Notice the impact from the first part of the weight that, without relevance information,
is nearly the same as an idf weight (as we discussed in section 7.2.1).
Because the term “lincoln” is much less frequent in the collection, it has a much
higher idf component (7.42 versus 2.44). Table 7.2 gives scores for different numbers
of term occurrences. This shows the importance of the “lincoln” term and
that even one occurrence of a term can make a large difference in the score. Reducing
the number of term occurrences from 25 or 15 to 1 makes a significant but
not dramatic difference. This example also demonstrates that it is possible for a
document containing a large number of occurrences of a single important term to
score higher than a document containing both query terms (15.66 versus 12.74).