Why isn't a 2nd degree a better fit?

Please use the “pencil” icon in the thread title, and move the thread to the correct course.

Currently, you’ve posted in “General Discussions”, so we don’t know which course you’re asking about.

In classification, the job of the decision boundary is to split the dataset between true and false regions.

It appears to me that degree=2 does a pretty good job. There appear to be about the same number of examples that are mis-classified, and by about the same amount.