Cute. Except:
1) Professor Lichtman absolutely hates Trump's guts, and openly admits his being a piece of shit could mean his model doesn't work any more.
He is quoted explicitly, here.
2) His model is not a valid statistical model. He is working from a small sample size from every angle, with subjective judgments on each "true/false" hit at best.
3) He is a historian, not a statistician. Historians -- and let me say this as a formal student of History --
do not make absolute predictions. Ever. We make hypotheses, theories, and capture our predictions with likely, maybe, the nuances are XYZ, and so on. Professor Lichtman was very clever in couching his language in very narrow, hedged terms.
4) He's selling a book.
Spin some more.