Race above TR FTP

Big question, im very confused….. hope to find the answer here…

I read that TR ftp detection is great. Im struggling with this for a long time and now I have some proof to show.

This weekend raced Dutch Championships 6 hours race (MTB) finished 3rd and raced with 206 NP. Last 3 hours stept of the gas a bit as I doubled #4 and # 1-2 were a size too large for me :grin:

But…. TR detected my FTP at 219 (Im 60+ yrs) but I raced for 6h at 207 NP and

1 hr - 230 / 2hr 225 / 3 hr 219 (all NP)

So is TR ftp detection correct? Dont think… hope to find the answer here..

Thx and KR Jan

2 Likes

What are the average power values?

1 Like

NP is generally not a good measure for this because it is a rolling 30s average, and short accelarations (as are common in MTB etc) can overly inflate it. Especially if you have high anaerobic capacity. Usually it’s better to compare to average power instead.

2 Likes

thx for your help… good point..

all power data:

6 hrs NP 206 avg 196 - 1 hr 231/221 - 2 hrs 225/216 - 3hrs219/210

Hope this helps…

Set FTP off a steady 40–60 min average power, not NP from a long MTB race. MTB files are spiky, lots of coasting and surges. That makes NP read high vs what you can actually hold steady. Whole-race avg power reads low. Neither is perfect for FTP unless the effort is steady.

From a race file you could grab the best 40–60 min continuous section with few zeros. Use avg power for FTP, use NP only as a sanity check. Or take your best 40–60 min avg power, compare to NP 60 for the same window. Set FTP near the lower number (or NP 60 minus 2–5%).

NP is great for TSS/IF and comparing how costly a ride was. Not great for setting FTP unless it’s a pretty steady 40-60-min segment and even then you still adjust down a bit.

2 Likes

Absolutely true…. Fully understand… my point is that T0 gives me an AI ftp of 219 en I can ride 219 avg in a 6hr race.. this is in my opinion not correct…

Do you have a limited number of rides with power that AI FTP uses as a base for assessment? If so, that might be the issue. I had similar experience with AI FTP just recently. I have returned to TR after 12 months or so with no structured training, and few rides with power. About four weeks prior to starting again with AI FTP I had my last rides with power. 6 hours at avg 190, and 6 hours at 180 avg (or something like that). AI FTP gave me an FTP of 153 :slight_smile: I manually upped it to 230 for my first wo, and then after my first wo I adjusted it to 245 based on my RPE. I have a fairly good sense of my legs after several years on TR.

Long story short: If you have limited rides with power, that might be the case. If not, I would reach out to support.

1 Like

Thx ohjelm for your help… to answer. I think started using a power meter in the early 2000 (SRM) and never not used a power meter since. Im using TR now for 3-4 yrs I think and was of for 3 months beginning this year, but did not stop training (with pm). When coming back to TR, it analysed all my pm data and gave me a 255 ftp. which I think is good. Then, I dropped to 219 and did not get of that number. Now I raced a 6 hr race and did not raced at power as I believe 219 is not correct, so rode as good as I could, which resulted in the shown numbers, which confirmed my conclusion that 219 is not correct. I don’t blame TR, I dont blame the AI model, I just want to know what did I do wrong, or where did this go wrong as TR is very fond of the accuracy of their AI model.. hope this info helps…. KR.

I assume you mean TR (TrainerRoad) not TP (training peaks)?

Do your workouts feel productive?

Have you emailed support to take a look?

Are you using the same power meter for TrainerRoad workouts that you used for the race?

I agree with others that NP isn’t the best way to test FTP but it can be a good sanity check if you understand the limitations. Based on these NP numbers you give it might be a good idea to reevaluate your FTP. You’re obviously already doing that but I think you’re too hung up on this race and what the AI did or did not do right or wrong.

The only question that really matters is “When you go out and ride at this FTP number, does it feel like FTP?”. Anything other than that is pure guess work.

Go outside (or inside I guess, it doesn’t really matter), and ride 10-15 minute blocks around what you think your FTP is.

  • Ride 5-10W below, Does this feel easy? Okay, bump it up a bit
  • Ride at what you think it is, Does this feel like you’re working but still suspiciously easy? Okay, getting closer but try bumping it up again
  • Ride at 5-10W over, Does this feel like your RPE is quickly outpacing the power you’re putting in? Okay you’re likely not over FTP and that second step is pretty close.
  • Continue/repeat until you feel over FTP

We can sit here and speculate all day about NP calculations, the AI model, your FTP, and your anaerobic capacity but until you just go out and test it it’s all going to be guess work.

Yes… stupid me.. dont know why I keep writing TP instead of TR … changed it…. Thanks for noticing..

Thx… good point…. Yes I do use the same power meter…

Thx and I fully agree a 1000% on that.

If you want to know your FTP, go out and test…. Simple is that….BUT…. TR advertises that the AI model is so good that ftp test are not needed anymore and I find (already for some time) that for me the number is wrong and I want to know why.

To answer your question… when I ride and use 219 as ftp, training as waaaay too easy. My thought was that after this race (I did many more in the past with high numbers then 219) it would suggest a higher ftp, but it did not.

So for me it is not clear why I ride these numbers and why is TR giving me this ftp.

Note: I do my training 240-250 ftp and then I still stay on the top side of the scale…

Contact support.

1 Like

Definately contact support but I suggest you go to them with the problem that is the workouts are too easy rather than your FTP is lower than your MTB race.

TRs AI FTP doesnt use max efforts to calcualte the number and is solely meant as a number to set the correct intesity of your workouts.

Having a lower number is not and issue but the workouts being too easy is an issue :+1:

Based on that you weee only slightly above the AI FTP for first hour and then your power output dropped. What are the values for the 4,5,6 hours and are you showing your average in each hour or the average up to thst point?

How flat was the course? Your Ave. FTP to NP ratios feel more like solo gravel race values than a MTB race. So something feels off

1 Like

+1 on the prior comment about those numbers looking “off”. For pretty much anything besides a steady-state time trial, you would expect to see a much bigger difference between AP and NP. Is there a chance that you have your computer set to ignore zero power numbers? When you look at the race/ride in the Trainerroad app, are you seeing the same values for AP and NP that you are showing here?

Hi AlphaDogCycling…

Thx… here in NL all is flat :laughing:… we live below sea level and the race was also flat and moslty single track. It was an 8km loop with some very short steap up and downs…

Hope this helps…