One of the important benefits of Data Cloud is its ability to unify profiles the exist in multiple data sources. Using rulesets,  a user can create match rules which act as criteria for deciding whether data from one object ought to be 'unified' with data from another object(s). (eg. row from marketing cloud consolidated with a row from service cloud into a new unified profile object). Many match rules can be created which, when chained together become a giant filter having 'or' logic. This means that if rows between objects satisfy any of the chained rules, then those rows are deemed to be a match. Imagine if there a a total of 500 rows from 3 objects and Data Cloud Admin determines that rows with matching last names with identical emails be deemed a match (criteria 1), then adds 2 more criterias-- say... same pasport numbers (Criteria 2) and same Drivers license numbers(Criteria 3)...after running the identity resolution process, it results in 500 rows being unified into 400 rows in the unified profile object.This results in a consolidation rate of 20% (1 - 400 / 500). During UAT, the end-users think that the 3 criterias were too aggressive in defining what constitutes a match, the Admin then can takes off the 2 criterias and find that 500 rows only gets unified into 490 rows which results in a 2% consolidation ( 1 - 490 / 500).Intuitively it makes sense...the less criterias, the stricter the rules, the lower the consolidation rate. The more criterias, the looser the rules (remember its using OR logic not AND) , the higher the consolidation rate.Snipped above are 2 rulesets--the one on the left only had 1 criteria, the one on the right had 4 criterias.