While McLaughlin et. al. (2018) argue for similar trends in the medieval Irish historical record and the archaeological radiocarbon record, part of their results are due to an ad-hoc bandwidth being used to calculate the kernel density estimates (KDEs).  This contribution looks at using a data-driven bandwidth to re-calculate the KDEs and also look at the first derivative of the KDEs. The results here indicate the radiocarbon record declines much sooner than the early 9th Century and not recovering again until the late 11th Century.  Comments are also noted on the Irish annals and the approach, for at least one region, on the use of radiocarbon dating.