🏅 Top traders

I count 73665 (using Julia's Dates default, I believe UTC).

NOT PRIME

@ScottLawrence will you be resolving this?

@Noit Working on it now. (This is one of ~5 markets I made last year which require actual effort to resolve. I'm trying to make time to resolve one per week.)

Might feel unimportant now, but resolution details will be crucial later:

-- Timezone?

-- Can you share a snippet of API requests that do the thing you plan to do?

@Boklam as it says in the market description, timezone is UTC.

I'll include all markets accessible from the API, but I'm not writing the script in advance

@ScottLawrence Oooops I didn't read, thanks

@Boklam This seems like a fun target for last-second manipulation...

By the prime number theorem the probability that a given number is prime is roughly 1/log(n), so if there are ~10k markets created then the probability is around 11%

@jonsimon There's the additional chance that someone intentionally creates markets at the end of the time range to ensure that there are a prime or non-prime number of markets. Assuming that the likelihood of this happening on either side is equal, I think that this may push the overall likelihood of the number of markets being prime very slightly closer to 50% by some amount.

@lukalot Nope, because they'd be just as likely to miss as not (best case scenario, supposing the number of markets close to the closing time is ~10k, you're looking at the density of primes in a small range of numbers just above 10k). For that and in general, the Logarithmic Integral is a much better approximation, and the anti-prime-market camp has better odds of accidentally missing a prime than the pro-prime-market camp has of getting a prime outcome. With 10^4 markets, there are 106 prime numbers in the next thousand and 103 in the following (112 in the previous thousand). The pro-prime camp has to create many markets to reach the next prime (8 or 9 on average) and the anti-prime camp only has to create one to avoid a prime, so it's not going to push results towards 50% until the pro-prime camp is more efficient than the anti-prime camp by a factor of 10 or so, because the anti-prime people don't really need to coordinate with each other to take advantage of the naïve odds while the pro-prime folks can't create ~20 markets simultaneously to each other and get better odds than picking a random number 15-25 higher than the current.

they'd be just as likely to miss as not

Huh? They'd be targeting a specific number, why would they be just as likely to miss it?

@IsaacKing It's an oppositional game, and if we assume that all prime and anti-prime predictors are playing at once with nonzero latency, it becomes essentially uniform random.

How do you know how many markets are created in the year? Is there a way to check or are you going to count them all?

@Fion Presumably he intends to use the data on https://manifold.markets/stats

@jonsimon not sure the precise figure I want will be available at end of the year. I intend to use the API.