Garmin Training Readiness – Not Accurate – Here’s Why

Garmin Training Readiness Accurate
Which is more precise?

Garmin Training Readiness – Not Accurate – Here’s Why

Despite the inaccuracies & confusions that I’ll detail further down in this article, Garmin Training Readiness remains a valuable tool to maximize training efficiency and prevent counterproductive overtraining.It assesses an individual’s readiness level using a multi-layered analysis based on five primary factors: sleep (score & history), recovery time, acute training load, HRV status, and stress history. These factors are key in understanding the body’s preparedness for intense workouts and the need for adequate rest.

The first factor, sleep score, evaluates the quality of sleep based on duration, sleep stage distribution, and autonomic nervous system recovery data.

The second factor, recovery time, indicates when the body is expected to be fully recovered from the last activity, considering its strenuousness and other contributing factors. The third factor, acute training load, measures the combined impact of recent activities on current fitness levels.

The fourth factor, HRV status, assesses the balance of heart rate variability, reflecting the body’s ability to handle stress effectively.

Finally, the fifth factor, stress history, takes into account stress levels from the past three days while an individual is awake, which can affect training readiness and overall performance.

Training readiness is updated throughout the day, with significant changes occurring in the morning as sleep score, HRV status, sleep history, acute training load, and stress history are updated. These insights provide valuable guidance in determining the best times for intense workouts and when it might be necessary to allow the body to recover.

While training readiness is an essential aspect of optimizing training efforts, it should be used alongside other considerations such as individual feelings and experiences. Garmin specifically states that it may not directly predict race-day performance.

 

Sources of Inaccuracy

There are potentially many sources of inaccuracy ranging from the quality of the input to the correctness of the calculations and through to how you use the device.

Garmin’s latest Elevate 5 sensor is found on the Epix 2 Pro/Fenix 7 Pro and will be on most future, higher-end Garmin devices and this new generation sensor is more accurate than the previous generation. The likes of dcrainmaker have found it to be accurate during a variety of endurance sports and Fit Gear Hunter is impressed with the accuracy during gym/X-Fit activities. The Quantified Scientist is less convinced stating that the improvement over Gen 4 is modest and that deep sleep correlations are not great.

Q: Does any of that matter?

A: Yes. Mostly

Your acute training load (ATL) is a short(ish) term measurement of the load placed on your body by exercise. Its data will come from either a power meter or your heart rate monitor. So long as your optical sensor is ‘about right’ and your power meter calibrated then the ATL will be ‘correct enough’ in my experience.

Next, we look at your HRV which is a measurement of how your body is handling stress in its widest sense ie including exercise stress and general daily stress. Garmin takes this data from its optical sensor and the key measurement is based on the average of your night’s sleep. Broadly speaking this IS a scientifically correct method to use. There are quite a few factors that can make this reading incorrect, however:

  • Latency – how long do you take to get to sleep? How accurate are the start sleep and end sleep points that Garmin uses to delimit the averaging period? If you lie in bed reading for an hour Garmin may well assume you are asleep… that’s really your fault and not Garmins, in my opinion! I DO expect Garmin to be accurate but I don’t expect them to perform miracles.
  • Skewing of HRV – your late evening activities do impact HRV more in the first few hours of sleep. This could be from drinking alcohol or simply eating late; in both those cases and other cases your nightly HRV average will be impacted (lowered). Latency and skewing can both be reduced or even eliminated if you take a 1-3 minute manual reading every morning with a chest strap and a proper tool like HRV4Training, however, Garmin has the advantage of fully automating the data collection process overnight and, for most people, that is practical. In any case, it is also accepted as a scientifically accurate averaging method for the purposes of HRV.
  • Incorrect determinations in sleep stages – The Quantified Scientists analysis against a polysomnograph (PSG, below) tool raises concerns that Garmin is much less reliable at accurately recording HRV during Deep Sleep. This is the physical regenerative phase of sleep and so I would assume it is the most important one to get right. Note that I’m not talking here about Garmin incorrectly identifying Deep Sleep, it’s more that in that period as defined externally by PSG, Garmin is less accurate. I’ve not seen any analyses that show how impactful the calculations made in deep sleep are on the overall average, I would imagine it’s quite important and I believe that WHOOP used to focus its data gathering more on those periods of sleep before changing to a similar method as Oura/Eight Sleep.
  • HRV algorithms – Garmin uses widely accepted algorithms where its nightly HRV average is assessed against a multi-week moving average (your baseline range)
Quantified Scientist – See Garmin really isn’t as accurate as you’re told

Then we have the tinkering in the daytime. Garmin appears to use HRV readings from during the day to potentially boost or degrade readiness. My understanding is that this is somewhat of a made-up concept and has not been widely tested. It’s wrong! Although to get an alert that you might be recovering faster is also useful.

Then we come to the single-digit assessment of your entire training readiness. It’s great that it’s a simple percentage number. Everyone can understand that. However, I would challenge Garmin and all the other tech vendors to show the science behind how they add together all these disparate measures. AFAIK they are not additive. It’s wrong.

I think Garmin has also missed a trick by excluding ‘feel’. At the end of every workout on most high-end Garmins, you can now assign your RPE and FEEL to each workout. Garmin could use a similar approach to assess your feel when you wake and add that to the readiness score. But then, of course, it’s just yet another factor that can’t really be combined with the others!

Garmin Outdoor Sales Still Declining – Q2.2023 results are in

A Mis-Interpretation of One Reading

Garmin Training Readiness AccurateOK, I showed you this image at the top and the corresponding one below comes out of Connect both places give you the breakdown.

I think that Garmin’s algorithms are working as intended when this is produced. However, you would have thought that this is basically saying ‘Stay in bed’. The readiness of “1” is starkly clear, right? Well, no. It means that I shouldn’t train hard and I would say that is a correct reflection of what I should avoid today. If you look at the HRV Status then you will see it is ‘High’ or within my normal range (which it is). I am specifically trying to overload my body at the moment to elicit a certain endurance training response and today is the end of a hard training block. Thus, as my body is coping, I’m going to do something similar again today.

Or as Garmin say…

If you are using a block-based training model, then a lower training readiness score during a deliberate overload period likely echoes that work. Against that background, you should still check to see which factors are driving that lower score. Higher than normal acute training loads and longer recovery time estimates can be expected during block training. Reduced readiness caused by other factors (e.g., sleep history or high stress) might be a reason to re-evaluate your approach.[Garmin]

I would thus argue that the simple Garmin number is also misleading, adding to other sources of inaccuracy. Perhaps there needs to be a ‘Readiness to train AT ALL’ number and a ‘READINESS to train HARD’ number. I am TOTALLY ready to train today but, in agreement with Garmin to a degree, I am absolutely not ready to train hard.

As a counter-example, HRV4Training clearly and correctly (IMHO) tells me to stick to the plan…

I think Garmin has the presentation and interpretation wrong. Although I’m pretty sure they do understand the data they work with.

Tip: If you want to ‘just train’ check your HRV score

Tip: If you are following a plan and have low readiness on an intense day…replan

Take Out

Hopefully, the information above gives you an idea of why your Garmin readiness reading might be incorrect and/or confusing.

Readiness may well turn out to be useful for you so keep using it if that’s the case. FWIW, I would say that Garmin Readiness is useful to me but that it is not definitively correct and I treat it with caution.

I do use Garmin readiness each day but I also have other HRV/readiness tools including WHOOP, Oura, HRV4Training, Athlytic (Apple Watch) and Eight Sleep. They’re all pretty good and, in my opinion, even better when they all agree with each other 😉 WHOOP and Oura are more easily worn and HRV4Trainign is probably the most correct. I especially like Eight Sleep as it gives scientifically validated nightly HRV averages and adjusts bed temperatures to boost HRV and recovery

 

 

Fit Gear Hunter

Quantified Scientist

 

Reader-Powered Content

This content is not sponsored. It’s mostly me behind the labour of love which is this site and I appreciate everyone who follows, subscribes or Buys Me A Coffee ❤️ Alternatively please buy the reviewed product from my partners. Thank you! FTC: Affiliate Disclosure: Links pay commission. As an Amazon Associate, I earn from qualifying purchases.

13 thoughts on “Garmin Training Readiness – Not Accurate – Here’s Why

  1. “that’s really your fault and not Garmins, in my opinion! I DO expect Garmin to be accurate but I don’t expect them to perform miracles.”

    I don’t expect garmin to perform miracles, but in my view they are never the less trying to perform miracles.

    I think we all, or most, or at least some, of us agree sleeptracking with a watch isn’t very accurate. If you know that, then don’t call it advanced sleeptracking and don’t depend important advice you give your user on it.

  2. A bit of an aside, but I wish HRV4ttaining would accept readings from right sleep. Because as you said, of all the options they might be the most accurate at this time

    1. Chrus one option is to change HRV4Training to manual input, and then in the morning manually input HRV and RHR from any device that you trust.

      1. Is there any way to just delete this from my 965? The algorithm is hot garbage.

  3. I bought an Enduro and started to use just a week ago. I kept my F6X for a while. After I set wrist options in settings to left and right I started to use them parallelly as a test. Both were strapped solidly.

    And the good news that they run on the same software. The OHRs of these 2 watches are said to be the same piece of hardware.

    Even if they measure more or less the same total sleep time, the durations of the sleep phases are measured with huge differences, sometimes more than an hour. So I can get sleep scores with a difference of 10-15.

    Certainly the different cpu speed of these 2 watches may count, but if so then it is also a clear sign that the measurement of sleep phases is still a humbug provided we speak about Garmin watches.

    If somebody has 2 pieces of the very same Garmin he can test what sleep scores he gets for the same night.

    1. different arms can certainly give different readings eg if you wore 2 oura rings. maybe you have to alternate devices and wrists for consecutive nights. I’m not entirely sure it’s worth doing unless you are bored!

      there might even be machine-level variation from one watch to the next even of the same model (I’m just speculating that as a possibility)

      things like eight sleep and emfit get around the left/right thing by measuring whole body, as does hrv4t with a single morning reading

  4. Not really on this article , but as a shift worker i found Garmin useless as my venu2 cannot track sleep.
    If i dont put in the time i go to bed and supposed wake up time it won’t track the sleep correct, if you ask me , Thats not tracking at all.
    So without the correct sleep data, the stress and body battery won’t be correct either.
    I can use it as a pulse meter , standard Watch and count steps, thats it.
    And their support on this matter is nothing but a joke.
    That said , i will never buy a Garmin again , over priced piece of sh*t.

    1. yeah it should get bed in and bed out as correct.
      the body battery is interesting rather than correct. it’s based on your current hrv and so that will be shown correctly throughout the day but what it means is another matter entirely (Garmin can’t know what it means)
      i cant recall exactly how stress is recorded. its probably linked to workouts and/or low hrv. that should be about right

      1. Stress, body battery and hrv are all calculated from the watch routines , witch has to be learned all over every time you change sleep routines.
        Thats according to Garmin support.
        They told me that this is being looked at for future modells , thanks very much , my old used amazfit watch i bought for 25dollars handles it just fine.
        Also the accuracy is not that good , compared to other watches in the price range.

  5. The only thing in your example that is making it say 1 is your acute training load and estimated recovery time is over 2 days. These only reason it would say you recovery time is that high for that training load is if it has no previous baseline to know your average workout training load. For example if it sees your training load is usually 1000 then you do 1135 it will say keep going. But if it is working if nothing. Like if you did this review with only one week of training and are confused why it’s recovery time and recovery metrics is off then it’s because it has no baseline.

    The manual actually says give 1.5 to 3 weeks to calibrate to your usual exercise routine and it will take longer for advanced athletes.

    So the fact your writing a whole review about it being inaccurate without actually following instructions just means to me this is a user error. Many watches for the past 8 years has had training load and recovery so for you to not be used to this in a new watch is concerning as a reviewer.

    Please repeat review 3 to 4 weeks after using it consistently and let me know how this feature works

    1. hey Julian thank you for chipping in

      it’s usually best to ask and clarify before telling someone why they are wrong

      Baseline: My Garmin Connect data goes back over a decade and i have had the 965 since launch as my primary training device for every single workout. each one recording correct hr and cycling power
      yes you are probably partly right about the data cause in the sense that high acute training load is what is giving me the very low readiness. however, my HRV clearly shows I can handle this level of training as I have in the past eg when Ironman and HIM training. A readiness metric is about true, physical “readiness to train” not modelled fatigue (a la Bannister’s ATL/CTL/TSB) which is what Garmin is almost certainly relying on to determine this readiness reading. you’re actually highlighting another error in Garmin’s methodology rather than an error in my article. thank you for pointing it out

      Again, there is no such thing as a definitive ‘readiness metric’. You can ask Altini or Chuck Hazzard or any other of your favourite leading global scientists on the topic and I’m pretty sure they will agree with me. Actually, I did run some of the content for this article by one of them before posting 😉

      And, if you read more closely above I literally quote the Garmin manual that you also refer to.

      My exposure to and use of HRV/readiness goes back well over a decade. I was even involved in the development of an app for it. I don’t claim to be an expert in the subject but probably am one of the more knowledgeable reviewers on this particular topic.

      you can believe me or not. your call.

      sorry for taking you to task on this

Leave a Reply

Your email address will not be published. Required fields are marked *

wp_footer()