follow-ups on predictions that speedrunners might eventually do things

smb1 tas tie is coming any day and sm64 1 key already happened

During my first daily blogging spree last November, I made a post arguing that one should generally expect even very difficult techniques to be implemented in speedruns. This post ended with two predictions, both of which have seen appreciable developments since that post came out. Given that it is currently 11:30pm and I need to somehow speedrun getting a blog post out within the next half-hour or else fail my impromptu daily blogging challenge that I committed to at the last minute, I figured now might be a reasonable time to revisit those predictions.

They’re gonna tie the Super Mario Bros. TAS any day now

Niftski is going for it. I don’t know exactly what his stream schedule looks like, I haven’t been paying that close attention to the grind, but it’s pretty frequent where you can tune into a stream and watch him trying to tie the TAS. Just the other day I personally watched him get a run to 8-4. The run, of course, did not end up tying the TAS, and wasn’t even a record. But it’s not like I’m needing to watch speedrun attempts all day to catch a run that makes it into 8-4, I just happened to open the stream when one was about to happen. So they must be pretty often. And of course frame-perfect 8-4s aren’t easy to get very consistent at, but Niftski’s done them several times!

LeKukie was also going for it. I say was because he hasn’t been doing attempts for a little bit. I think he might get back into it at some point, but he’s ran into some personal issues that have been making it a little difficult to focus on the grind:

Heartwarmingly, the SMB1 speedrunning community has come together to fund repairs to LeKukie’s family’s house, raising substantially more than the original goal. I’m not totally sure when LeKukie will be able to get back into doing speedrun attempts, but if Niftski doesn’t get the TAS tie before then, LeKukie is also going to be a serious competitor.

We’re not quite there yet. But it’s no longer the ambitious¹¹ Okay maybe it wasn’t that ambitious. It was maybe like, ambitious by normal people standards, but not really by the standards of anyone who’s particularly well-calibrated about these sorts of things. prediction that it was a year ago. It’s coming. We could hear the news come in any day or any week now.

They did 1 Key RTA all the way back in February

As far as I can tell, the current RTA record for SM64 1 Key is 3:10.

That is, of course, three hours and ten minutes. You couldn’t do it in three minutes, the TAS takes four!

Now, this probably shouldn’t quite count as a fulfillment of my prediction. I was definitely thinking of someone doing 1 Key fast enough for it to actually be a new Any% record, not someone managing to get Moat Door Skip by hyperspeed walking for three hours.

But three hours or not, this does establish that specifically the parallel universe movement aspect of the trick is RTA-viable. So the only missing piece is preserving the speed out of the Vanish Cap stage, and then getting a viable PU setup that uses speed from there instead of from the hyperspeed walking. (And then, of course, managing to string everything together in one run). This really does increasingly look to me like something it would be totally reasonable to go for.