OK. This analysis shows how far from the original OFCL (NHC's best track: corrected consensus forecast track) the 3-day OFCL track runs are, from Thursday (the inception of the hurricane), to Friday, to finally, Saturday (today). This analysis isn't meant to trash any hurricane prediction science. It is meant to demonstrate that in just 1 day, the predicted official NHC consensus track can change drastically, putting one in a situation by deminishing your time-window of escape. The models with the best long-term track record (2014 NHC analysis; and these may change from year-to-year) are: the FSSE Superensemble and OFCL (If I recall, they vary by only ~~ 30 nautical miles after forensic analysis). There are individual model ensembles that sometimes outperform the top long-term product, but these are anomalies. So, this is why the NHC OFCL is the only track shown to the public (to minimise confusion).