A lot of pop-celebrity-educator types like Attia or Stephen Seiler say that. Then you look at how the professionals train and you see pyramidal distribution almost universally. Something like 85% below first lactate threshold, 12% in "sweet spot" and 3% in "zone 5".
I've spent a lot of time reading a lot about opinions and then looking at logs of professional athletes [1] (in cycling as I am most interested in that). My conclusion is that training comes down to:
1)do a lot of volume, the more the better
2)do some "hard stuff" - if those are hard intervals, longer "sweet spot" intervals or a a mix of those (like 5 minutes at threshold and then 15 seconds sprint, repeat n times) matters little
3)at pro level do some training specific to what you are going to do a lot in racing
1)is by far the most important and the most reliable predictor of overall fitness
[2]https://www.youtube.com/@sportscientist - Stephen Seiler's youtube channel; he has done work on analyzing how pro athletes train but his conclusions are very simplified and it seems made to sell "polarized" training idea. When you look at the details in the data no one trains like that, the final distribution is almost always pyramidal, not polarized
I've spent a lot of time reading a lot about opinions and then looking at logs of professional athletes [1] (in cycling as I am most interested in that). My conclusion is that training comes down to:
1)do a lot of volume, the more the better
2)do some "hard stuff" - if those are hard intervals, longer "sweet spot" intervals or a a mix of those (like 5 minutes at threshold and then 15 seconds sprint, repeat n times) matters little
3)at pro level do some training specific to what you are going to do a lot in racing
1)is by far the most important and the most reliable predictor of overall fitness
[1] https://www.trainerroad.com/forum/t/pro-elite-training/14046 - is a nice thread with a lot of rides from whole weeks or months of training posted with power/heart rate data for various World Tour riders
[2]https://www.youtube.com/@sportscientist - Stephen Seiler's youtube channel; he has done work on analyzing how pro athletes train but his conclusions are very simplified and it seems made to sell "polarized" training idea. When you look at the details in the data no one trains like that, the final distribution is almost always pyramidal, not polarized