If a market submitted here resolves in the same direction as /duck/will-donald-trump-win-the-2024-pres , either both Yes or both No, that submission will resolve Yes here.
If a submitted market resolves in the opposite direction as the Trump Market, it will resolve No. If one of the markets resolves N/A or %, I may N/A that option here unless I think it was a mis-resolution.
To make effective bets here, my advice is that you should first figure out the chances that the two markets resolve the same way if the probabilities are accurate and they are independent events. For a market very likely to resolve yes, this would be close to Trump's current % odds. For a market very likely to resolve No, this would be close to (1 - Trump's current % odds). For anything else, it'll be somewhere in the middle. Call this P(I).
Then, if you think the event is correlated with Trump's election, you should bet higher than P(I). If you think it's anti-correlated, you should bet lower than P(I). If you think it's not very correlated or anti-correlated, you should try to bet against the people betting it higher or lower than P(I).
You can submit your own markets in the format of [Market Title] - [Link to market]. If the market's end date isn't obvious, please also include that information in the title like for the market about a recession.
Guidelines for what good submissions should probably have:
About something that someone might reasonably think is correlated or anti-correlated with Trump's election.
Is not directly about the election. Nothing like "Will Trump win Pennsylvania".
Is not extremely likely or extremely unlikely to occur (this makes the correlation much less worth betting on, even if it exists).
Has well written criteria, and so is unlikely to resolve N/A or to %.
The market has a decent chance of resolving in 2024 or shortly after.
I may N/A any submissions that I view as not meeting these conditions, or which simply aren't interesting enough to attract traders. But these are guidelines, and not strict rules. I'm open to suggestions for improving these criteria, please let me know any you think of.
@BoltonBailey That’s what’s so hard about this market. For some pairs you think, ‘surely there must be a relation’, you just don’t know which way it goes.
Has well written criteria, and so is unlikely to resolve N/A or to %.
FYI, the math works just fine if one or even both underlying markets resolve as a fraction. Just resolved the derivative to the same fraction x in the Trump yes case, or 1-x in the no case.
And if they both resolved to a fraction (presumably not in this market, but maybe in some other related market...), say x and y, then the derivative market just resolves to x*y+(1-x)*(1-y).
@EvanDaniel Hmmm I'd still prefer if people tried not to submit such markets but that's good to know if anything does end up resolving as a fraction.
Oops it did not register to me that if Biden wins a second term he is much more likely to die "during his presidency" @EvanDaniel you might want to put back the limit orders I filled.
@BoltonBailey Yeah I didn't notice this at first either 😅
I'll N/A other submissions that might resolve after 2024
@BoltonBailey yeah, that's why I put them there... Wasn't sure if there was confusion or actual disagreement going on. It's out then back up if I still had mana free but I've been sinking it all into markets that resolve at the new year.
@Joshua Just to confirm: even though the market extends into his second term only death before the end of 2024 would count for this market?
@NicoDelon Uhhh this is getting into the weeds but I'm not sure my rules as written actually say that. I encourage people to not add markets that can resolve after 2024, but did not fully ban them and didn't notice this one, and it actually does seem like an interesting question.
So now I'm not sure what to do. Should I N/A it? think it doesn't work to only count market resolutions this year for markets that might resolve in future years because then the displayed probability of the Biden Death market is misleading. Also, Evan and Bolton already traded with Evan assuming future deaths would count, which the rules don't say they wouldn't.
This is additionally complicated by the fact that I think the Recession market is the best correlation we have... but actually it resolves as late as 2026 because we don't necessarily know we're entering a recession as we're doing so. So I don't want to blanket-ban everything with a close date after 2024.
This is complicated...
@Joshua How about this bit? I understand it’s about the rules of submission not resolution but I’d just like confirmation.
Is about an event that will take place before the end of 2024.
@NicoDelon Okay I think it was a good market suggestion by Matty and a very savvy bet by Evan so I don't want to N/A it I think, since the guidelines are stated to not be strict rules.
Instead I think I'm just going to try to make it more clear in the options when a market resolves. Doing an update now.
@Joshua Yes, I agree, I lost the mana fair and square and I think for me it was only about 30 mana EV anyway.
@Joshua I'm fine with (almost) any of these options. I'm a big fan of trying to get the market to say the right thing, especially if problems are found very early in the market's life. I appreciate the effort going into careful consideration here!
I think the only one I don't like is the "only count death during his first term / before the election" one; that makes it a derivative of some market other than the listed market, IMO.
@EvanDaniel Yeah I think it's important to use the same rules as the linked markets, so if the linked market goes beyond 2024 then that needs to count for this market. If someone wants a 2024-only version that market probably also exists and could be added.
@BoltonBailey Yeah I think someone should certainly try to do something like this for directly election related markets. But I'm trying to figure out here is if there's anything indirect but still predictable.
If the markets are "uncorrelated" that means that P(A) = P(A|B) -- the probability of A is the same as the probability of A given that B happens.
By definition, P(A|B) = P(A&B)/(P(B)). A little algebra will get you to:
P(A) = P(A|B) implies P(A) = P(A|~B) and also P(B) = P(B|A) and so on.
"Market A resolves the same as B" = P(A&B) + P((~A)&(~B)) = P(~(A XOR B)).
In the uncorrelated case, you'd expect P(resolves the same) = P(A)*P(B) + (1-P(A))*(1-P(B)).
So, for example, this pair:
/chrisjbillington/will-the-us-enter-a-recession-by-20 (43% as of writing)
/duck/will-donald-trump-win-the-2024-pres (40% as of writing)
The uncorrelated case would be 0.40*0.43 + 0.60*0.57 = 51.4%.
And since the probabilities of the two markets aren't identical, the available trading range is reduced (100% that they both resolve the same way is contradictory).
@NicoDelon Yeah it's really very hard. Almost nothing is monocausal enough, except of course directly election-related markets. But if someone can find something I'd be very impressed.
@Joshua It’s the resolving before the end of 2024 that makes it so hard. Otherwise there’s a bunch of nearly trivial markets (about, say, SCOTUS retirements/nominations).
@NicoDelon Yeah, if this market worked we could identify "bell-weathers" that would help us predict the presidency. Which is very difficult.
I just made another version for Trump as president, for events through 2030.