The only real comparision is power at different times so 1sec, 1 min, 5 min etc. W/kg do not tell anything about rider profile (take for example track sprinter or cyclocross racer) and as a value it is only useful if you ride in a very hilly environment with many long climbes.
And you can always use intervals.icu that have this comparision on some bigger samples of riders.
Bigger? Are you sure about that? I just checked and intervals.icu has 9017 male athletes, 3222 in their 30s and 1711 in my 35-39 age group. While I approve of the implied Dragonball Z meme from that first number, I’d be surprised if TR didn’t have more samples.
Not biger in comparision to TR but bigger than anecdotal samples from friends. Sorry for my lack of precision.
God bless TR’s sweet-spot work and long Z2 rides (which TR doesn’t emphasize much, but should). In July, I was right around the 1st percentile for all durations past 1 minute.
Still 99kg and 212W FTP. But while a little running/swimming reduced my bike time, six months of 1 or 2 “sweet-spot progression” workouts and one two-hour Zone 2 ride each week have gotten me well on the way to that 5th percentile:
I should hit the 5th percentile across all time intervals pretty easily next year, and getting up to 8th percentile or so looks achievable as a stretch goal. A bunch of you have helped teach me how to do this… many thanks. 
At 94kg and ~2w/kg I have a pretty messed up chart - but I am new to cycling (2 years - 99.9% of it on an indoor trainer) - At least I am slowly losing weight 
I did manage a couple of recent form sprints in Fletcher this week where I surpassed 200 RPM so the legs can move fast enough - they just don’t have the power or stamina (yet).
Keep in mind that the intervals.icu population is not a normal cycling population. By its nature, it attracts people who will be on then high end the bell curve.
Yes, both TR and IICU have userbases that clearly don’t represent the global cyclist population. That’s why I’m quite happy to be moving from the 1st percentile to roughly the 3rd, and to have (I believe) a reasonable hope of hitting the 8th or even 10th percentile with another year of training.
I’m thrilled to see these results PRECISELY because the userbase skews to the high end of the population’s bell curve, and I’ve been training for less than 2 years, and I’m still 50 pounds overweight, AND I have only 2-4 hours a week to spend on the bike. All things considered, I’m ecstatic. 
Unless you are accounting for age, I think that it is very much on the LOW end of what is possible.
ETA some data.
Here is maximal power for non-sprint trained kids and adults across a wide range of ages. On average, it peaks in young adulthood at about 18 Wpk, which would top the ICU chart.
Similarly, mean power for young untrained men during a 30 second Wingate test is usually about 10 Wpk, which would also top the chart.
Coggan famously calculated that the the average young person could achieve 4 Wpk for FTP, which would also almost top the ICU chart.
Coggan famously calculated that the the average young person could achieve 4 Wpk for FTP, which would also almost top the ICU chart.
https://forum.slowtwitch.com/forum/?post=2830698#p2830698
He was saying that even with training, most people will top out at 3.9W/kg due to genetics.
As for the paper you mention: given that we can safely assume that people on ICU have at least had some training, and your numbers are supposed to be averages, there is no way to reconcile both datasets, because this would mean that any average untrained person would top the ICU chart. And that clearly makes little sense.
This brings me hope because after reaching 4wkg lat October I’ve declined to 3.7wkg and am having huge issue at Vo2max power which of course curbed my 20 min tests. Yet, my 1 min power went up from 377 to 544w (7.6wkg) - a huge jump and my 5 min power went up 10w to 341w (4.6wkg)
I’ve been Z2, then Z3 since October with 1 x weekly low cadence 1-6 min efforts (50rpm @3.6wkg) with the exception of some failed VO2max efforts between races although am much better a climbing lower hills and have greater endurance but my top end has gone.
My chart looks top 20% ish, but I never feel or race as good at that indicates.
From the graph it looks more like 1/3rd, but anyway, the “problem” here is of course that if you go race you’re going to encounter more people from the top of that graph than the bottom!
I wonder how many people really top out at 4 genetically vs those who see diminishing returns and give up when they only gain 3-5w per year
I wonder how many people really top out at 4 genetically vs those who see diminishing returns and give up when they only gain 3-5w per year
He wasn’t looking at the distribution of W/kg numbers to make the prediction, so what you mention is largely not relevant to that estimate.
(Given the assumptions, I wouldn’t read too much into the exact numbers, I think his point was more that you definitely can’t assume everyone will be able to reach 4W/kg or whatever benchmark you have in mind.)
So where did the estimate come from?
So can I assume that I’m genetically gifted if I had 4 w/kg before I even started structured training? When I started structured a month ago I had a 260 ftp at 65kg. Up to that point I was riding unstructured for a year, 2-3 times per week, but some weeks with time completely off the bike. Mostly 40km rides, with the occasional 80ish ride mixed in.
32y/o, male, no real backround in endurance sports, only bodyweight excercise 3 times a week.
Really looking forward to my next ftp test in about 10 days!
My apologies, I was referring to my 2/3hr which are not show there as I haven’t done anything that long with a PM for 42 days. I’m an MTB Marathon racers so thats my target watts - 84 day graph highlights that better.
Not sure if I’m a slow-twitcher or a middle of the roader who has turned himself into a TT’er over the last 40 years.
I have not done a 20min test this season, or anything hard under 20min. Bests last year were 420w for 5min, 390w for 10min, 350w for an hour (that PM read a little high – might knock about 10w off each of those).
I changed my Strava last fall. I no longer have the data from '12-'20.
Definitely gifted, provided those numbers are accurate and not based off of virtual power or a bad power meter. Well done 
Yea I was worried about my smart trainer being off at first, but I recently also got a power meter for my bike and the reported numbers match
So where did the estimate come from?
Read the original discussion. It’s linked in my post.
