problem_data dataset has many missing data. I tried to understand the relationship between level_type and tags with the points and I found a small strange thing. For example: when the level_type is: B and tags is: implementation the points differ from a problem_id to another.
The first question: why would the points differ even though the level type and the tag is the same ?
The second question: for level types above H ( from I to N), all the points are missing. how can we get an idea about the points ? for the other types ( A to H ) we at least can see the mean and median which gives us an idea about them.