I don't consider Alan (r.i.p.) to have been an expert on gambling - I guess he was just speaking about his observations - observations are not proofs
in America, everything, even trivial things, are studied to death
so, in thinking about it - I would have to believe that if dice influencing were possible it would have been proven somehow and somewhere with thousands of rolls
academics love this kind of thing
and guess what - I found an academic study - see link
these PhDs built a machine they called "Lucky Lil" and calibrated it to throw the dice in a way that they believed might produce non random results
they tested both short rolls that did not hit the back wall and legal rolls that did hit the back wall
but alas - "Lucky Lil" failed to produce any results that were significantly different than random results
the is I believe because the action of the dice becomes chaotic once it hits the table and then even more so when it hits the back wall with the pyramids
https://digitalscholarship.unlv.edu/...5&context=grrj
.