Clinton's approval ratings were sinking into the low 40s during the midterms, but they rose to the mid-to-high 50s by the time of the general election. A different candidate might help, but the best way to make the Republicans competitive is to stop Clinton from regaining his popularity.
From I remember (without doing the correct thing and looking it up) Clinton's approval numbers climbed as did the personal attack (right or wrong).
A strange political dynamic came out of his post impeachment numbers. CW usually has it that when a President seeks re-election (or his party's replacement seeks election come retirement) his popular percentage will mimic his current approval rating. Therefore CW predicted that Gore in 2000 would produce Clinton's approval numbers, but there was a problem, which numbers?
There were two sets of poll numbers regarding Clinton's approval rating floating about: as for his presidency, after impeachment, for the traditional question, his numbers stayed high, and CW should have predicted an easy victory for Gore, but for the second question, one asking the on the approval of Clinton's 'character,' (a second question not usually posed by pollsters seeking an approval rating) his numbers sank to the low 40s.
In a way CW did turn out to be correct: Gore did mimic Clinton's approval rating(s), his popular percentage in 2000 (around 50%) split Clinton's two poll numbers down the middle. (High 50s with low 40s).