[Retros] Fastest selfmates in 1?

[François Labelle]

Hi Noam,

I'm happy that you like the computer solutions. It must be especially 
interesting to you since you've spent time thinking about the problem. 
Like you, I noticed that the human-discovered positions have large 
counts. If the human discoveries are a random selection, then they seem 
to be randomly distributed among *games*, and not among positions.

I don't have a Stack Exchange account, so please report my findings 
there. I think you can do a better job than me making it human-readable. 
In your post, you can link to 
http://www.pairlist.net/pipermail/retros/2020-December/004895.html as 
the original source. I guess the unique SPG should be considered 
published with the source being the Retro Mailing List. You can of 
course tell Tim Krabbe or others about the findings, and they can 
republish if they want.

     François