NBOs Explanation

I get many similar questions from NBOs so have tried to put the common answers into a standardized format.

Use Anonymized Data

I STRONGLY suggest using anonymized data. I have seen far too much confirmation bias if names are used - in both directions. "I know this pair is shady - they have brown eyes - therefore assumption is cheating", "I know this pair is clean - they have blue eyes - therefore I don't find anything". I have even had top experts testify (in cases) that this pair is clean, but if I give them anonymized data as a 'placebo' for another case, they are adamant they are cheating. To me: anonymized data is a must. But, just my opinion.

Use Placebos

If I give you a pair, "anon-101", and say, "these are cheating" then the investigator will "find" cheating actions.

If I give you two additional placebos, e.g. anon-102 and anon-103 and say to an investigator, "one or more of these three pairs may be cheating" then the investigator does not know who is cheating.

Cases are Different

Some cases are easy. You can use the unusual_leads concept or the special_boards concept.

Some cases are hard. You need to analyze individual boards.

I can generally help by indicating what type of case it is likely to be.