# Estimated FTP calculation

I just joined Strava Premium and imported 40 Peloton rides. My power curve looks normal but the estimated FTP is crazy. My 20' average is around 200 W but the estimated FTP is given as 300 W, which looks more like 95% of the 20 s average and not the 20' average. What could cause this?

### Opmerkingen

14 opmerkingen
• Same here, my estimated FTP is shown as 300W, but I only sustained 300W for 1 minute... A more realistic FTP would be around my 45 min power curve reading (~180).

I don't think this feature works.

• I wondered this too. Maybe Strava's Fatigue Score on day when best efforts take place is taken into account. Or maybe repeated efforts in same ride affect it, eg Average power for long ride 200W but could include 4 or 5 efforts of say 280W for 20 mins or so, or towards end of ride. thus giving you FTP of 300W as your FTP should mean falling off bike after an hour with nothing left in the tank. As you can sustain repeated efforts throughout long ride maybe Strava somehow uses this in some complicated equation to provide FTP estimate.

Could be miles out but that's my thoughts!

• I've heard back from support, there seems to be something wrong with the data imported from Peloton but they are still looking into it. 300W is not what Strava intends to calculate as my FTP.

Richard - that's pretty generous thinking haha

• Mine seems too low though.

I did a 20min effort of 214W in the later half of a 130KM ride, I didn't feel easy but also didn't feel this effort painful. Also earlier this month I did a 2hr ride with weighed average power of 183. But my estimated FTP is only 190W.

• You can't calculate 20 minutes of effort on the back of ride that long.  For FTP calculation, be fresh, properly warmed up and then do your best for 20 minutes.

• My estimated FTP calculation seems to be better now. BUT the power curve itself is broken.

By definition the curve should always go down, mine goes up in the middle.

For example, it says my max sustained power for 17 min is 220 watts, but then it says my max sustained power for 14 min is 217 watts. ???

It kind of defeats the purpose of collecting all this workout data if you can't trust the analysis.

• @Daniel what I tried to say is that as my 20min effort after a long ride still can hit 214W, my FTP should at least be better than 214*95%=203W. However the Strava estimation is 190W. which is even lower.

@Patrick, yes I also see this, but usually in single ride power curve. My overall power curve is fine.

• @Patrick R.

"By definition the curve should always go down, mine goes up in the middle."
This is not strictly true, logically it may seem that way, but think of it in this way:

- If you did a 4 minute interval @ 300w, then 4 minutes of easy spinning @ 100w. This would give you a best efforts 8 minute average of 200w for that ride. (300*4+100*4)/8=200

If however, in the same ride you did another 4 minutes at 300w immediately after. This would give you 300x4, 100x4, 300x4. (300*4+100*4+300*4)/12 = 233.33) Which is an average of 233w over 12 minutes, however the maximum 8 minute average across that period is still only 200w, because no matter what continuous 8-min range you choose to average over, the maximum you can get is 4 minutes at 300w with 4 minutes at 100w. Though anything above 8 minutes will once again have a slightly higher average.

Over time, the power curve will converge, but individual rides can have longer intervals of a higher average power than shorter ones for that exact reason.

• @Michael P - I see what you are saying.

I thought the power curve was plotting "the maximal power you have sustained for a given time interval" which is why I said it must have a strictly non-positive slope. If it were defined like this, then in your example, the max sustained power for 12 min would be 100w (bc 100w is the highest power you sustained for the whole 12 min window), and then the max for 8 min would also be 100w for the same reason.

The support thread says it's plotting the "best average power", which is what you are saying Michael, and I agree in that case the curve could go up.

• @Patrick R. Ahh yeah, I see where the confusion may come from. Tbh, for certain zones, that could be more useful, as for example, its easy to skew up 20s-30s power by having an all-out 5-10s sprint and then riding at a lighter pace for the next little bit, which isn't that useful. In those shorter zone cases, an alternative of calculating average might be more appropriate, though I guess the reason they dont do it that way is because if your power ever dipped, it would get capped by the lowest point it dipped to.

(For example, my 20s power is recorded as 950w, but that was more as a result of doing 1100w for 5s and then fading very quickly down to 500w, and I dont think if I paced it at 950w exactly that I could hold it for 20s)

Funnily enough, I actually had a curve blip in a race last weekend from the same thing, average power for 40 minutes was lower than for 41 minutes, because the final sprint booped up the average :)

• nice hah

well back to the issue at hand in this thread. My estimated FTP is off again. Says 300W for the last six weeks which is definitely off.

Interestingly the data from 2017 seems to be right it shows my FTP at 212W.

Screenshot attached. It's funny bc my power curve is higher in 2017, but the FTP is definitely lower than 2018's reported 300W.

• Same has happened to me - ftp hovered around 190W for the last year (no reason to disagree with that) then a few weeks ago jumped up to 300W.  I wondered if this was down to a couple of weird turbo sessions (both recorded over 40 miles travelled in 50 minutes, and top speed over 70mph...) so deleted these, but no change to ftp.

• Bringing this back up.

I think what they have done is best 20min ftp / 0.95 when it should be 20min x 0.95

Appears that way for me anyway.

• I don't think they are using best 20*.95... my best 20 is 1 watt higher than my estimated FTP.

U moet u aanmelden om een opmerking te plaatsen.