Amateurs don’t train enough for zone 2 to work
Part of Roadman Podcast discussion with Lidl-Trek head coach. Interesting discussion. I haven’t listened to the full podcast straight through but lots of nuance to the discussion I’m sure
I enjoyed that podcast & what I would say is that among coaches training elite athletes, zone 3 training never left, but it may have existed only as a coincidental part of workouts that targeted other zones. Either way, I like the idea that a given intensity might fall into a ‘zone’ relative to a given athlete’s physical ability…but the absolute thermodynamic cost of that intensity is experienced the same for all athletes. I rode behind Chad Haga during one of his LT1 rides (‘light riding but light riding with purpose’) & the KJ required for him to do that workout were ENORMOUS. Yikes.
Also, I would say that for mere mortals such as myself, the exact split of anything over zone2 doesn’t matter too much most of the time. If you want to do some serious training, start by accumulating 12 hours of zone 1 & 2. Make sure Zone 2 work doesn’t exceed 2 hours.
After accumulating those hours of work, you have two one-hour chits that can be used for:
1.) More zone 1 or 2
2.) Sweetspot/Threshold work
3.) VO2max work
You can go a long way just doing that. That training approach is adapted from the “Training Session Models In Endurance Sports” paper. That paper is a good read for no other reason than the coaches’ comments.
Zone 3 training never left for time crunched athletes.
I train on 6 hours a week, with the majority at SST (2x30s and 3x30s).
OG TR plans got criticized heavily for having too much intensity, but the basis for that criticism was always overestimated FTP, IMHO. Those plans, when done to completion, were extremely effective.
I’d still argue they had too much intensity, but that depends on your own goals. Those plans are good for rapid fitness gains, but maybe not the best for long term growth.
Even now at max weekly time ytd of 7 hours and 20 minutes this is my hr based tid. Zones here are based off running hr, so may skew more towards zone 1. This is 2 hard, 2 moderate, 3 very easy.
OG TR Plans FTW - I’m still following them.
Zone 3 never left for time crunched athletes. Depends how time crunched you are. Time crunched for me is less than 6 hours a week or TR low volumes plans 3,5 hours. Best bang for your bucks. I tried polarised approach for time crunched…like in everybody else it did not work. Maybe if you are fatigued and change to z2 you will recover and feel stronger for some time, but this will settle fast and you will need intensity
Is there a typo in the end here? Z2 work not to exceed 2 hours but you’re suggesting accumulating 12 hours in z1/2? So you’re saying build a week of 10 hours of z1, 2 hours at z2, then tack on intensity?
I’ve been focusing on just ‘hours’ lately as its an easier metric than specific time in zones. What i’ve found is if I want to increase volume, it needs to be easier riding. I’ve done this by putting a soft cap on HR when doing my ‘add volume’ rides as to not get too punchy or put a ton of stress on the system.
Did z3 training ever leave? It’s fairly common in most people’s training at some point during a season.
It’s actually not common any more. Z3 got labeled junk miles and everyone focused on SS, and then the whole Polarized fad happened. Even most TR training seems to go high Z2 and then jump to SS (or at least mine did). It’s been out of vogue for quite a while now, but is definitely creeping back in. For me, where Z3 really adds value is when I get on a long hill and try to keep cadence low, but not drift into steady Z4.
Yes, you got it. If you look at the actual, real training distributions for elite athletes in that paper they are spending the vast majority of their time in Z1. And then really just a small bit of time in Z2. Often more time in Z3 than in Z2.
And then you can just let TR pick the exact workout. Or a lot of times TR will pick and workout but tell you, ‘no, that’s dumb. Do this instead’. It’s really slick. You can start with a big, high level structure and then TR AI is like putting those inflatable bumpers in the gutter at the bowling alley to keep you in the lane.
Were they using a 3 zone or 7 zone model in that paper? I’d be surprised to see that much time in Z1 in a 7 zone model. Not saying it’s wrong, just not typical from what I’ve seen. If it’s a 3 zone model in the paper, it would make more sense to me and the Z1 time would roughly map to 2 time in a 7 zone model.
Not 100% certain this is the right thread, but it felt semi related. I’ve been messing around with Claude, mostly for analysis. I fed it my entire cycling training history, literally every ride I’ve done in the last decade, when I started cycling. I then had it run various analyses to figure out what worked and what didn’t. I was slightly surprised by the results actually.
Quick background. Started cycling in 2017 just for fun. Started with TR in 2018 through 2021ish. I had a year or so of stagnating results for most of 2021, and wanted to try something different, so I stepped away from TR in 2021. When I was doing my own thing, it was basically a repeat of a SweetSpot block, VO2 max block, Threshold block over winter, then lots of riding in the summer months along with racing. 2023 I had a huge breakthrough and gained 10% on my FTP, and continued to get stronger. Peak FTP was winter 2025, funnily enough, right before I tried the new TR AI system which I ended after 4 months when I had dropped back down from my all time FTP. With the TR AI I dropped almost 15W on my FTP but this isn’t about that.
So, I fed this into Claude and from my entire training history, Threshold was where I got stronger. SweetSpot to a lesser extent. Volume and Z2 had zero correlation to me getting stronger. I pushed back and did more analysis, but again, volume did almost nothing for me. I had it break down weekly hours into groups: <6, 6-8, 8-10, 10-12, 12-14, >14. Zero relation to breakthroughs, periods of stagnation, drops. Nothing. But, I did show a positive correlation to consistency, or % of days ridden. So I had it breakdown a rolling 4-week block and compared blocks where I had <50% days ridden, 50-75%, >75%. Not surprisingly the >75% won out.
So for me personally, it’s less about total hours on the bike, and way more about # of days on the bike. And that SweetSpot and Threshold work do way more for me than lots of Z2. My best blocks were when I did 2-3 intensity workouts extending out SweetSpot and Threshold, with some easy spins on other days. My total volume for those periods was only around 8-10 hours.
That’s interesting as it lines up with Frank Overton’s approach in general. I was going to ask when you mentioned threshold - was it in general at 100% ftp, more upper sweetspot (97%), or over ftp towards Z5?
So, funnily enough, I dove a little deeper to find the cutoff point, and Claude was counting a lot what I called SweetSpot as Threshold. I used a lot of the “More Sweet Spot” team workouts (IYKYK) and would do a ton of time at or around 90%. Claude counted this as Threshold. So after a little recalculation, my biggest gains come from SweetSpot. I would do months of nothing but these workouts, pushing out to 140-160 minutes in zone for a single workout (4x40 was my longest). It was boring, but simple. And it was easy to plan, just add time in zone each workout. Interesting to note, that most of these blocks I’d rarely break 10 hours in a week, usually just at the end. My average weekly total was around 6-8 hours. And yet this is when I would have the biggest gains.
When I did Threshold blocks, it was always 100% FTP.
That’s the best thing about sweet spot: most of the training stimulus, a lot less fatigue, which allows more consistency (greater number of sessions per week in the long run) when paired with low intensity training.
High end of tempo is the same thing, though a bit less stimulus per hour.
Excellent. Thanks. I pointed Claude to my Strava account and asked about training blocks with associated ftp increases. Claude stated what you found. Sweetspot/upper tempo/lower threshold emphasis produced the most for the effort. I have also found that it increases what is called today, “durability” on the bike.
Here’s the thing about biology though… stuff you did years ago is connected and has an impact on performance years later.
I was a college athlete… high mileage and double days were a staple of training. When I graduated I got a job. Focused on longer races like the marathon. My training went from doubles to singles but much more volume per individual session. I got some solid performance gains.
Was it I was older. Training harder in single sessions. Or was it the doubles previously gave the foundation to what was happening in that moment?
That’s what I love about biology, it doesn’t always fit neatly into data sets.
Yea, it would have been a lot easier to just link Strava but I don’t use Strava. I ended up having to download all my rides and manually upload them.