Hi @Tyler_Carrier, the –p-error-rate should only count towards mismatches.
To clarify, cutadapt makes use of IUPAC codes by default, i.e. --p-match-adapter-wildcards. Thus, if a base in the sequence matches given the IUPAC letter of your primer / adapter, then it should not count as en error / mismatch.
For example, if your primer contains an M then the bases A or C will be considered a match at the corresponding position of your read. Other bases, will be considered a mismatch. Assuming the sequence is the same strand as your primer sequence in this case.
Note, you can also enable the --p-match-read-wildcards to make use of any potential IUPAC bases in your sequences.