We have a small edge case in our index when searching for the word "Moon". It matches to a company that we should call "Full Hardware". Would it be right to guess that since "Full Moon" appear together in some documents, that this would cause "Moon" to find records with "Full"? It scores low, but we were not expecting to put in a min_score filter.
Below is the explain output when hitting the explain api (version 1.6)
{"_index":"search-company_1","_type":"company","_id":"123456","matched":true,"explanation":{"value":4.793683,"description":"weight(name:moo in 46422) [PerFieldSimilarity], result of:","details":[{"value":4.793683,"description":"fieldWeight in 46422, product of:","details":[{"value":1.4142135,"description":"tf(freq=2.0), with freq of:","details":[{"value":2.0,"description":"termFreq=2.0"}]},{"value":7.7477617,"description":"idf(docFreq=2956, maxDocs=2519805)"},{"value":0.4375,"description":"fieldNorm(doc=46422)"}]}]}}