Polarized training/flawed studies?

With apologies for another PT thread…

Having had my ear bent by a couple of ex racers at the LBS over the last couple of visits (specifically, that TR-style training is inferior to polarized, with arguments of varying merit), I did a bit more research, and came across this article: https://philwilks.com/polarized-training-and-my-experience-1-year-on-b334eb99e440

I was particularly interested in this statement made by Wilks:

The flaw in the scientific studies

OK, it’s not a flaw as such, but it does limit how applicable the studies are to the majority of amateur riders.

I noticed something subtle about all the studies Dr Stephen Seiler referenced. In all cases they kept the training stress constant when they switched an athlete from whatever training they were doing onto the polarized model.

In Training Peaks language that means if you were doing 500 TSS per week then you would continue doing 500 TSS per week but mostly in zone 2. Because training stress is proportional to the square of normalised power (double the power = four times the training stress) this means the time you spend riding has to go up significantly to compensate.

My theory is that this is where most of the athletes’ gains came from. They were simply putting in more time, albeit at a much lower intensity.

Can anyone with a bit more knowledge of this comment on the author’s identification of this ‘flaw’?


The ‘flaw’ in Seiler’s polarized prescription is that retrospective studies show pro cyclists actually train pyramidal and not polarized.

However both pyramidal and polarized distributions have most time spent in low aerobic zones. That should be your key takeaway.

My training has been pyramidal for a year now, on roughly 8 hours/week average, and it’s working. So a lot of z2 with some tempo and sweet spot thrown in, and then a sprinkling of sprints and vo2 intervals. The biggest surprise was increase in short power performance - anything up to 5 minutes.


This was also my experience, and surprise, this past summer after a 2-month POL block. But it makes sense. And as KM says, a lot of people produce 1min power PRs after doing a lot of aerobic work.

Volume is king, after all.


While you are correct that they must increase time, the thing you are likely overlooking is that past a certain point the only way to add more time is at low intensity.

Think of it this way - you can handle X hours of high intensity training in a given week. This number is relatively low, 3-4 hours. If you want to add time you can’t just add more high intensity work. Thus, it isn’t really possible to compare a 20 hour training week in a Polarized or Pyramidal approach to a 20 hour training week of HIIT because no one can handle that much intensity.

You could get that same TSS with 3 hours of HIIT and 7 hours of sweet spot, or 3 hours of HIIT and 17 hours of zone 1. But you couldn’t do 3 hours of HIIT and 17 hours of sweet spot.

For the pendants - numbers are just for an example - no need to nitpick on whether 3+7 is possible, etc. etc.


Yup. And why, among other things, amateurs shouldn’t look to replicate pro-type training.


Just to clarify, I was quoting Phil Wilks (article linked); these aren’t my conclusions. I was just seeing what other people thought of them.

1 Like

Could you put the part that is a quote into a block quote or something? My response stands either way, but if you put it into a quoted section it’ll be clearer where the quote begins and ends

The flip side of increasing hours is ‘how low you can go’ and at some level, say 3-4 hours/week you have intensity in every workout.

It’s the bit in quotation marks…

but will try and edit

1 Like

It is not as much a flaw as it is a premise. I think of some of the logical conclusions you can draw are:

  1. At some point adding more intensity work is needless or even detrimental
  2. The amount of intensity work being beneficial is likely lower than most people think (probably 2-3 sessions per week)
  3. Adding more low intensity can help you increase training stress
  4. There might be interactive effects between high volume / low intensity and low volume / high intensity work

I think the 80/20 rule in polarized training according to number of sessions is too restrictive. But I also believe that Seiler is slowly changing his tone as to whether high(er) intensity work has to be zone 3 (so it could in reality be more pyramidal). Data from pros seems to confirm this.

Also, remember, that the article you are linking to is fairly old.

At one point, you just have to add more volume instead of intensity work in order to improve.

Yes, and @RecoveryRide I think this is a far better article on the subject:


The strange thing is that people try and make this into a religious war. Seiler is just not that dogmatic about it and he’s an exercise physiologist, not a coach. I’d recommend listening to 5 or 6 hours of podcasts (Fast Talk, That Triathlon Show, etc) with Seiler. Look at the big picture of what he is saying to amateur athletes.

His main message is make the easier days easier and the hard days harder.

Three years ago I tried a Seiler style base block. I rode 10-12 hours per week at an excruciatingly slow 125bpm (70% HRmax). My only intensity was a Saturday group ride. I didn’t even do vo2max intervals. By week 7-8 I was breaking all my benchmark PRs on Strava week after week. (My two benchmarks were an 8 minute climb and a 20 minute climb.) My FTP was up by 20 points.

I saw HR at every power across the board drop. My group ride started getting way easier because I was now running 10bpm lower at every intensity. So I proved to myself that very low intensity really works. I was not fatigued day to day.

Now I tried this same base block the following year and didn’t get the same gains. I was already adapted and carried that fitness through the year. And I wasn’t willing to push a base block to 15+ hours per week.

This year, I’ve had less time - 6-7 hours per week and I’ve been house bound due to covid and weather. I’ve done 6-7 hours of mostly Z2 on the trainer inspired by ISM. I’ve been mixing it with high Z2 and then low Z2 when I was fatigued. I sometimes take one day off per week but not every week. I’ve had zero time for a 3+ hour ride. I’ve totally maintained my fitness on less hours. Now I’m mixing in 2 days per week of SS/threshold and I quickly got a 10-15 watt bump in FTP.

The big picture message from Seiler is that low intensity training is the foundation of aerobic fitness. If you go out and thrash yourself in the tempo/SS zone that feels good every day you’ll be tired all the time and plateau.

Also remember that Seiler studied skiiers … elite skiiers with really high VO2maxes. They are a different animal.


I was also going to say that if you race or do group rides you’ll get a lot of time at tempo/SS. Even if you train polarized you’ll probably end up with a pyramidal distribution when you include the events.

1 Like


So go listen to coaches… after all Seiler looked at history, what coached elite athletes had actually done. My position is that you are better off taking your lead from a coach that has seen what works and what doesn’t work. Coaches have their athletes doing a lot of base aerobic work and spicing it with mid and high intensity work. The art of coaching is recognizing patterns and applying appropriate progressions for a given athlete and time constraints.


If I were a pro athlete with only time to ride that I would do polarized training. But, I am an amateur and can only ride for a limited time. So, I believe that sweet spot training is better for me.

And this.

And…it’s been pointed out several times across this forum that “polarized training” is the result of high volume training, not the prescription. If you train only 5hrs/week you can do all the SS you want. If you train 25hrs/wk, guess what, you’re going to get a lot of low intensity hours – because you can’t do 25hrs of SS+ per week!

Put another way, the athletes who observed as doing POL training were probably actually doing HIIT and then recovering by doing low intensity work. POL isn’t flawed, the way we popularly think about POL is flawed.


Ahem. Aaaaaaackshually, it’s spelled “pedant”.