Anyone have an opinion on alternating between hard and easy weeks?

Due to weather and injury, it turns out I've been alternating hard and easy for a few weeks now. I don't know if it's helpful or not, but my "Efficiency Factor" in trainingpeaks has gone up.

I vaguely remember scanning a study that tested this, and I think the alternating group had better results.

How hard is hard, how easy is easy? Are you distinguishing in terms of volume, intensity, or some intensity-weighted volume? And how different are hard and easy according to your metrics?

I think it's easy to distinguish hard from easy, but hard to know what the training consequences are.

Well, for simplicity, let's say that "hard" is a Training Peaks 7-day Fitness (CTL) Ramp Rate over 10, and "easy" is under 5.

My 7-day ramp rate is currently 12, but it has been up and down for a while now.

I'm sorry, I don't use TP. Does a ramp rate of 12 over a week mean, e.g., that your CTL was X one week ago and it's X+7 today? Does a ramp rate of 109 over 365 days mean that you were doing nothing a year ago and today your CTL is 109? (That is, is RR calculated as the difference between your CTL today and your CTL X days in the past, so to get the daily average rate take that difference and adjust it for the elapsed time?)

This sort of reminds me of interval training on a macro scale: are you better off doing intervals or steady ("sweetspot"-ish) training?

I think a ramp rate of 12 means CTL was X a week ago, and X+12 yesterday. CTL:
Saturday, 5/6: 100
Friday, 5/12: 112
Ramp Rate: 12

My CTL on 5/14/2022 was 2, and today it is 111. And as you suspected, it appears that RR is simply yesterday's CTL minus the CTL from X days ago.

I found the paper!

While this study is not precisely "alternating hard and easy weeks", this 4-week program preloaded the first week with 5 days of high intensity (block periodization), followed by 3 weeks with just one high intensity day. And they report power increases over the traditional "two intense days per week" training program:

Block periodization of high-intensity aerobic intervals provides superior training effects in trained cyclists

...The present study suggests that block periodization of training provides superior adaptations to traditional organization during a 4-week endurance training period, despite similar training volume and intensity.

Ronnestad et al, 2014

I feel bad about myself.

My CTL has increased 12 points.

In 5 weeks.

After more reading, the increase in VO2 max isn't as compelling as power increase at 2 mmol lactate, which looks really good.

https://www.tandfonline.com/doi/full.../OAJSM.S180408

Whoops, sorry, I had a typo in my first sentence: I meant, of course, that I was wondering if RR over X days was simply CTL(today) - CTL(X days ago), i.e., it's not a rate at all, just a difference.

Since CTL is a weighted mean of past CTLs (weighted by a constant decay), I'm thinking that both an increase of 12 and an increase of 5 over the course of 7 days is a lot. So rather than "hard" and "easy" I'm thinking you're doing "hard" and "even harder."

How often do you update your FTP? What's your average TSS per ride during a +12 CTL week vs during a +5 CTL week? (Or, what were the TSSes for rides during a +12 CTL week?)

I don't have access to that paper. Training load during the "on" and "off" weeks was "similar." If you do have access to the paper, is it possible to see what the CTL accumulation might have been like during the on and off weeks?

Mine is up about 10 in the last month and 20 since mid-January and that’s a nice, sustainable, build for me.

Seems to be OA. Here

Doesn't show up as Open Access to me. What did they use to measure training load, and how did they keep it "similar"?

I use Skiba LTS rather than Coggan CTL, but my LTS has increased by 34 since Jan 1, so on average a bit more than 2 points per week. I'm pretty impressed that Terry has been seeing gains of +12 per week for his "hard" weeks and +5 for his easy ones. (To be fair, I also use CP rather than FTP as the base, and my CP is probably a tad old, so those are other ways that my delta LTS isn't exactly comparable to Terry's delta CTL).

Most training plans follow a 3 week hard, 1 week recovery cycle. For older riders plans are often modified to 2 weeks hard, 1 week recovery. At 55 I tend to find the latter 2/1 cycle more sustainable.

Yeah, you're right; I forgot I was on work VPN. I don't have the time or inclination to read the paper right now, but if you (or anyone) DMs me an email address, I'll send the .pdf.

terrymorse Define intensity. Do you consider an FTP high intensity? And above FTP, there is VO2, anaerobic and neuromuscular. One can do a 20-30 second maximum effort which is technically not sprinting and it will really cook ones legs. Rest 10 minutes and do another.

The next question is did the participants do 5 sessions of "HIIT" efforts and what was the level and number of efforts and what were they doing between efforts - maximum or VO2, resting or riding z2?

It seems like a lot of variables to standardize, control and correlate effectively.

I am not being negative. Just trying to understand how one sequence is better than another and against what goal.

The intervals in the linked Ronnestad study were either 6x5 or 5x6 at the maximum the cyclists could output with 3 minutes rest between, IIRC. So, around VO2 max power.

I looked because I was curious if the intervals would be the traditional ones as above or would be the shorter and even more brutal Ronnestad intervals.

Here's a link to the full pdf: Block periodization of high-intensity aerobic intervals provides superior training effects in trained cyclists

There's a graph showing the training for the two groups:



The zones are based on heart rate: "(1) 60–82%; (2) 83–87%; and (3) 88–100% of maximal HR".

I update FTP sporadically, and I don't do a traditional FTP test. I get notices from intervals.icu that it detected a new FTP, so I then go in and change it. It's probably lower than it should be.

My recent, haphazard, "block periodized", weekly TSS:

Week 1: 665
Week 2: 114 (illness)
Week 3: 1012
Week 4: 472
Week 5: 1276
Week 6: 645
Week 8: 956

Long ago on the old Wattage List, when Andy Coggan first proposed NP, IF, and TSS, we discussed rough categories of rides by their TSS. Here's the graphic from back then. This was his description of how we might label single rides.

How does one do a 400 TSS 4 hour ride with an IF of 1.0, yikes. That must be pro level.

The worst I can recall doing was a 2:45 Zwift race, I did just a hair under 0.90 and with the 20 minute warmup, it was a 90%. I was absolutely wrecked. Last race I ever won although zwift racing doesn't seem real.

I put that chart together from things Andy had said so he could react to it. I think the general response was that although the math was consistent with what he had said, he imagined that a more common pattern would be that high TSS rides would be composed from lower IF and longer duration, so the chart should have been extended off to the right. But he liked the term "epic."

Given the definition of IF = NP/FTP, wouldn't it imply that your FTP was artificially low? I tend to record an higher IF on rides that involve repeated hard efforts with periods of recovery (e.g. Zwift races) where my NP is often higher than my steady state FTP. But 4 hours at at an NP equal to my FTP is never going to happen. I very much doubt pros can do that either, although they would get a lot closer - maybe 0.9 ish.

Just looking at some pro data, Paris-Roubaix was won last year with an NP = 341W over nearly 6 hours. I would expect Van Baarle's FTP is well north of 400W, so that would be at an IF = 0.85 at best. With a shorter 4 hour effort, maybe he could hit 0.9?

Dr. Chung explained it.

I don't have the paper at hand but a well known exercise physiologist showed that W' could be recycled faster than expended and that this asymmetry could result in higher IF than might be predicted by FTP alone, ostensibly (my guess) due the the nature of the VO2 slow component in the moderate, heavy, and severe domain. I was thinking Pros might be able to go briefly into the severe domain before the slow component rises much and then recover quickly in the moderate domain. Let's say a 2 minute or 120 second hard effort at 100 watts into the severe domain (beyond FTP range)or 12Kj. The idea is this 12kJ could be restored quicker than 120 seconds. Rinse/repeat. Since Pros probably have closer to 30kj, this effect would be greater than what I wrote. ( IF is based on Normalized Power). My interest in the topic started when I tried to explain obvious differences on a very reclined recumbent compared to an upright. I wish I could remember the guy, he posted sometimes in wattage and on slowtwitch. Anyway, I am wrong

Pros are known to be able to go really hard above threshold and recover, I thought this might explain the chart.