I will resolve any answers to these questions as soon as I see the xkcd posted for the given day and have time to resolve. If it ever appears that the post was skipped (I don’t see anything before the next usual post day) I will resolve all answers to N/A, with the exception of any questions regarding whether or not the xkcd is posted late. If the answer is ambiguous, I will wait to see thoughts posted on explain xkcd, and will either resolve or conclude N/A on the next regular post day based on the possible consensus there.
Others are free to add options. I will delete duplicates and ask regarding resolution criteria for potentially vague options. If you believe that there is potentially a vague option, feel free to comment so we can address the issue as early as possible.
If you have a fun idea but don't feel like adding the question yourself, either comment it or message me your suggestion, and I'll probably add it to the next xkcd market.
I may bet in this market.
DUE TO MANIFOLD INCREASING THE COST OF QUESTION CREATION, IT IS POSSIBLE I WILL NOT BE ABLE TO CONTINUE TO CONTINUE MAKING THESE UNLESS THERE IS ENOUGH ACTIVITY TO PAY OFF MOST OF THE COST.
1,000
3.00
@FH7979e I definitely read the question as binary. I suppose I could have resolved to the percent, but I’m still kinda new to making questions.
@EBurk on Wednesday. Back in the day, all comics were posted at exactly midnight; now this is less common
@AndrewG I'm going to take a Snip of the image including the border as closely as I can and then use this website: https://mk.bcgsc.ca/color-summarizer/?analyze
@EBurk Why take a Snip and not just download the image (avoiding imprecision from including area outside of the image)?
BTW I have used this command to download an image and measure the black percentage using ImageMagick:
curl -s "$(curl -s "https://xkcd.com/$number/info.0.json" | jq -r '.img')" | magick - -format 'Black fraction: %[fx:(1-mean)*100]%%\n' info:-
On the last 20 images i got these numbers: 9.43031%, 13.0816%, 9.61491%, 11.8382%, 11.8977%, 9.92533%, 10.9274%, 12.947%, 11.2207%, 8.13073%, 11.7813%, 14.7108%, 6.72917%, 6.10852%, 12.6851%, 11.1894%, 10.8807%, 10.8561%, 8.30181%, 8.77909%
@FH7979e Fair enough, I’ll download the image instead. I’ll probably just use the website I found. Hopefully the image is just black and white, so I can just refer to the non-white space.
@EBurk Actually, it turns out the precise measurement method used makes a big difference. There is a lot of gray pixels due to anti-aliasing:
My method counts a grey pixel as partially black. But if we count grey pixels as fully black, it significantly inflates the number of black pixels. On the latest image, my method gives 8.8%, but https://mk.bcgsc.ca/color-summarizer/?analyze gives 15% non-white pixels.
This is made worse by the fact that the website by default first downscales the image to 50x50 (it can be increased with the "precision" radio buttons). The effect of this is that an area containing both black and white pixels is turned into a single gray pixel, further inflating the percentage to 32%
@FH7979e Thank you for pointing out these issues. I made a rule for myself that this market wouldn’t become a job, so I am more likely to just resolve N/A at this point. If you would like to approximate the comic, I will default to your judgement.
@EBurk I'd be happy to give you the results i get form my command.
Note: In case the comic is not just black and white, i may modify the command (because currently it counts colorful pixels as partially black).
