A method of error detection for a data packet, the method comprising the steps of: i) identifying a set of non-compliances (N), the non-compliances being illegal bit sequences according to a coding standard; ii) identifying a first subset (N+) of non-compliances that are to be treated as errors; iii) identifying a second subset (N.) of non-acceptable near-compliances; iv) decoding the data packet according to the coding standard; and v) adaptively deciding based on the first and second subsets whether to treat a detected non-compliance within the decoded data packet as an error or as an acceptable near-compliance.