Similar to the Fuzzy level % for column but this value is used to determine when the group should be considered similar enough to be presented as potential duplicates or matching records. 

For example, group 1 contains two columns – column1 with result 90% and column2 with result 80%. If both columns have weight 100% then group result will be 85%. If column1 has weight 100% and column2 has weight 50% then group result will be near 87.5%