Figure 7:
Linear regression between n of small letters and total n of errors by imitators.
