I know I should be able to predict the OLL in easy scrambles like this, did you practiced with unlimited inspection time or that ability just came with time?
Of course I practise with unlimited inspection most of the time but I try not to spend A LOT more than I would have in comp. It's no use to be able to predict PBL if it takes 2 minutes. Let's say I don't activate csTimer 15 second timer but most my inspections don't go over 30 seconds I'd say, unless I see an interesting case I want to learn something from.
It does help a lot but, anyway, don't get me wrong, I'm not able to predict OLL unless the face is 3-4 moves like this one. You have to admit that this one was kind of easy to predict. Other OLLs with no adjacent stickers can be a bit trickier. In particular I tend to predict wrong sune/antisune, mistaking one for the other, because I don't identify correctly whether the sticker close to the one on the U face is to its right or its left. To be honest I don't know how people that does CLL/EG can one-look 80% their solves. I think it's just practise but young minds are better at spatial reasoning, I guess.
EDIT: I forgot to comment it in the video, but here is my nice rule of thumb for predicting AUF on the double diag swap PBL with 2-side recognition, just in case you don't already do that (probably you do):
- If you have vertical bars, then no AUF,
- if checker pattern of opposites, then U2,
- in any other case, see what you would need to have bars (U/U') and do the opposite or just align the bars before doing the PBL, but I prefer to post-AUF as it gives me some time for my brain to process what I saw while I perform the PBL.