Race The Training Thread

[Tsuga]

Time is really sailing along. Already been back from the US for a couple of months. I did eat entirely too much cheese and bread, but I don't think that has slowed me down too much. I'm still "training" although thankfully during this time of the year that means no intervals and just riding how I feel like riding. Lately that has been endurance rides like the one below. Sometime in June I started losing all of my top end power due to neglect, and it has only eroded more. 400+ watt efforts crush me, but I have developed a lot of comfort cruising at 230-270.













This ride put me over the 500 hour mark for the year, so that's really the only goal I have left. Just trying to see how much time I can accumulate in the next couple of months. I'd like to hit at least 600. But other than that I have no event related goals for the rest of the year or even into 2024.

 

[baribari]

Now that the weather has finally cooled down, my power seems to have shot up quite a bit... That, or both of my power sources are inaccurate. Haha.
I did 264 watts for just under an hour up a climb (I thought it would be more, but altitude is a factor) the day before yesterday and apparently 274 watts for a bit over an hour in a Zwift race last night. Neither could be described as an all-out effort.

I wonder if I need to do a power check on my new Kickr Core, since my old one also read quite high until I did a factory spindown. This one is less than a year old, though.

That, or maybe TrainerRoad just isn't very good training compared to Zwift rides and races. Glad I got a new Apple TV 4K to play it again (my old Apple TV was not really up to the task).

 

[nateberkopec]

I found out last week that my indoor trainer reads 10% higher than my Assioma pedals, but only for ~200+ watt efforts (at zone 2 pace, the difference is like 2-3%). I looked at some watt calculators online and I think the Assioma is the much more realistic number.

Made me think my FTP was shooting up...turns out it wasn't!

I think power meters are generally a lot more inaccurate than we think.

 

[baribari]

I found out last week that my indoor trainer reads 10% higher than my Assioma pedals, but only for ~200+ watt efforts (at zone 2 pace, the difference is like 2-3%). I looked at some watt calculators online and I think the Assioma is the much more realistic number.

Made me think my FTP was shooting up...turns out it wasn't!

I think power meters are generally a lot more inaccurate than we think.

Last time, after doing a factory spindown (which is a hidden menu option), it basically lined up with my power meter, dragging me back down to earth...

This time, if it wasn't for seeing similarly high numbers on my power meter, I'd think it was reading high. It might still be.

I could put my road bike on the trainer and see how the numbers compare, but it would be too depressing to have a third trainer lying to me.

Zwift feels easier than TrainerRoad, but you can chalk that up to motivation levels being different. Going somewhere (virtually) versus going nowhere.

There is some physiological basis for power increasing in optimal temperatures (e.g. 20 C and below), especially if you've got an extra fluffy build like moi...

Why Cyclists Can Handle Heat Better than Runners

An analysis of power data from pro cyclists quantifies the effects of hot and cold air temperatures on performance

www.outsideonline.com

 

[Tsuga]

I think in the current era of equipment power meters are reliable. But when you get too much distance from what is "accurate" for you, it can get a little wild. If you're really curious you should look at https://www.dcrainmaker.com/ & read his power meter reviews. For example, I will never own a Shimano branded power meter.

I have used Stages, Powertap, Pioneer. Currently I am using a Quarq. When looking at my "all time" peak powers, I know not to trust anything below 1 minute from around 2016-2020, because I was on a stages. The accelerometer for cadence could not handle gravel, causing massive spikes in power. I have tried to go through my TrainingPeaks and remove any peaks when I was on my gravel bike. And even then, I only real like comparing my numbers within meters. So basically when I buy a new meter, my history "resets".

 

[baribari]

I think in the current era of equipment power meters are reliable. But when you get too much distance from what is "accurate" for you, it can get a little wild. If you're really curious you should look at https://www.dcrainmaker.com/ & read his power meter reviews. For example, I will never own a Shimano branded power meter.

I have used Stages, Powertap, Pioneer. Currently I am using a Quarq. When looking at my "all time" peak powers, I know not to trust anything below 1 minute from around 2016-2020, because I was on a stages. The accelerometer for cadence could not handle gravel, causing massive spikes in power. I have tried to go through my TrainingPeaks and remove any peaks when I was on my gravel bike. And even then, I only real like comparing my numbers within meters. So basically when I buy a new meter, my history "resets".

Dual sided purpose built power meters are the only thing I will ride. I think maybe trainers, which usually only estimate power instead of measuring it, tend to average high because of inertia. It's harder to zero out the power. Actual power on a real climb is quite erratic unless you are riding to power.

Even a good power meter probably needs to be checked at the factory every few years… but sending it back to Germany for who knows long is not feasible.

 

[Tsuga]

Dual sided purpose built power meters are the only thing I will ride. I think maybe trainers, which usually only estimate power instead of measuring it, tend to average high because of inertia. It's harder to zero out the power. Actual power on a real climb is quite erratic unless you are riding to power.

Even a good power meter probably needs to be checked at the factory every few years… but sending it back to Germany for who knows long is not feasible.

Dual sided reputable power meters have also come down so much in price in the past couple of years. I know it's still a big investment, but it's so worth it. I'm ok with non dual-sided depending on how it's measured. We have a Force AXS Crankarm meter on my wife's bike, but it measures in a different manner to a Stages.

Trainers are a little trickier. I have a Wahoo Kickr, which I think is comperable to my Quarq. It's at least close enough that I don't lose any sleep over it. It's nice that if I use it with a program like Zwift I can have the power come from my Quarq, but the program controls the trainer's resistance. So I can get the best of both worlds with a smart trainer. But I'm too cheap for Zwift at the moment... and I like ERG mode to set it and forget it anyway. And like you mentioned earlier, it's also so dependent on your environment. I can't ride hard indoors without nice fans, dehumidfier & cooler outside temps. I had a coach that would advocate for Indoor & Outdoor FTP... I could never be bothered to do an indoor test.

A whole other barrel of monkeys is what settings you are using to collect & read your data. By default Wahoo trainers have some kind of power smoothing enabled which makes it look like you can ride at an unrealsitic flat power. ERG mode is good, but not that good. Personally I also found the smoothing rounds up... so if I set the ERG to 190, with smoothing it will be exactly 190, without smoothing it will probably be around 187.

Follow that with the ability for some head units set to record power without the 0s averaged in... yikes. It's like when people only want to talk about their Normalized Power. NP Can be another helpful metric, but a lot of people use it soley to inflate their "stats". I know a guy that posts every ride with his NP & TSS... which is another personal metric. I don't know if he realizes his TSS is high because he's not often doing hard rides, where a different cyclist could do that same ride (and sometimes does...) and get a much lower number.

I guess it's a rambling way to say that the best power is your power. And as long as the measuring device you are using is accurate & consistent, that's all that really matters. And of course, don't forget to calibrate your meters regularly!

 

[baribari]

Dual sided reputable power meters have also come down so much in price in the past couple of years. I know it's still a big investment, but it's so worth it. I'm ok with non dual-sided depending on how it's measured. We have a Force AXS Crankarm meter on my wife's bike, but it measures in a different manner to a Stages.

Trainers are a little trickier. I have a Wahoo Kickr, which I think is comperable to my Quarq. It's at least close enough that I don't lose any sleep over it. It's nice that if I use it with a program like Zwift I can have the power come from my Quarq, but the program controls the trainer's resistance. So I can get the best of both worlds with a smart trainer. But I'm too cheap for Zwift at the moment... and I like ERG mode to set it and forget it anyway. And like you mentioned earlier, it's also so dependent on your environment. I can't ride hard indoors without nice fans, dehumidfier & cooler outside temps. I had a coach that would advocate for Indoor & Outdoor FTP... I could never be bothered to do an indoor test.

A whole other barrel of monkeys is what settings you are using to collect & read your data. By default Wahoo trainers have some kind of power smoothing enabled which makes it look like you can ride at an unrealsitic flat power. ERG mode is good, but not that good. Personally I also found the smoothing rounds up... so if I set the ERG to 190, with smoothing it will be exactly 190, without smoothing it will probably be around 187.

Follow that with the ability for some head units set to record power without the 0s averaged in... yikes. It's like when people only want to talk about their Normalized Power. NP Can be another helpful metric, but a lot of people use it soley to inflate their "stats". I know a guy that posts every ride with his NP & TSS... which is another personal metric. I don't know if he realizes his TSS is high because he's not often doing hard rides, where a different cyclist could do that same ride (and sometimes does...) and get a much lower number.

I guess it's a rambling way to say that the best power is your power. And as long as the measuring device you are using is accurate & consistent, that's all that really matters. And of course, don't forget to calibrate your meters regularly!

The problem with trainers is that they're usually pretty accurate until they're not!

I guess double checking with a power meter once in a while is the only way to be sure.

Two of the three trainers I have owned have had accuracy problems at some point (one after weeks, one after years).

I wish more would use strain gauges like Elite instead of just guessing.

The Minoura Kagura apparently is/was notoriously inaccurate (reading too high, making it a "happy meter"). Firmware updates may have fixed this.

Actually, if a trainer seems like its reading high, it probably is, because power meters should be a few percent higher due to drivetrain loss...

I wouldn't mind checking my power meter with some Assiomas at some point, either...

 

[OreoCookie]

Dual sided purpose built power meters are the only thing I will ride.

Going from a one-sided 4iiii to a Quarq made a big difference to me, too. My Quarq power meter is one of my favorite bike-related purchases ever: zero failures, zero dropouts in 2.5 years. Well, it "dropped out" once, because the battery was empty.

I think maybe trainers, which usually only estimate power instead of measuring it, tend to average high because of inertia. It's harder to zero out the power. Actual power on a real climb is quite erratic unless you are riding to power.

Agreed, trainers can measure power accurately, but some of them can become inaccurate over time (see your comment above!). My Suito was dead on with my Quarq initially, now both, slope and offset are markedly different. But since I use my Quarq power meter when I am on the trainer, too, this doesn't impact me.

I wish more would use strain gauges like Elite instead of just guessing.

Small nitpick: trainers don't "estimate" power, they measure power differently. Strain gauges infer force from a voltage and other parameters like temperature, and need cadence. Power (at the crank) is then equal to torque (force times leverage and cadence). Alternatively, power is change of energy over time, which is what most trainers use. Factory calibration and spin down tests tell the trainer the built-in friction at a given flywheel speed, so that the remaining change in energy at a given flywheel speed is due to rider input.

Importantly, both measure power indirectly and are based on different equations for power. I think the reason why the second method has a bad reputation with some is that you could buy flywheel speed sensors for dumb trainers and then get "virtual power". The crucial difference is that those speed sensors were not calibrated for one particular trainer and e. g. did not take temperature (of the surroundings and e. g. the fluid) into account.

You are right that some Elite trainers use optical strain gauges, and they have a great reputation for accuracy. I think some early spindle-based power meters were based off the same principle. They have some inherent advantages such as not being impacted by temperature (unlike strain gauges).

Even a good power meter probably needs to be checked at the factory every few years… but sending it back to Germany for who knows long is not feasible.

I would like that option.

 

[OreoCookie]

Dual sided reputable power meters have also come down so much in price in the past couple of years. I know it's still a big investment, but it's so worth it. I'm ok with non dual-sided depending on how it's measured. We have a Force AXS Crankarm meter on my wife's bike, but it measures in a different manner to a Stages.

You mean the latest second-gen Force? Do you have experience with the first-gen Force groupset as well? How do they compare?

Trainers are a little trickier. I have a Wahoo Kickr, which I think is comperable to my Quarq. It's at least close enough that I don't lose any sleep over it. It's nice that if I use it with a program like Zwift I can have the power come from my Quarq, but the program controls the trainer's resistance.

That's what I have been doing with TrainerRoad for years, works great.

I had a coach that would advocate for Indoor & Outdoor FTP... I could never be bothered to do an indoor test.

If you are used to indoor riding, outdoor riding will feel hard and vice versa. It is just a matter of getting used to it. I wouldn't want to do an FTP test outdoors. I don't think I could find a suitable stretch of road that is both long enough and safe enough.

 

[baribari]

@OreoCookie

I understand how strain gauge power meters work, I just think typical trainers measure power more indirectly (measured torque versus an assumed resistance curve). The fact that there are two different kinds of spin down tests says a lot.
New trainers get a factory spin down, but new power meters actually get checked for accuracy at the factory!

 

[Tsuga]

You mean the latest second-gen Force? Do you have experience with the first-gen Force groupset as well? How do they compare?

That's what I have been doing with TrainerRoad for years, works great.

If you are used to indoor riding, outdoor riding will feel hard and vice versa. It is just a matter of getting used to it. I wouldn't want to do an FTP test outdoors. I don't think I could find a suitable stretch of road that is both long enough and safe enough.

I have ridden 3 SRAM AXS groups. Previous gen Red & Force, current gen Rival. Between the (now) old Force or Red, I'd prefer to have the new Rival. This is mostly down to the shape of Rival's hoods. Force was OK, but I prefer Rival's smaller shape. I typically wear a generic "medium" glove so my hands are not small, but also not big. Weight is probably the biggest factor between all of the groups. And I guess also the issue with SRAM Red's crank / power / chainring being integrated without the ability to just install new chainrings. That's another dealbreaker for me.

I'd imagine that the current gen Force is very nice, but I'm not sure how much nicer compared to the price jump from Rival.

My go to FTP test is outdoors and 1x 20 minutes. When I was first getting into cycling that was the "normal" test so it's what I learned and what I've stuck with over the years. Briefly I gave the 2 x 8 minute protocol a try, but I wasn't a fan. I guess I don't see it as any more dangerous than standard riding out on the road. My biggest issue is the +/-60 minutes it takes to ride to my preferred testing road. Ideally I could do a 20 minute warm up, test, then keep riding.

I have been curious about giving an indoor ramp test a try, so maybe I'll do that sometime this year.

 

[baribari]

I have ridden 3 SRAM AXS groups. Previous gen Red & Force, current gen Rival. Between the (now) old Force or Red, I'd prefer to have the new Rival. This is mostly down to the shape of Rival's hoods. Force was OK, but I prefer Rival's smaller shape. I typically wear a generic "medium" glove so my hands are not small, but also not big. Weight is probably the biggest factor between all of the groups. And I guess also the issue with SRAM Red's crank / power / chainring being integrated without the ability to just install new chainrings. That's another dealbreaker for me.

I'd imagine that the current gen Force is very nice, but I'm not sure how much nicer compared to the price jump from Rival.

My go to FTP test is outdoors and 1x 20 minutes. When I was first getting into cycling that was the "normal" test so it's what I learned and what I've stuck with over the years. Briefly I gave the 2 x 8 minute protocol a try, but I wasn't a fan. I guess I don't see it as any more dangerous than standard riding out on the road. My biggest issue is the +/-60 minutes it takes to ride to my preferred testing road. Ideally I could do a 20 minute warm up, test, then keep riding.

I have been curious about giving an indoor ramp test a try, so maybe I'll do that sometime this year.

Ramp tests are very practical, and will give you a good number to train by, but they're really tests of MAP and to some extent, anaerobic capacity.
I always make huge "FTP" gains after doing a really hard race.
I haven't done a proper indoor 20-minute test probably since around the time I started this thread. Haha.
For a 20-minute test to be accurate, you have to do an all-out five-minute effort first, so the whole ordeal is rather unpleasant, while a ramp test only really hurts for a handful of minutes.

 

[Tsuga]

Ramp tests are very practical, and will give you a good number to train by, but they're really tests of MAP and to some extent, anaerobic capacity.
I always make huge "FTP" gains after doing a really hard race.
I haven't done a proper indoor 20-minute test probably since around the time I started this thread. Haha.
For a 20-minute test to be accurate, you have to do an all-out five-minute effort first, so the whole ordeal is rather unpleasant, while a ramp test only really hurts for a handful of minutes.

That might be part of your disdain for the 20 minute test haha. My coach described it like this:

"Warm up for 20-30 min. 1 x 20 power/HR test. on a flat to rolling road, with no stop signs or interruptions. After you can ride up to 2 hrs total.

Pacing, start the first 5 min hard, but on the conservative side, focusing on your cadence and form, HR should not be anywhere near max. At 5 min, evaluate and go up or down with your effort. Hold that through 15 min. At 15-20 min, bring it up slowly. Maybe you can raise the effort with 5 to go, maybe with 3 to go. but finish at a 10/10. leave it all out there."

As I have a benchmark, I usually start with a 5 minute effort targetting my previous test's average. If that felt easy, I try to increase that a bit for the next 10 minutes. If that increase was too much I'll be fighting the effort in the final 5 minutes to keep it above the previous test's average. If it was "easy" I'll try to do another small bump in the final 5, basically as described above.

Typically (on a good day) I finish a 20 minute test with a tiny amount of gas left in the tank. Maybe in the last 1-2 minute I can muster a relatively big increase in power before hitting the wall. So maybe I go from 350w to 380w as a "sprint" finish.

Without a coach looking over our shoulders we probably are guilty of picking the format that we excel at. Which I don't think is a bad thing, it's nice to feel strong on the bike. If you look at my power profile my strongest efforts are 5-60 minutes long. I really enjoy long tempo / steady state efforts. So a 20 minute test is "fun". That's probably the biggest reason I have never done a ramp test.

 

[OreoCookie]

New trainers get a factory spin down, but new power meters actually get checked for accuracy at the factory!

The factory "spin down" is a calibration just like when the factory calibrates crank-based power meters at the factory. According to this GCN video (6:00 onwards) the calibration takes "up to 45 minutes" and you can tell from the rig that they put known loads on the trainer with an electric motor. I reckon crank-based power meters are calibrated in a similar fashion.

That is why trainers can and do achieve the same accuracy that good power meters do (typically +/- 1.5 % or thereabouts).

Anyway, don't want to derail the entire conversation.

 

[baribari]

That might be part of your disdain for the 20 minute test haha. My coach described it like this:



As I have a benchmark, I usually start with a 5 minute effort targetting my previous test's average. If that felt easy, I try to increase that a bit for the next 10 minutes. If that increase was too much I'll be fighting the effort in the final 5 minutes to keep it above the previous test's average. If it was "easy" I'll try to do another small bump in the final 5, basically as described above.

Typically (on a good day) I finish a 20 minute test with a tiny amount of gas left in the tank. Maybe in the last 1-2 minute I can muster a relatively big increase in power before hitting the wall. So maybe I go from 350w to 380w as a "sprint" finish.

Without a coach looking over our shoulders we probably are guilty of picking the format that we excel at. Which I don't think is a bad thing, it's nice to feel strong on the bike. If you look at my power profile my strongest efforts are 5-60 minutes long. I really enjoy long tempo / steady state efforts. So a 20 minute test is "fun". That's probably the biggest reason I have never done a ramp test.

You can't really apply the 90-95% formula to a 20-minute effort unless you do the five-minute effort first.
Otherwise, your anaerobic system is contributing too much, so your result will get skewed upward.
I don't mind a 20-minute time trial or hill climb at all. I don't necessarily think I could do 95% of that for (the better part of) an hour, though.

 

[baribari]

As expected, the trainer reads slightly high when it's warm. Especially at low power levels. Sprint power is bang on.

It was closer when the unit was cold. I think this just be due to my pain cave being colder than the Kickr was designed for (supposedly 15C). Elite is said to be good down to 5 C.

Edit: Tried it again for about seven minutes in 15+ C temperatures, and it was basically correct, if very slightly high on average. I suspect it will start to drift apart after 10 minutes.

 

[OreoCookie]

It was closer when the unit was cold. I think this just be due to my pain cave being colder than the Kickr was designed for (supposedly 15C). Elite is said to be good down to 5 C.

On my Elite Suito I can tell a slight drift in resistance mode as the trainer warms up. It isn't as drastic on my Elite Volano (a fluid direct drive dumb trainer), but you can tell. If I weren't using my Quarq power meter, I reckon that'd be one potential source of reading lower at lower temperatures (my thinking is: colder parts —> more friction at given flywheel rpm —> lower flywheel rpm at same power —> lower power readings). Trainers should have built-in temperature compensation, but I don't know how good that is.

One big downside of the calibration on trainers is that you should only do them after you have warmed up the trainer. In practice that means I only do spin downs after a workout. My Quarq can be calibrated any time. (In principle it features autozeroing, but I still zero it every time before I do a workout or a ride.)

 

[andywood]

I've done a few months of just riding. No structure. No intervals. Rarely entering tempo. Never seeing threshold. But lots of good solid riding.

Last month at SDA Otaki I was surprised to find myself leading the race, only to be foiled by a puncture.

On Friday, I did a steep 10 minute single-track climb. First time to go hard in a while.

Today I went to the trail. A few laps just to get used to the technical / mental side of things. I pushed on the 4th lap and was surprised to be just 5 seconds off my best time at the end of a full CX season 4 years ago...

I wasn't planning to do any CX racing this winter but that gave me the motivation to enter a race.

In addition to that I've started hitting the gym again.

The next target will be the SDA Otaki spring event (motivated to try and win it!) in May whilst preparing for a 4th event here in Kashiwazaki at the end of June.

Ride on!

Andy

Blog

ssskashiwazaki.jimdofree.com

 

[baribari]

On my Elite Suito I can tell a slight drift in resistance mode as the trainer warms up. It isn't as drastic on my Elite Volano (a fluid direct drive dumb trainer), but you can tell. If I weren't using my Quarq power meter, I reckon that'd be one potential source of reading lower at lower temperatures (my thinking is: colder parts —> more friction at given flywheel rpm —> lower flywheel rpm at same power —> lower power readings). Trainers should have built-in temperature compensation, but I don't know how good that is.

One big downside of the calibration on trainers is that you should only do them after you have warmed up the trainer. In practice that means I only do spin downs after a workout. My Quarq can be calibrated any time. (In principle it features autozeroing, but I still zero it every time before I do a workout or a ride.)

The spindown time varies greatly depending on the temperature of the trainer, taking much longer when it's hot.
More heat means the grease in the bearing is less viscous, and the rubber belt is more supple.
The wax in my chain also takes a while to heat up, too, so it's quite loud until it warm.

I wonder if that isn't a factor in why the perfect power (averaging well within the margin of error) starts to drift upwards slightly (3-4% high on average) after about 10 minutes. Probably not enough to notice.

Supposedly 10 minutes of warmup is best (I have seen it written that after a ride is the second best option), but I think it should be more like 20 minutes if you're riding in a cold room.

Wahoo confirmed that Kickrs are most accurate between 15 to 30 C. Curious whether they calibrate it to compensate for drivetrain loss...