In the talk when he showed the 3 similar maps getting equal score, and then the green one getting a much higher score, that could easily be because people couldn't decide between the 3 similar maps, to they went with the one different map. To do a proper comparison the test should be 1-1 or 3-3, not 3-1.
Comments
In the talk when he showed the 3 similar maps getting equal score, and then the green one getting a much higher score, that could easily be because people couldn't decide between the 3 similar maps, to they went with the one different map. To do a proper comparison the test should be 1-1 or 3-3, not 3-1.
They wanted green because it's a python library. He goes into the details in the talk