Just wanted to check something - Legs's run has him exit cashier then move the cursor before going back in. I'm using Bizhawk and I don't have to move that cursor at all to get back in to the cashier, is there a requirement to do so or is that just free time to improve the record?
EDIT: When timing a run do I divide frames by 60 or by 59.73 or what?

Hi!

It's free time. I'm holding left preemptively to be moving left in the menu. If I was mashing faster I would enter back into the help desk faster

You should just have to divide by 59.7275, but I retime my runs in Premiere. I could time it in Premiere if you have a link and want to compare the times.

Legs

Gameboy Framerate is 59.727500569606 (according to tasvideos.org). With a run like yours with the emulator framecounter shown, that's what's used.
If it's a recording without framecounter or from console, I use Avidemux or Virtualdub to time the run.

EDIT: Since your video is 30fps, it's possible that the real start and/or end frame got skipped, but it's not an easy thing to fix...

I was curious if a 60fps recording of this existed too; Id love to more closely see how many frames are left to save. It's a super good run.

Well thankyou so much! I hope my initial question didn't come across as boastful or crass, I realised after asking it that it might. I'll do a TAS of this tonight, and get back to you with the results. And i_o_l, thankyou for the info on the GB divisor! EDIT: is there a way I can record that will reproduce at 60fps?

Here's the TAS! The best possible time seems to be 1.021 seconds.

So I press START on frame 621, 628 and 638. The cashier comes up starting the timer at 645 and I immediately press A, then again 3 frames later. 9 frames after that I re-enter the cashier. You may notice that the cursor is initially positioned outside of the YES/NO area - so you wouldn't gain any time by holding left beforehand. 8 frames later I start moving left, after 26 frames I press A. 3 frames later A again, then after 2 frames as the final input. The taxi appears and ends the timer at frame 706.

Important question:
RNG for the GB casino games seems to be entirely frame based. I typically keep the frame counter on my emulator. I found the entire distribution of the Cleo/Caesar wheel and I'm going to actually hand produce the results for each frame to try to find a good clump for betting on.

So would you feel like using frame counter during an actual run would count as tool assistance since it is not information a casual player would have? Or would you feel like it is difficult enough to hit specific frames consistently anyways. I feel like frame count should be okay simply because otherwise one could spend a million tries never getting the combination of spins you want.

I feel like using a framecounter like that is no different from using a timer like LiveSplit that you start at the same time as you hit reset.

It's nice that we both feel the same about that, though I would welcome (and encourage) discussion of this and comments from NES and other version runners too, since logically if we allow its use here then use on other platforms is inferred. I submitted a slightly sloppy run but which succeeded in the frame marking aim. I certainly don't intend to gazump other platforms' runners in this, though my buddy CoKami might be slightly miffed that I nicked his nascent WR