Unlike Reference or Choice fields, List fields can store more than one value (sys_id).
When a breakdown is applied to a List field, each sys_id in the List field is evaluated to find a matching breakdown element. The number of evaluated sys_ids is therefore usually higher than the number of records that contain the List fields. As a result, breakdown scores can be higher than the overall indicator score.
This behavior can be seen most easily with indicators with the Count aggregate. With a breakdown of a Reference or Choice field, the total breakdown scores are always equal to or less than the overall indicator score. Each Reference or Choice field has one value, there is one field per record, and the count of all the Reference or Choice values cannot be more than the Count of records for the indicator. However, if you have a breakdown on a List field, each record can have multiple values in its List field. Each of these values that matches a breakdown element contributes to the Count for the breakdown score on that element. You could end up with an indicator with a score of 10, representing 10 records, where one of the breakdown elements on that indicator has a score of 15, representing 3 records with 5 List field values each.
Similarly, if you define breakdown relations on the elements of a single breakdown that is applied to a List field, the scores of the child elements can be greater than the score of the parent element.
For Field breakdown mappings, this behavior is expected. To work around it, create a Scripted breakdown mapping where the script treats all the values in a List field on one record as a single score.