I finished a run with the royal regatta and everything went according to plan. Then this happened -

A few resets later I got the Regatta, woman combo again. However this time India 1900 was simply the right path and not the ending location. Any idea why this would happen?

I had this same sort of thing happen with a few other routes... the Russia 500 Hadrian's Wall Woman combo in particular seemed to vary on me despite the villain being the same. I got it often enough that I noticed the hours allotted were different, but those hours seemed to align with the different results I got playing it. But that doesn't help here since I believe I had the same hour quantity in my pb and it was the same mission. I never got the Regatta thing to show something different... saw this combo three times and it was all the same way it was in my pb.

My wild theory is that the result is modified in some fashion not just by frames until game start but perhaps each frame point (initial start, then elevator start, then mission start) might impact things in slightly different ways, which would explain maybe why you can get virtually identical cases but with just the location (and/or the final combination of searches) to be different.

In some instances I couldn't get the provided end location to show up at all like both Nobel prizes and Russian hadrians wall

Yeah, I had trouble doing some of the ones that Ucci had no problem with, but then others always seemed correct. I don't know enough to understand it all but it is definitely something that's tricky to frame-perfect get the same thing over and over.

This is still a very fresh strategy working only with the limited information from the TAS publication and what I was able to ascertain to make it work. As I've mentioned to aaron previously, there were instances where I've had...

• ...some case starting locations on the spreadsheet bring about a different gender suspect.
• ...some case starting locations have the correct gender suspect from the spreadsheet but a different set of locations.

The strategy is following as close to the observation published by Bobo in his TAS about the frame from the title to the elevator, and then minimising the frames to go to the second floor and, finally, entering the department to input your name (the second and third inputs capable of being buffered, but also interact on button lift rather than press). These are the three key input areas to influence the layout of the first case.

Probably doesn't hurt to point out as well that it is from a hard restart, but I am assuming that this is always followed. I haven't tested to see how the manipulation interacts with a soft restart.