Study on % 20-min TT power to predict 60-min TT

For those that like FTP pinned down to a specific duration (60-minutes)…. Or those bored and haven’t participated in a rousing FTP debate recently :rofl: I hereby present the following straight from X formerly known as Twitter.



  • professional: PO60min=PO20min x 0.96

  • well trained: PO60min=PO20min x 0.95

  • trained: PO60min=PO20min x 0.92

  • recreationally trained: PO60min=PO20min x 0.88

Somewhat reminiscent of FasCat’s 0.9 to 0.95 multiplier.

Coggan in the 2003 USAC paper didn’t say it’s exactly 0.95, he suggested a multiplier and not to be lulled into a false sense of precision IIRC.

Hey, it’s a slow Saturday and I’m in the car outside TJ Maxx waiting for my wife that was quickly going to go in and buy some pumpkin scented soap (40 minutes ago lol). Need some entertaining threads and I’m done looking at The Bee on X.


Thanks for giving us your own abstract. I’ve been having to read a lot of academic journal reports this week and a little tired of them at this one. Just happy to have the key info pulled for me!

1 Like

An update, an hour later…


I have a copy of the full text if anyone is interested, DM me.


By the abstract it’s not 100% clear if the 20min test was done with pre 5min deload….I assume yes since they introduce with „The Allen & Coggan protocol“. Personally I’ve had much better training using 92-93% so all in line for me.

:popcorn: :eyes:

EDIT: @The_Cog Lets see what the Dr. thinks

1 Like

It’s an interesting paper, but useless to fully understand or implement without defining how they determined “recreationally trained,” “trained,” and “well trained.” What were the criteria for those categories?

1 Like

I’d imagine discipline plays a significant role too- 0.95 is a touch low for me IME, but I’m no pro, just have that nonexistent triathlete power ‘curve.’ There’s so many factors that I think it would be difficult to come up with just a few ‘boxes’- could be a good application for analytics software, though it seems like most are leaning away from the 20-min protocol and testing In general.

Also, we just got TK Maxx in Australia. I don’t know why it’s so easy to spend ages in there, but I think it’s all the choices. I didn’t know there were that many scents, let alone ones you could put in a candle :exploding_head:

Abstract says they categorized them by maximum oxygen uptake. I’m not sure if that’s appropriate. But to use their proposed multiplier I would have to have a good idea of my VO2, which I can’t say that I do.

Correct indeed. I have never been a big fan of Hunter’s way of estimating FTP, as there are other, less time-demanding/disruptive ways of obtaining an equally (im)precise value.

Other than that, I would say that this study is a good example of #sportsscienceatitsfinest


They used previously defined categories as seen here: Guidelines to classify subject groups in sport-science research - PubMed

I would advocate for newly defined classifications: Defining Training and Performance Caliber: A Participant Classification Framework - PubMed


As far as I’m concerned, the biggest issue with the paper is they provide no demography/anthropometrics for the sample. Presumably they’re all men? How old are they? Height/weight/BMI? It’s not in the supplementary material, either. If I’d reviewed the paper, this would’ve been my first question–and I’m almost shocked the journal published it without this info. I can’t generalize the results or evaluate external validity without it.


More #sportsscienceatitsfinest

1 Like

Good choice of thread title, BTW. It aptly sums up what was actually measured, and illustrates why the study wasn’t worth the effort required to conduct it.

(Is it surprising that the better/more highly endurance trained you are, the flatter your power-duration relationship in this region tends to be? No. Does it tell us anything new about the physiological traits or adaptations that lead to this? No. Is such population information directly applicable/useful to an individual athlete? Not really. So, why bother doing the study in the first place?)


Excellent point. Could you expand a bit on these other ways?

Google “Coggan seven deadly sins”.


Couple of follow up Q’s:

  1. from inspection of a ride file

What specifically are you looking at?

  1. based on normalized power from a hard ~1 h race.

Sadly, this can be mich less accurate than 20m x .95

  1. from the power that you can routinely generate during long intervals done in training.

What interval duration or time in zone as a minimum?..…I’d say 3 x 10m minimum

I’m a bit concerned. It’s been 16 hours (the next day here PCT) and not one post has argued that FTP is not one hour power. You guys are slacking.

Not to mention no update from @WindWarrior… he very well could still be in the parking lot waiting for his wife. :rofl:


I think the only predictor is probably a race as I have a real hard time going all in without a number. That said it’s not a million miles away for me. This year I did 293W for a 21 min 10 mile TT on the road bike and 271W for a 60 min 25m TT on the same bike (much tougher course and of course I didn’t do the pre 20min effort as I wanted a good result) On the TT bike I did 270W recently for a 22min 10 (again slower course) and 254W for 25m so again pretty close to 95% of 20 ish mins…I will leave my FTP at 270W for the winter while I do a bit of running and do some maintenance cycling…I’m sure my FTP will drop a few watts but not enough to make TR workouts unmanageable.

Cuz there are people who need to pad their CV and publications willing to enable it?



We’re busy arguing over the pumpkin-scented soap.