If I had to guess, I'd say it's one of two things.
In the video you provided, it's a probe compensation issue. The probe and the scope need to be properly compensated. I talked about that a while back during Scope Month here:
First, it's possible that the compensation network changes at 200 mV/div, I'm not sure how this scope's front end circuitry is setup. You could test it with a smaller Vpp signal (signal should stay completely on screen) and see if it changes.
Most likely, you're violating the oscilloscope's dynamic range . With a 5V signal and 200 mV/div setting, you're looking at 25 divisions of signal. That's surely saturating the front end amplifiers and can cause weirdness.
If you want to learn more about front end stuff, check out this talk from one of our designers:
https://www.youtube.com/playlist?list=PLzHyxysSubUmxGOMVpiKLxouweh2AAlG1 Feel free to skip video #1 if it looks too basic.