cross-posted from: https://programming.dev/post/6822168
I was watching a twitch streamer play the game pogostuck (A game similar in frustration and difficulty to Getting over it with Bennett Foddy — Don't Fall!).
They were also reading chat at the same time (usually out loud, as well). Multitasking.
Lots of sources (here's one) say that true multitasking is impossible. Rather, it's very fast switching, where there is a degradation of performance.
Knowing this, I naturally made it my mission to trip the streamer up with seemingly benign messages.
I was sharing some actual information about another streamer who beat another game, but a made a typo something like:
I remember a streamer beat the game a game ...
And I noticed how much more the streamer struggled to read this compared to previous, accidental typos (missing spaces, extra spaces, etc.). He spent a good 5 seconds on this message, and during the process, he fell really far. 😈
So I decided to do some testing. Inserting words, swapping them around, and whatnot, to see what tripped him up the most. Most typos didn't affect him.
There was one typo that tripped him again, where I said something like:
If it wasn't for a for
So it seems to be repetition? But I couldn't always replicate this with other forms of repetition.
Later on, I copied the two guards riddle, with an alteration:
One of the guards always lies and the other always lies as wekk. You don't know which one is the truth-teller or the liar either. However both guards know each other
Sadly, I didn't cut the part about "don't know which is truth teller or liar" out.
The streamer spent a good 5 minutes interpreting this puzzle, and eventually interpreting it as the original puzzle. Then, he was trying to solve a riddle, game, and read chat all at once.
He was stuck on the bottom until he gave up on the riddle (I revealed that I meant what I said when I said both guards lie). 😈
Anyway, that was a bit off topic but still relevant.
I'm wondering if any studies have been done on this? I know studies have been done on human's ability to read words with the letters partially scrambled, but what about typos?
How can I improve my distraction game (with plausible deniability of course)?
I've been thinking about this for a very long time. The only platform that actually passes all of Soatok's tests is Simplex. But it's only briefly mentioned in the article about session
Wait no. I think only server builds are reproducible.
But there is an Android build on F-droid and F-droid does offer reproducible builds, so that's a way to get them.
A big problem, however, is that the creator of simplex is an antivaxxer, anti-DEI, anti-abortion, climate change denialist, etc. I can understand why people would be hesitant to push a platform like that, given they may also be pushing the creator's beliefs. I think it's okay to do the recommendation with a little asterisk of "please be aware that the creator of this app is crazy".