Periodic Fitness Tests

Tagged: , , ,

Viewing 15 posts - 1 through 15 (of 21 total)
  • Author
  • #9939


    Is it better for the athlete to do a full 30 minute fitness test in Z3 (with warmup and cooldowns too of course) periodically (3 weeks, 6 weeks), or to do 20 minutes and do the 95% calculation? I see the 20 min Z3 test on the maintenance plan in the forms of RT13 and CT30.

    Does what you train with (power, HR, or pace) make a difference in this decision?

    All of my structured cycling and running workouts are driven by power, but I capture and analyze HR for both, as well as pace for running – not necessarily because I need HR and threshold pace, but just because I have it from the 30 min test. With the 30 min test I get LTHR which I wouldn’t with a shorter test.

    Does what I am training for (what duration of tri) have an impact on this decision? So if you are training for longer tri, you’d be more inclined to do the 30 min test over a shorter test, and shorter tri do a shorter test?

    Does your age have an impact on the decision? As 50+ I recover from hard efforts more slowly, but also know how important moderate and high intensity efforts are to staying strong and fast as I get older (Faster After Fifty by Joe Friel)

    I find the 30 minute rFTP and cFTP tests are absolutely brutal to do. I dread them. They are lung busters and mentally hard to prep for and execute. I am mentally and physically exhausted for a day or two after each. But I am also so psyched when I have finished them with a solid effort. I know a) the data from those tests is critical to making the most of my training, and b) training at that intensity (in an 80/20 balance) is very good for my fitness, especially as I get older.

    Because I dread the 30 minute test I have been doing them every 6 weeks generally, rather than every 3 weeks – replacing whatever tempo workout is in the schedule with a 30 minute Z3 test with warmup and cooldown. But would I be better off doing the 20 minute power/pace test every 3 weeks instead, and just rely of the CT and RT workouts in the plan (whichever plan I am doing at the time)?

    A related question – what should I do with the data of a 24 minute Z3 effort (CT31)? That’s neither 20 nor 30 mins for the calculation, but CT31 appears to be the testing workout (scheduled for recovery weeks) for cycling that appears frequently in the Half Iron Level 2 plan.

    Thoughts, feedback? Thanks!


    Hi @tedc

    I am approaching the big 50 and completely understand your mention of the mental and physical affect it has on you. Because of this I also either skip every 2nd or 3rd one. I thought it might have just been me, so I am interested in seeing what Matt/David say.

    David Warden

    Ted and Dan,

    We are slowly moving away from the 30-minute test to other options. First, check out our newly designed 80/20 Zone Calculator and you’ll see two new option for bike and run: 20-minute and Talk Test. These will be the primary tests going forward. The Talk Test can be done in any workout. The 20 can be done in place of any RT or CT workout.

    It’s going to take up some some to completely re-do the documentation, update workouts and plans, etc, but the testing protocol is definitely improving.

    The Talk Test is super easy can be used right now for the run, check out Intensity Guidelines for Running and we’ll have it in the Tri doc ASAP.

    Matt thinks the Talk Test can be just as good as the 20 if executed properly. I’m not convinced at all yet.

    To answer your direct questions:

    – I now recommend the 20/95% over the 30/100%.
    – Does what you train with (power, HR, or pace) make a difference in this decision? Hmmmm, I need more time on that one. I keep writing “no”, the backspacing. I have a nagging feeling there is a difference. Come back to me on this one after we’ve had time to work out these new protocols.
    – Does what I am training for (what duration of tri) have an impact on this decision? Same. Maybe. The Talk Test and 20 are new recommendations for us, I can’t in good conscience give a yes or no yet.
    – Definitively yes. This was one of the major decisions for us to re-do the testing. A revolt among the 50+. Shorter test or Talk Test for sure.
    – I think the 20 every 3rd week is doable for most athletes, not all.
    – I’d round the CT31 down to the 20-minute test from now on. Take 97% of that value for now if you have a full 24 minutes.




    Just curious. Do you feel the 20/95% is more accurate than the 30/100% or just easier to do and fit into a training plan and the results of the 20/95% are good enough to train against?

    I’m with you on not being convinced of the talk test. How do I know I’m “talking comfortably”? Am I actually talking comfortably if I think I am with regard to trying to gauge a threshold pace. Like I mentioned in another topic, the talk test is another one of those nebulous concepts that’s too difficult to judge accurately. Just my opinion.


    I have done 2- 30 minute tests each for bike and run. The first bike test went poorly (went out of the gate too hard), but the runs were both solid. I just did a 20 minute run test today. Here are the 3 run tests.

    First 30 min run test – 10/14/20 (TP link)
    rFTP 292 watts
    rTP 7:52/mile
    rLTHR 182 bpm
    Elevation gain: 119 ft
    Elevation loss: 135 ft

    Second 30 min run test – 11/17/20 (TP link)
    rFTP 302
    rTP 7:30/mile
    rLTHR 180
    Elevation gain: 92 ft
    Elevation loss: 171 ft

    Today’s 20 min run test – 01/19/21 (TP link)
    rFTP: Normalized Power for 20 mins was 327… 95% would be 311
    rTP: pace for 20 mins was 7:06/mile… 95% would be 7:31
    rLTHR: avg heart rate for 20 mins was 183 bpm…
    Elevation gain: 125 ft
    Elevation loss: 128 ft

    A few things come up:
    1) a 20 min test is still pretty brutal on the body and the mind, beforehand, during, and after. It felt awful. It’s a harder run than the 30, just for 2/3 the time, so I’m not sure if it’s any easier on either the body or the mind as a training tool. David/Matt – is there any data to suggest a faster 20 min run is easier on the body/mind than the slower 30 min run? I will say after doing the complete maintenance program (with power), I feel like I was less familiar with the zone I had to run in for the 20 min test. Running with Stryd power, I took my current rFTP (302w), divided it by .95 to get the amount that would be for a 20 min run (318w), and then added 15w for training progress I was hoping to see (333w), then built a modified “zone” of 326w-341w for the specific workout (5min @Z1, 10min @Z3, 20min @ 326-341w, 5min Z1, 10min Z2, 5min Z1.) I feel like the training plan has had me doing a fair amount of lengthy Z3 work (284x-302w), but this would be 20min in the middle of Z4 which is very hard, and not something the plan has me doing anything close to. Considering the 20 min test is a intensity and duration that isn’t practiced much, is this a better test than the 30 mins?

    I know a) checking my fitness regularly, and training at this intensity are both really important, so that’s on balance just enough motivation to put myself in such an uncomfortable situation regularly.

    2) The Zone Calculator at 8020endurance says take 95% of the 20 min Avg HR to get the LTHR… That would be 174. That doesn’t make sense to me. I have done 3 tests now. 2 were the full 30 mins. And the avg on the first was 182, avg on the second was 180, and the avg for 20 mins on the 3rd was 183. The 95% just doesn’t seem like a valid formula for HR in this specific case. Looking at the bpm chart for the 20 min test on 1/19, it took 4:01 into the 20 min interval for my heart rate to rise into a range it then held and slowly gained over the duration of the 15:59 remaining, from 175 to 191. A better way to estimate LTHR? If I look at the 15:59 period after the HR ramp up, the avg HR is 185. 98% of 185 is 181. I dunno… But 95% formula for 20 mins just seems clearly problematic with my data and perhaps it’s not a good solution for a broader group either. 174 seems pretty far off from my LTHR.

    3. Do you see many 50+ year olds who get up to 192? So much for the formula 220 minus your age as your max HR! LOL)

    4. I was surprised to see the gain in watts from an rFTP of 302w on 11/17/20 and then a calculated rFTP of 311 from the 20 min today, but NO increase in rTP from 11/17/20 (7:30/mi) to the calculated one on 1/19/21 (7:31/mi). In the earlier comparison between 10/14/20 and 11/17/19, in discussion with David, seeing a bigger change in rTP (from 7:52/mi to 7:30/mi) than in rFTP watts (from 292 to 302), we pondered that perhaps I had gotten more efficient as a runner – running faster without exerting lots more watts. So then seeing a bigger increase in watts from 11/17/20 to 1/19/21, but essentially no change in pace, made me question that explanation and search for a better understanding. Also – Stryd power is basically just factoring forward pace (distance over time), runner weight, elevation gain/loss and headwind I think, right? Since it’s just a high end accelerometer, I don’t think it can illuminate running efficiency – covering more ground with less exertion – it’s calculating a watts of exertion with pace as one of the inputs…

    David, Matt, et al – Do you agree with that assessment about Stryd power and efficiency?

    While all 3 tests have been done in the same area (Prospect Park in Brooklyn), they have not had exactly the same course, and there are hills up and down. I looked for the first time at a comparison across all 3 of elevation gain and loss, and can see gain and loss clearly show differences between all three. Unlike cycling, where I tend to just look at elevation gain, it seems both gain and loss might play a factor in pace. I know going down hills is exhausting on the legs (hello first few miles of the Boston Marathon!) So it seems like my fluctuation in rTP was impacted in part by the differences in elevation gain and elevation loss in all 3 tests.

    David, Matt, et al – Any thoughts on how to view and factor elevation gain and loss data? It seems to me that a run with 200 ft elevation gain and 200 ft elevation loss is not the same as a run with 0 ft elevation and 0 ft elevation loss – that’s it’s not about the “net elevation gain”.

    Perhaps this is a good testament to the value of Stryd running power data in training – with varying terrain on these 3 tests, it was able to give me the best, I think probably most accurate, metric for my fitness gains. 10/14/20 rFTP 292w, 11/18/20 rFTP 302w (3.4% gain in 5 weeks), 1/19/21 rFTP 311w (3% in 9 weeks but the holidays

    Please feel free to share any other thoughts/feedback.

    David Warden

    @Bstarr The 30/100% is definitely more accurate than the 20/95%. But, a VT test is more accurate than an LT field test. We have to give up a small level of accuracy for convenience and practicality. I have total confidence in the 20/95% test to be really close to the 30/100% and still recommend it.

    We’re going to come up with some better instructions on the Talk Test, like a video example to make that one more clear. In my experience, if Matt recommends it, it is worth while to persue.

    @Ted First, contrasts on your progress. I’m showing 95% of 7:06 is 7:38, not 7:31. See and enter in 7:06 in the 20-minute Method.

    – No empirical data that 20 is easier than 30 other than 20 is less than 30. This is good feedback. In general, the feedback has been relief and positive that the 20 is less brutal, but doesn’t mean it is easier for everyone. I see your concern with moving from the 30 to the 20, but I think the advantages outweigh the disadvantages. You can keep the 30 if you want. Note that the updated calculators have the 30 and the 20 still there

    – If you have a 30-min result: use it. The 30 is better than the 20. No need to take the 30-min 302 and make it a 20-min by dividing by 0.95 and then add 15…. just use either 100% of 302 you did on Nov or the 311 you did today (95% of 327). I propose your thresholds have been 292, 302 and now 311 over the last 3 tests using the recommended methods. Use 311 until the next test.

    – Noted on the 95% of HR. I don’t have an answer for you at this time. It’s linked to the decision to bring all tests down to 20 min. We may adjust again. I hope you’ll recognize that we are always trying to improve the process and this means there will be some bumps as we deploy new protocols. Again, you can use the old peak 20 of 30. You’ll also note my objection to HR in general is based on inconstant results from HR. You’re HR could have just been depressed today, or artificially high the other days. It really does not matter if we use 20, 30 or something else: HR will do what it wants based on the environment.

    – Agreed an age-based HR calculations. They are not helpful.

    – Again, I show your 95% of 7:06 is 7:28, so a slight improvement. but your point is still valid. Any decrease in watts for a given pace or increase in pace for a given watts is an increase in performance. Or, a change in weight 😉 But you are exactly right and this is a normal phenomenon when increasing fitness.

    – Yes, regarding elevation, use the NGP value in TP when looking at a run: Normalized Graded Pace. Unless you are running pancake flat, NGP will difference than average pace, and you’ll want to use NGP which factors in elevation changes.

    Need to go soak my swollen fingers in some ice.



    I am interested to read about the talk test and I’ll definitely be trying it out.

    But I have a question around testing. I haven’t historically done much explicit run testing. I generally just used my most recent 10k time as my threshhold, whether it was last week or last year didn’t matter to me (and then only when motivated, admittedly, I have done 99% of my lifetime runs no higher then low Z2). I have only started testing regularly since I got an 8020 training plan — and a power meter — this past Fall.

    Test1: 342W, 6:58 min/mi, 176 bpm
    Test2: 340W, 6:52 min/mi, 176 bpm
    Test3: 318W, 7:27 min/mi, 174 bpm. I think I gave up at 28 minutes here after I couldn’t climb back out of Z2 at the end).
    Test4 (20 minutes, just this morning): 302W, 7:48 min/mi, 172 bpm

    The first two tests are within range, if slightly better, of my 10k PR. But since I’ve been using a real training plan for the first time ever, suddenly my average daily effort was way faster than it had been, and I just couldn’t sustain the effort. Every workout is suffering. Z4 efforts have fallen to Z3, Z3 efforts to Z2, …

    Is it possible those first two tests are really super-max anaerobic results, and my FTP should be closer to today’s result? Should I reset my training zones to go slower on a daily basis? I was hesitant after the first bad test 4 weeks ago, but it feels like the evidence is telling me 340W is just too high.

    Conversely, bike and swim efforts haven’t been amazing, but they’ve probably been pretty average. So I don’t feel overtrained in general, just on the run.

    FYI, I have done more bike testing historically because it’s harder to get appropriate-distance bike races to gauge. But I hate them, and maybe I’d do two annually. The 30 minute test has been a bit of a relief to me, 30 minutes at ZY feels less brutal than 20 at Z4. But talk test here I come, I think it sounds the best of all.

    David Warden


    I don’t like the trend for sure. Something is “wrong.” If you are not experiencing similar issues on the bike and run it stumps me further.

    I don’t think the first two tests were wrong. You had two solid tests, consistent results, and then the last two just are off the map. I’m used to seeing one test, even two tests in a row decline from significant training, but that’s a tremendous drop from test 2 to test 3.

    HR is still relatively high, lower, but high enough that I’d expect a much higher pace for tests 3 and 4.

    I may not be able to help remotely on this one, but I recommend the following:

    – Let’s rule out overtraining. Get some bloodwork done. I rarely recommend this, but it’s appropriate here.
    – Switch from Power to Pace for the next 4 weeks, I want to simplify the problem and just reply on one measure.
    – Use 7:20 as your threshold pace and adjust pace zones accordingly and report back here after one week of those zones.



    Hey, thanks for the tips, David! The 7:20 threshhold this week felt pretty good. I didn’t have a lot of high intensity efforts on my calendar this week, but my intuition is that the effort was where it should be — I’m glad I didn’t reset to 7:45.

    RCI1 3x(5min Z3, 3min Z1): I targeted a 7:30 – 7:40 here, middl-ish of the Z3 range. Avg HR for the 3 intervals was 163/166/169, maxing at about 172 each interval. No problems meeting or holding the pace.

    RAe12: 3x(15 min Z2/X, 5 min Z1): I targeted an 8:30 – 8:40 here, right at the border of Z2 and ZX. Avg HR was around 150, no problems meeting or holding the pace. The Z1 finish was pretty rough (10:35 pace for 65 minutes), but not unusually so.

    My last run was just a 40 minute brick off the long bike (CAe21, 4.5 hours) for the week. I usually have trouble hitting even the middle of Z2 for these, but I did about a 9:50 over 30 minutes on a hilly route, so pretty acceptable for me.

    Overall, I felt fresher this week both during the runs and after. To tie it back to the original theme of the thread, how does it affect how we interpret the fitness testing?

    David Warden


    OK, good progress on figuring this out. It’s still a bit of a mystery, but we’re one step closer. I’d feel better if you got that blood test to rule out overtaining.

    Let’s drop your threshold down to 7:15 for this week and report back again. It may take a few weeks of this to come to any conclusion.

    This is not ideal training, we’re sort of wagging the dog here and trying to reverse engineer your thresholds. But, I think we can get you to the right place.



    David/Matt – For cycling FTP, have you evaluated the effectiveness of the ramp test compared to the 20 minute and 30 minute all out effort tests? Zwift, TrainerRoad, and SufferFest, among others, recommend it as an effective way to regularly test FTP with less impact than the extended all out efforts.,p1z,missions,locale_es_eu,gift_card_end_of_life,NEWTAGSYSTEM

    Ramp Test Makes FTP Testing More Efficient and Less Stressful

    David Warden

    Ted, I’m good with the ramp test, I think it can be a acceptable protocol. For example, if you perform the 20-minute test one week, and do the ramp test the next week and they come up with a very similar result, then it works for you. In fact, the ramp test is so easy you can perform the ramp test on a Monday and the 20-minute test on a Tuesday without fatigue and really compare the two.

    The ramp test does not work for everyone. And, the truth is, the longer the test the more accurate the results.

    We may eventually add it to our list of approved protocols if I can gain more confidence in it. I suppose if we have now introduced a talk test, a ramp test is not an unreasonable addition.



    I really find those FTP test discussions quite interesting. I actually tried the ramp test and for me, the ramp test didn’t work – or in other words, the result compared to a 20min test differ significantly (in my case the ramp test would give a much lower result) and as David knows, I was doing a different 20min protocol (Coggan) which potentially gives different results than the 20min protocol on this website…how confusing 🙂
    David, I would assume that the ramp test is more tailored towards a cyclist that can put out relatively high power for shorter durations (or, I hope I say this correctly, more towards riders who can ride at VO2 for extended (whatever that means) periods of time). Cyclists who are better at maintaining threshold power for longer periods of time but lacking higher output are underrated when doing a ramp test?
    The ramp test from Trainerroad also differs from the Zwift ramp test. I think that the Trainerroad test increases every minute by 6% (?), Zwift increases by a fixed 20 (?) watts. I also think that the starting point in Zwift is different compared to TR but I cannot remember exactly (I do believe that TR starts at a certain % of your FTP (maybe around 50% – Zwift starts the test at a certain wattage maybe around 100 watts).
    At the end of the test, 75% of your highest 1 minute effort is estimated to be your FTP.



    Thanks, David. Great to know you think the ramp test could best useful.

    RAMP TEST PROTOCOL: What does the best ramp test protocol look like? 5 mins easy, start at 50% of recent FTP for 1 min, increment 6% for 1 min and repeat to failure. Then take 75% of the last completed full minute interval’s watts, and add some more watts based on how far through the final interval you got (maybe # of seconds/60 * interval watts * 75%)

    With your blessing, I’d like to give the ramp test a try in comparison to the standard 20 or 30 minute FTP test. I have done the ramp test many times before back when I trained on/by the TrainerRoad platform, though never really a closely timed comparison to a full 20 or 30 minute FTP test.

    DOING BOTH RAMP TEST AND 20-30 MIN TESTS IN SAME WEEK AS A COMPARISON: Interesting you are suggesting I could do the ramp test one day and a full 20 minute test the very next day. Can you give me an idea how those both would fit within in a recovery week of the Maintenance Plan (weeks 3,6,9,12) and the Half Iron Level 2 Plan (weeks 3, 6, 10, 13 etc), with other workouts that week. Within that question, is the ramp test a complete workout or part of a longer workout?

    VARIATION BETWEEN RAMP TEST AND 20-30 MIN TEST: How much variance between the 20-30 minute test and the ramp test would be acceptable? And perhaps, if it comes back below or above the number from a 20 or 30 min test, but is consistent in that way, might there be a way to make an adjustment to the number from the ramp test to get to the most accurate number? Maybe this idea is a stretch… looking for a reaction…

    OBSERVED VS NON-OBSERVED TEST: I know that the 20 minute test, using 95%, or 30 min test using 100% of the calc, is a “non-observed” test and estimate to replicate what you could do in an “observed” setting, right? So then what if you are doing a ramp test or 20 or 30 min test that is “observed”? Should you account for that somehow? TrainerRoad launched Group Ramp Tests a year ago. Obviously doing a ramp test in a live group is going to lead to a harder effort than alone.

    FITTING RAMP TEST INTO A TRAINING PLAN – ADJUSTMENTS: OK, let’s say the numbers in a comparisons of these are quite close, close enough that the ramp test is a suitable fitness test option, then what sorts of adjustments to the training plans (let’s focus on Maintenance Plan and Half Iron Level 2 as specific examples) would you suggest. Your guidance on those plans, the first which I did to completion and the latter which I am in currently, would help me to understand how to include the ramp test and make proper adjustments on any plan or modifications I need to make on the fly “as life happens while I am training.”

    If the ramp test is accurate, isn’t as debilitating after, and doesn’t require as much rest in advance of the fitness test as a 20 or 30 minute test, what workouts would it replace in the existing plans?

    Could the ramp test be done on a Tuesday after a Monday off in a microcycle weekly or somewhere else in a step cycle other than a recovery week – moving/removing a cFTP test from recovery weeks leaving only rFTP and swim test? Is the ramp test typically the full workout for the day or part of a longer workout? What kinds of workouts would you pull out when inserting a ramp test, of course assuming that the calculations for the weekly microcycle add up to 80% zone 1-2 overall for all disciplines (SBR), and 20% Zones X-5 overall for all disciplines. maybe you have other ideas for how you’d integrate the ramp test into the training plans? If you were able to rely on the ramp test for FTP tests, would you still do workouts like CT31 on recovery weeks in the Half Iron Level 2?

    David Warden

    @winoria, I think you’ve identified an important genesis of not only the testing protocols used by different coaches, but the genesis for many applications of endurance training. Coggan identified his test as the best test from years of experience coaching pro cyclists. I’ve use the 20 and 30 based on my 1.5 decades of coaching elite triathletes. TR developed their test based on other factors, but I’m sure that there was a significant commercial factors in their decision (“what test is accurate, but also most likely for athletes to remain on our platform”?

    Each of us is like the blind men and the elephant, using our own experience to identify the “best” part of the threshold testing beast.

    I’m confident that all 3 would work, and as you pointed out, I’m also confident that some tests are better for certain athletes. Ramp Test 1 might be best for Athlete A, the 30-minute test for Athlete B, the 20-minute for Athlete C, and Ramp Test 2 for Athlete D.

    But Coggan, 80/20 and TR all face the same problem: we really need to keep the testing simple and consistent for our athletes. We can’t just say “pick you own threshold test.” So, we pick and endorse our best method for the masses, and the take our lumps from the 10% who struggle with the method we chose and say it does not work for them.


    “What does the best ramp test protocol look like?”
    I can’t answer this, I just don’t have the competencies or experience with ramp tests. This may be a failing on my part, I never trusted it because it felt like witchcraft or a free lunch. I am not crazy about the talk test either, it’s still on probation. But, more importantly, I’ve never seen empirical data to support ramp tests like I have with the 30-minute test. My recommendation is similar to what you suggest: try them all and see which one comes closest to your actual 20 or 30 minute test.

    The ramp test is so easy it’s not a matter of when you should do it, it’s only a matter of when you should not do it. Don’t do it on the same day as your formal 20 or 30 test. Otherwise, any day rest week or no rest week will work. It can be at the beginning, middle, or end of any given bike workout or as a separate 20-minute workout. There should be a deep warmup (any ramp test that does not have you perform at least a 15-minute warmup is already suspect in its credibility).

    If you are pinning me down to something more specific, perform the ramp test within the first cycling workout that occurs each rest week after a 20-minute warmup. Don’t add the ~5 minute test onto the workout, it’s now part of the workout. Your scheduled CT test will usually occur the next day.

    Whatever you choose, just make it consistent. Always perform it in the same way.

    I don’t know, but I’m guessing anything within 5% of the 20 or 30 would be acceptable. That’s not a statistical 95% confidence level unless you perform the test like 20 times to get a decent sample size, but 95% is close enough.

    Oh, wow. This is a great question. Observed and non-observed performance results are real. I can’t answer this specifically as it applies to the ramp test, but I would say that the observational phenomena that impacts the 20-minute test will equally impact the 5-10 minutes ramp test. I don’t believe the observational phenomena is disproportionately represented in the ramp test.

    If you find that the ramp test is accurate for you, perform it weekly as part of the first cycling workout of each week. Continue to perform the CT workouts as written, as they play an important role in the 80/20 ratios, but the CT tests are no longer have a test component, they are just solid Zone 3 work. I know you asked a lot of questions in that last paragraph, but it really is this simple and no major changes are required. Just add it within the first cycling workout of each week and continue to perform the CT workouts as written.

    Here’s a question back to you: if you try the ramp test for the bike, why would you not also try it for the run?


Viewing 15 posts - 1 through 15 (of 21 total)
  • You must be logged in to reply to this topic.
Posted in


8020 Ambassador