Skip to main content
MANIFOLD
Will someone find a bug in the *fifth* version of my Time Weighted KBC market code by market close?
6
Ṁ150Ṁ482
resolved Nov 2
Resolved
YES

Context: /DanMan314/time-weighted-keynesian-beauty-cont

I'm creating this market both to see if the code is good enough, and as an incentive to find bugs. If I find a bug, it doesn't count for NO or YES (I'll keep it as the same version).

I will use my discretion on what counts as a bug, but generally it should be capable of impacting the final resolution of the market - I'm not interested in typos.

I won't bet on this market.

Fifth version of the code: https://pastebin.com/c3EMy9p8

Market context
Get
Ṁ1,000
to start trading!

🏅 Top traders

#TraderTotal profit
1Ṁ67
2Ṁ32
3Ṁ20
Sort by:

If final percentage is somehow exact, script will output e.g. "100% 2, 0% 2".

predictedYES

Also, some ideas to improve code:

L77, L85, L90: divide by total_time only once, in the end.

L90: sum([a["weighted_sum"] for a in market_info["answers"]]): remove [] within sum, as it's capable of summing over generators.

@AnT Thanks! I’m away from my computer right now but I’ll count this.

@DanMan314 I’m surprised so many bugs are in the final linear weighting, which I thought was simpler.

Joshua boughtṀ1NO

@Joshua my hero

predictedNO

😭

@Joshua I let you down 😭