Forecasting the Metaculus Conditional Cup
Read my rationale about various questions, mostly geopolitics.
In this newsletter, I often write how important conditional prediction markets are for decision making. Still, I couldn’t bring myself to take part in Metaculus’ Conditional Cup so far. Writing about it might help, so here we go with some stream-of-consciousness forecasting some of the cup questions.
I’ll hide the community prediction at first and might adapt in a second step.
If Khamenei out as Leader of Iran in 2024, US Iran War Before 2025?
Iran does some proxy-fighting against the US, but so far the country itself pretends to stay friendly. I don’t see any obvious reason why Khamenei should be out (peeking - 13% - confirmed). Neither do I see any reason Iran wants to escalate to a war and the US doesn’t seem keen on any shenigans either. On the other hand, the situation there is complex, so I’d but the general risk of a war at “probably not” 20%. Metaculus says 14%. Ok, I’ll update to that as they probably know better.
What about the condition? If another leader emerges would they be more or less likely? No clue really. To differentiate a little, I tend towards a more aggressive successor, so let’s say +1% war chance if YES.
Reveal! 5% if yes and 8% if no. Metaculus you are inconsistent. They cannot both be significantly below the unconditional 14%. I’ll keep my numbers.
If Gang Leader Fito in Custody by April 2024, extended State of Emergency in Ecuador 2024?
No clue who this Fito is or what is going on in Ecuador. The description explains:
On January 8, 2024, Ecuador's president Daniel Noboa declared a 60-day state of emergency. The state of emergency followed a wave of violence related to the drug trade in Ecuador and the escape of gang leader José Adolfo Macías Villamar (also known as Fito) from prison.
The link between the escape and the violence is unclear. The violence was increasing for years, so the escape feels more like a symptom than a cause to me. It also means catching Fito or not will not make much of a difference. In more recent news, things are happening but it doesn’t go towards restoring peace. I’m leaning slightly towards an extension in both cases: 60%.
Reveal! Community prediction is If yes, 50%. If no, 60%. Apparently, catching the gang leader would be considered progress. Maybe I have seen too many movies where gang leaders continue to lead even if imprisoned, but it doesn’t really change my estimate. Ok, I’ll lower it to 59%, so there is a difference in predictions.
If Biden and Trump Debate in 2024, will Trump win?
Metaculus currently has a Trump win at 50% unconditionally and the chance for a debate at 65%. The question is open for a while, but recently Trump actually asked for a debate and Biden rejected. That 65% is surprisingly low to me. I’m at 75%. Isn’t it a tradition that the two candidates debate after the nomination and before the election? I think the current win chances assume a debate as usual, so the NO case remains. What could happen?
Biden drops out. Then a Trump win is more likely just because of the ensuing chaos among the Democrats.
Trump health issues make him drop out. Then a Trump win is impossible.
Trump prevented from running somehow. Then a Trump win is impossible.
Trump in jail but still running. A Trump is less likely in this scenario any ways.
Trump is significantly less likely to win if no debate happens: 15% imho. That does not imply that Biden should reject a debate. Instead the causation is the other way around: If Trump cannot win anymore, the debate will not happen. (Recently published: Conditional prediction markets are evidential, not causal)
Reveal! Apparently, the community is 50% if yes and 48% if no. I stay at my 15% and leave a comment accordingly.
If Fed Funds Rate below 4%, S&P goes up in 2024?
Apparently, I already predicted this one previously. I’m essentially the same as the community (YES improves the up chances) but slightly less optimisting in both cases.
If US Federal Min Wage Increase before 2025, Democrat Wins 2024 US Presidential Election?
I assume this question gauges if Biden should increase the min wage to improve his election chances. A Democrat win is at 50% and I agree with that.
A quick search gives me no news about any intentions to increase the minimum wage and it has been stagnant since 2009. Metaculus is at 40% that it will still happen this year but there are no recent comments why. I’m at 10% there, so the if-NO case is at 50% like the general chance.
So what about the unlikely YES case, where the Biden administation increases min wage in the next months? Not sure how the parties would spin that. For the heck of it, I’ll give it a +1, so a 51%.
Reveal! Community is at 50% both.
If LLM cyberattack/virus/worm, NVIDIA stock below $250 this year?
The LLM attack chance are at 68% and have been on an upwards trend for the last months. This feels too high to me. A virus or worm which relies on API calls to OpenAI seems like a bad idea to me. Big question is if “cyberattack” includes the social engineering parts or only technical attack vectors?
Nvidia is at $721 currently but it was below $250 less than a year ago. I see no reason for them to drop that sharply even nothing happens, so 5% if no. Maybe they drop a little if a big attack becomes public but I doubt there is much of an effect, so 6% if yes.
Reveal! 16% if yes and 8% otherwise says the community. Well, I still feel with a slightly more extreme position.
If Musk Chairman Of Twitter in May 2024, Twitter Public Company again?
Metaculus has the unconditional chairman question at 99% and the unconditional Twitter public at 1%. That holds for the YES case.
What remains is the very unlikely chance that Musk drops out until end of May. I’m still at 1% because that is too little time to take it public anyways.
Reveal! Hm, if no the community says 8%. Well, I disagree.
If UN Troops In Gaza end of May, Joe Biden wins the election?
The unconditional Gaza situation is at 4%. That means “if no” is equivalent with Biden’s unconditional chances of 50%.
So now we must consider the unlikely scenario that UN peacekeeping troops are in Gaza before June. That would be good news and reflect somewhat well on the Biden administration. They will probably claim a significant part of the achievement to restore peace. I’m giving Biden +5% chances here.
Reveal! 51% if yes and 48% if no. It seems the Metaculus community has higher expectations and rather discounts Biden if it doesn’t happen. The difference is small, so I keep my predictions.
If you disagree about my predictions, I would love to read some arguments.
Many people write how Trump will surely or surely not win the election. Robert de Neufville is someone with great forecasting credentials doing this: Haley Republicans Could Sink Trump.
Someone disappointed by health science writes about 10 technologies which he does not expect in the next 5 years. Here is a prediction market for them. One is at 79%, so there is hope.
Probably until next week, unconditional readers! 😊