Bongard Problems sorted left have the keyword "seemslike" on the OEBP.

In a "seemslike" Bongard Problem, some relevant information is left out by the way objects are displayed. Solutions to "seemslike" BPs sound like "Seems like a __based on the information available vs. seems like a__ based on the information available". The two sides are typically negations of one another (keyword @notso).

Since there is information missing, examples can only be sorted by trusting they appear in a way that hints psychologically at what they actually are (see @help).

Bongard Problems where some information is left out from examples and they are only sorted on the side they seem to fit vs. other Bongard Problems.

Bongard Problems sorted left have the keyword "seemslike" on the OEBP.

In a "seemslike" Bongard Problem, some relevant information is left out by the way objects are displayed. Solutions to "seemslike" BPs sound like "Seems like a __based on the information available vs. seems like a__ based on the information available". The two sides are typically negations of one another (keyword @notso).

Since there is information missing, examples can only be sorted based on what they seem like, trusting they appear in a way that hints psychologically at what they actually are (see @help).

Although the rule sorting the underlying objects may be @precise, the rule sorting the projections of the objects with lost information is @fuzzy and @subjective.

See also @left-couldbe (and @right-couldbe), concerning situations in which the information given is enough to determine when some examples fail to satisfy a rule but not enough to determine when they do satisfy the rule, the Bongard Problem's solution being "could be ___ vs. clearly is not". In contrast to "seemslike" Bongard Problems, @left-couldbe Bongard Problems can be @precise; both evidence of fitting right and absence of evidence of fitting right is clear-cut, so no psychologically helpful hints (@help) are needed.

Aaron David Fairbanks