Tomorrow nights guidance will mostly likely be the most important. What we're all watching for is the system coming out of the plains. The GFS has thus far handled Maria quite well, so I believe the euro will follow in its nature.
Per Don's post. This is the problem with only selecting a few hurricanes with previous tracks from origination (picture 1). While it is beneficial to select on tracks, you need to do a deeper dive into the data per storm.
September 19, 1955 would be the closest storm in terms of pattern, synoptic environment, and quick kick (picture 2.)
The most interesting factor in this pattern isn't the previous climatology, but the players involved (picture 3). Jose has now retrograded to the point where WV is having a hard time tracking him, and I suspect the ULL (small area of convection near Florida) is trying to go neutral. Jose is dying, he's on his way out, and it's obvious the high is rebuilding. However, the system coming from the plains is the kicker. You can see it advancing from Minnesota, to Western Texas. That ULL to the east of Florida, South Carolina, And Georgia is creating a weakness which is what is moving Maria closer to the N.C. Coast. The question now isn't Jose. Jose is now out of the picture for all intents and purposes, the question is: will Maria somehow finding a steering current fast enough to advance her into the N.C. Coast. While I personally think landfall around Kitty Hawk isn't out of the question, I'm not completely sold. However, ~20% on the landfalling scenario isn't accurate. The GFS has modeled Maria rather well, and reacts better/faster than the euro in these situations (12 v. 6 hours essentially). Anyone speaking of the euro until 0315 is dealing with old information. In order for me to give your 20% credit I would need the actual breakdown of hurricanes you selected, and why.
If you would like to play with your own hurricane tracks please go here: https://coast.noaa.gov/hurricanes/
I do enjoy everyones opinions, but please, lets use the sat imagery at the present time versus clinging to every run. I believe we can give models confidence starting at the 12z runs tomorrow.
Be well,