20min FTP Test Protocl/Pacing +Results

richa
richa Posts: 1,631
I want to know my FTP to (a) use it as a benchmark next season, and (b) consider for my pacing the Marmotte in 3 weeks. So plan to do a 20min test

I am having a shot at it tonight (friends Turbo & Powertap). But I am unlikely to get another chance before the Marmotte so want to make sure I do it right.

1. How should this be done: Just warm up (10 mins) and then do the test?

2. I read on another post something about doing warm up, 5mins test, recover, 20 mins test. Is this preferable? How long should the recovery be?

3. What about pacing? Can I use HR to help pace? i.e. if Box Hill is 7mins @172 (90%) how would this extrapolate into a 20 min test?

4. I understand that ideally power would be stready. However, what are the consequences of increasing/decreasing in the 2nd half if I have gone out a little too conservative/hard?

Any other tips?

Many thanks,
Rich
Rich

Comments

  • Anonymous
    Anonymous Posts: 79,667
    I used the Allen/Coggan test -

    Warm up, 15 mins with a few high cadence (>100) spins.

    Then, all out for 5 mins

    Then 10 mins recovery

    Then 20 mins FTP test

    15 mins warm down

    Then take 5% off the 20 mins average.

    There are many ways to do it and I actually tend to take my FTP from actual rides / efforts, if I see a new max peak 60mins (or 20 mins -5%) then then that becomes my new FTP.

    Allegedly, the most accurate way is doing an hour at 25mile TT pace.
  • Since FTP is defined as best average power for a quasi steady state effort of about an hour, then the best test is a maximal effort of about an hour.

    20-minutes is a bit short as a test on its own, since anaerobic capacity has the potential to skew the results. Your FTP might be 93% of 20-min power or it might be 96%. Usually you would want to use the result of a 20-min test along with other indicators and not rely solely on the 20-min test.

    If a full blown 1-hr effort is not possible, then I would recommend examing the critical power from from maximal efforts, one about 25-minutes long and another 3-5 minutes long. The CP model isolates your anaerobic work capacity from your steady state power capability.

    As for pacing....
    I would suggest starting an effort of 20-min at around about the power you think is your FTP, then after a few minutes you'll get an idea if you can increase it from there. If you are finishing strongly, then don't stop at 20-min, keep it going for a few more to lengthen the effort and reduce the impact of the opening minutes on the average. It won't be far out anyway.

    You can read up on FTP testing here (including use of the critical power model):
    http://alex-cycle.blogspot.com/2008/05/ ... -sins.html
  • liversedge
    liversedge Posts: 1,003
    Why not take his powertap outside and blast round RP for an hour as hard as you can?
    --
    Obsessed is just a word elephants use to describe the dedicated. http://markliversedge.blogspot.com
  • richa
    richa Posts: 1,631
    I like the idea of a repeatable test. i.e. 20mins on the turbo. Nothing to influence the ride i.e. hills, traffic, roundabouts, pacing.

    I also find that I struggle to push on the downhill bits in the park.

    Riding budy is riding the dragon ride this weekend with me so I will also analyse the data from that. We are same weight so power will be the same (?).
    Rich
  • RichA wrote:
    I like the idea of a repeatable test. i.e. 20mins on the turbo. Nothing to influence the ride i.e. hills, traffic, roundabouts, pacing.

    I also find that I struggle to push on the downhill bits in the park.

    Riding budy is riding the dragon ride this weekend with me so I will also analyse the data from that. We are same weight so power will be the same (?).
    Repeatable is good.

    Do you know your relationship of power you are capable of producing indoors on a turbo to what you can produce outdoors? For many, they simply can't sustain the same power indoors, at least initially.
  • chrisw12
    chrisw12 Posts: 1,246
    Do you think that doing a 20 min test to influence pacing for a long SPORTIVE is a little bit pointless/ott?

    Wont you just be better pacing it on feel since that is what you've been doing up till now?

    During the event, so what if you get the pacing wrong. The brain ime is very good at not allowing the body to go too mad. Ok if you do get the pacing wrong, what exactly are you going to loose?

    I ask because I'm on the verge of buying a power meter and I'm on the verge of doing my 6th 100 mile tt. I can't see how knowing my 20 min power NOW should have any influence on how I pace the 100. Perhaps next year, after a year of having a power meter then I'll set up a model, but I honestly can't see why one data point could/should be used for an important event.
  • chrisw12 wrote:
    Do you think that doing a 20 min test to influence pacing for a long SPORTIVE is a little bit pointless/ott?

    Wont you just be better pacing it on feel since that is what you've been doing up till now?

    During the event, so what if you get the pacing wrong. The brain ime is very good at not allowing the body to go too mad. Ok if you do get the pacing wrong, what exactly are you going to loose?

    I ask because I'm on the verge of buying a power meter and I'm on the verge of doing my 6th 100 mile tt. I can't see how knowing my 20 min power NOW should have any influence on how I pace the 100. Perhaps next year, after a year of having a power meter then I'll set up a model, but I honestly can't see why one data point could/should be used for an important event.
    My view is that a maximal effort over a much longer duration would be a better predictor of performance and combined with FTP, it can be a very helpful in modelling a pacing strategy. Certainly I have provided a pacing strategy before along those lines which proved to be effective.

    But you are right as well, as for instance, if a rider doesn't know how they are going to be affected by riding at altitude, then how hard they think they should go on the upper reaches of Galibier etc is really going to be dictated by PE as much a by power.

    Then there are those who end up walking the Alpe D'Huez because they did not pace themselves properly as the brain does not always moderate effort so well early on. Temptation to go with faster riders or over stretch can be costly later on.
  • richa
    richa Posts: 1,631
    Ok. So I did it. And it was v.hard. Reckon that is about as tough as I have had it on a bike in the UK.

    Adopted protocolfrom Hunter Allen's book:
    > 30 mins warm up @ 70% HR (with some 100 cad spins)
    > 5 min FTP test
    > 10 mins recovery
    > 20 min FTP test
    > 5 min recovery

    The 5 mion test seemed to go quite well. However, on the 20min test I suffered. Started out at c300w and soon realised I was not going to sustain that so had to drop to c270.

    I would be grateful for any comments on how this data can be interpreted.

    Results:
    ftp11062009.jpg

    One thing i don't understand is that the formula implies I could go on forever at 260W? See the column I added with 400mins. Is this right? Would I therefore pace a 7hr piece at this power? It seems very close to my FTP/1hr power (265w).
    Rich
  • The critical power model breaks down for durations much longer than an hour - it is not a good predictor out that far. It works pretty well between a few minutes and an hour though. Provided that both efforts represent best maximal effort and are not separated by too many days (like within the same week or so).

    For FTP, I would use the CP value, rather than the 60-min predicted power.

    In this case that's 3.30 J/s/kg x 78.3 kg = 259 watts

    No point being that precise, usually I round to the nearest 5W, so call it 260 watts.

    As for long range prediction of power, well I would expect that mean maximal Normalised Power to decline by about 5-7% or so for each doubling of duration. Everyone however has a different shape to their mean maximal power curve and of course it very much depends on the nature of fitness you develop.