Setting > Data Management – Duplicate Detection Rule

• Rules can be something which define how the detection will work and what basic it will triggered.
• Duplicate Rules has Status as published or unpublished.
  Unpublished Rule will not take effect, only publish Rule will be triggered.
  Default When rule is created is always Unpublished state, because Only 5 Published Rules are allowed for each entity
• Rule has Base Record type, which is the Entity option set where we select the entity on which rule will be fired.
  On save of this Entity record, the rule will be triggered

• Rule has Matched Record Type , this is also an option set which is used to compare the value.
  In many case both Matched record type and base record type is same.
  In some case like Contact should not create if same email address is there already in lead entity. Here Base can be contact and matching can be lead
• Case Sensitive –  if this option is checked ,it make sure that checking values also check the case.
• Excluding Inactive matching record – System may have archival data which will be inactive state and that need not to counted in duplicated check, this can be implemented using this options
• Criteria Grid – This contacts a list of rules on each fields.
  > If Base record type and matched record type is same,  you will get single field column as both entity has same list of columns. And can be matched with same fields only
  > for e.g. if Account is selected in both record type and in field if you select email as field , it will be compared only with email no other field than email can be matched to find duplicate.
  you cannot match email1 should not be equal to email2, for this you need to write business rule or Script.
  > If based Record type is not same as matched record type,
  you will get two field selection option, base record field and matching record field.
  Basis on base record field selected option, matching record field option set will be filled, based on data type. As data type for both field cannot be different.
  > Criteria – This is also based on field selected, for e.g. if you select some lookup in fields , you will get exact option only but if you select some name field there are three option to set the criteria Exact match, Same first character or same last character.
  > No of Character – This is enabled only when Criteria is selected as same first character or same last character, as this takes input of length to be matched from start or end, length starts from 1 to 50
  > Ignore Blank values – To ignore null checks or while duplicate checks if admin don’t to check data which has null value this option can help in that requirement.
  ## If the duplicate detection rule contains only one condition, blank values are ignored during duplicate detection job.
  > There is something called Current Matchcode length , This is the length on which duplicate detection can be checked,
  450 is the value which is max value.
  We can have one fields of 450 character or multiple fields sum up to 450 character, we cannot add rule for more than 450 character field/fields.
  it will not allow to save the rule, unless is 450 or less.
  Character matched by CRM in backed cannot exceed 450 character, as giving more bigger length option can degrade the performance while creating/updating/importing records.
• Rules are not active unless its published
• Only 5 published rules are allowed per entity.
• Microsoft recommended to use duplicate detection rule for unique records such as email or some custom code etc.

Links to refer :  https://docs.microsoft.com/en-us/dynamics365/customer-engagement/admin/set-up-duplicate-detection-rules-keep-data-clean