildly annoyed if intraday causes the question to resolve differently to closing price). Small adjustment for 'momentum' (though I think the random walk prior is a good starting point).

Files
New Prediction

Coming back towards crowd given time remaining. We've seen slightly-relevant news (YouTube requiring AI disclosure) but it's not really the same as C2PA. The BBC Verify stuff from earlier this month is relevant, but the "only" in the question remains a very high bar.

Files
Files
Tip: Mention someone by typing @username