So newbies tend to focus too much on getting a model with a high R-squared. What a "high" R-squared number is depends on the field of application. Another thing is that R-squared close to 1 may mean your model is poor for interpreting relationship between IV and DV.
The video could be better for sure. Noise from building works in the building unsettled me.
Visit me at: http://www.statisticsmentor.com
...
https://www.youtube.com/watch?v=dV29oHF8Dhk