An attempt to fix the "Will this close above 50%" market. (READ DESCRIPTION)

1kṀ120k

resolved Nov 1

Resolved

YES

ALL

There is a popular form of market on manifold, with something isomorphic to "Will this market close above 50%".

In theory, this is the nicest possible market, it it always resolves in the direction of highest net gain. However, there are two major issues with this market.

First, it is very common for someone to enter the market right before it closes to be the last to influence the market resolution.

And the other issue is that people betting large amounts of mana ("whales") are heavily over-represented. So this market will focus on resolving these issues.

For the whale problem, for this market, only the square root of your position will count. For example, a user with 10,000 YES shares will contribute 100 points to the YES side. The side with the most total points will be the side that it resolves to. This will make it much more difficult for one "whale" to control the market.

For the first problem, after the market closes, a random integer N from 0 to 100 will be rolled, and trades made within the last N hours before the closing time will not be considered in the resolution. This will prevent it from getting sniped.

Update 2025-10-28 (PST) (AI summary of creator comment): The creator confirms the market will resolve (not resolve N/A).

Based on the random roll N (0-100), rolls of 72-100 will result in YES resolution.

Get

1,000

to start trading!

🏅 Top traders

#	Name	Total profit
1		Ṁ3,837
2		Ṁ1,305
3		Ṁ939
4		Ṁ805
5		Ṁ645

People are also trading

Will this market resolve below 50%?

Sort by:

N=3, resolves YES (in fact no trades were made jn the last 3 hours)

I apologize for the unexpected extension earlier on, I should have just resolved based on what the market said at that time

🤖

@JiaheLiu your random number is: 4

Salt: 5aqhx83wdxi, round: 5552966 (signature 893c0e5945a2bc21da687ff0137aa9e2bc59a31e7cea6a69d469a1ae8190a148eadbc5dd789aad367a2055b38ab0145d01b73cbe3661ff3aa78d9ea77339ae58b5c9aab160e7cf8010137fdd127afcac72d1ab80bf345c324ba3d5345ad937d5)

🤖

@JiaheLiu you asked for a random integer between 1 and 101, inclusive. Coming up shortly!

Source: GitHub, previous round: 5552964 (latest), offset: 2, selected round: 5552966, salt: 5aqhx83wdxi.

@FairlyRandom 101

reposted

https://manifold.markets/4fa/another-attempt-to-fix-the-will-thi?r=NGZh

Another attempt to fix the "Will this close above 50%" market. (See description…)

50% chance. This market will resolve based on the sums of the square roots of traders' positions. For example, a user with 10,000 YES shares will contribute 100 points to the YES side. The side with the highest point total will be the one the market resolves to. To discourage sniping, after each close, there is an 80% chance that the market will be extended for approximately 24h. (I will ask FairlyRandom for a number from 1-5, 1-4: extend, 5: resolve. This will be repeated every day until a 5 is rolled.) Inspired by: https://manifold.markets/JiaheLiu/an-attempt-to-fix-the-will-this-clo#0p0x2ozj3sne

Sum sqrt(YES shares): 729.130
Sum sqrt(NO shares):  453.687
Hours until close: 17.7

bought Ṁ25 NO

@traders there is still time to defect!

@ItsMe …but a 70% chance that it won't count.

Sum sqrt(YES shares): 762.148 Sum sqrt(NO shares): 345.372 Hours until close: 29.9

at 99.7%, that's good odds

opened a Ṁ44 NO at 99.0% order

@ItsMe Fair point!

Self advertising

Sum sqrt(YES shares): 818.668
Sum sqrt(NO shares):  400.661
Hours until close: 34.3

Sum sqrt(YES shares): 812.355
Sum sqrt(NO shares):  415.471
Hours until close: 54.7

Sum sqrt(YES shares): 556.249
Sum sqrt(NO shares):  272.949
Hours until close: 66.9

To clarify, the market will actually resolve this time.

72-100 resolve YES (correct me if I'm wrong here).

@JiaheLiu While I like the concept of this market, next time, don't just extend it if you feel it has a too few traders – instead, resolve it the way traders were expecting and make a new market instead.

@JiaheLiu I think this can now resolve YES independent of what number is rolled, but let's roll for completeness and maybe we'll even hit a timestamp on which I ran my script.