Let's stay with the previous example and suppose that our simplified similarity score is the count of identities. Between two words like BIRD and WORD we have two identities, R and D, so the score is 2. Whether this is much or little, depends on the other possible comparisons in the database. Comparing the word BIRD with a database of 10 words, we may get e.g. zeros in 9 comparisons and 2 in one case (for WORD), and in this case one "tends to believe" that the score of 2 is important.