>>Yep, memorizing the intermediate results is the hard part. Which is why I can't do it :).
>
>Maybe it's similar to silva mind control where they attach symbols to numbers and names...

Dunno, these things were always a mystery to me - how does it help at all? Our youngest is learning about those things now in her psycho(logy) course, and the whole thing seems to be based on forgetting, i.e. if I imagine that a spool of thread is to represent a 35 in my mind for the duration of exercise, I need to be able to forget it afterwards, or I'll have all sorts of things and numbers in my mind attached to each other.

It may be that attaching visuals to abstracts may help those who have trouble with abstracts, but this guy is mathematician, he shouldn't need that. I think it's just showmanship.

The part I don't get is guessing the missing digit when they multiplied a known 4-digit number with an unknown 3-digit. I have no clue how he got that. For everything else, there's a clear calculation which can be simplified and some bits of it memorized, i.e. where practice makes perfect. Calendar is an obvious example for that - you only need to remember the day 1800, 1900 and 2000 begin, then how many years does it take to go full circle, i.e. what's the period until you have the same day of the week start the year. Then you have to memorize the pattern of days-of-week on the 1st of every month, calculate a bit of an offset, modulo 7 and voila. You need to memorize maybe 40 numbers, of one or two digits size.

But this thing with missing digits is still beyond me.