my Garmin LOAD FOCUS is near perfect for once
I’m quite sceptical about quite a few of Garmin’s more advanced metrics, however, Load focus is one of the ones I love and regularly glance at.
Latest: Garmin Fenix 8 and Fenix “E” Leaked
Background
I was surprised to recall that Garmin /Firstbeat launched their method of determining Training Load over 7 years ago. All the more surprising when Apple announced their particular flavour of Training Load only a few weeks ago for WatchOS 11.
Training Load measure the work you have done in your activity and the more common methods of calculating it only work for endurance/cardiovascular workouts ie, not strength workouts. These methods can use power, heart rate zones and even perceived effort.
Garmin Training Load is based on EPOC
More: Firstbeat on EPOC
More: EPOC and Training Effect
Excess Post-exercise Oxygen Consumption (EPOC) measures the energy required for your body to recover after exercise. It is based on your maximum heart rate (HRmax). Specifically, the calculation assigns disproportionately higher scores for the time spent at higher percentages of HRmax.
Similar approaches might rely on TRIMP (Training Impulse), which uses Heart Rate Zones. These zones can be manually determined or dynamically adjusted based on your detected lactate threshold heart rate (LTHR). In the latter case, the zones are calculated using the threshold heart rate as the boundary for Zones 4 (ish) and 5, with percentages applied for the various boundaries above and below. Alternatively, the zones can be based on percentages of your HR ‘reserve’ (HRmax less HRrest)
To further complicate matters, Garmin incorporates HRV (Heart Rate Variability) in its LTHR estimate.
All these factors are interconnected. Some might describe it as a House of Cards. If there’s incorrect data anywhere, it can cause multiple metrics to fail. While I trust that Garmin has checks in place, it’s unlikely they cover every unusual event that could affect your numbers.
For instance, my HRmax seems to change weekly, sometimes increasing based on higher values achieved during a workout that supposedly triggered the increase.
An alternative approach would be to use power (measured by a power meter) rather than heart rate, especially for cyclists.
Training Focus
Sorry for all the acronyms.
Garmin determines an ideal range for your Training Load, primarily based on your training history and fitness. Your more recent workouts should add up and remain within that range. Over time, the goal is for the range to increase as you work towards your race goal. This progression allows your body to handle progressively increased loads over the weeks.
Almost there.
At any given point in your training plan, there’s an optimal balance of high, medium, and lower-intensity work. For instance, at the start of your training plan, easy workouts may constitute a higher proportion of your overall load. However, as you approach race day, the proportion of higher-intensity load should increase.
Garmin wisely avoids specifying a fixed amount of high-intensity training. Instead, it provides a range for you to aim for. Your actual training load, within the context of this ideal range, becomes your Training Focus.
Phew. We got there.
Why Is My Training Focus Near-Perfect?
I am mindful of capturing all my training data and ensuring its accuracy as best I can. In years gone by, I manually set my thresholds, maximums, and zones. However, now I simply go with the flow that Garmin dictates. The only intervention I make is stopping some of the automatic changes to thresholds or maxima that I believe are incorrect.
My goal is set for the English Standard-distance duathlon championships in a couple of weeks. However, due to family reasons, I suspect I won’t end up participating in that race. While I haven’t strictly followed Garmin’s training recommendations, I do consider any deficits highlighted by Training Focus when deciding what to do on a given day. I would also consider the effects of unusually low HRV reported by other products I use like the Ultrahuman Ring or Eight Sleep POD 4 – smart mattress.
My training has been significantly skewed recently, with some notably longer bike rides—like the Chase the Sun event last weekend. I’m unsure if a 200-mile ride in a single day is appropriate just three weeks before a duathlon that includes only a 40km ride. While endurance is valuable, such a long ride introduces the need for extensive recovery time. Perhaps that time could have been better spent focusing on faster intervals at this stage of my build-up.
Regardless, Garmin seems to agree somewhat, suggesting that I need to incorporate a bit more high-aerobic training.
So, in summary: Near Perfect By Accident 😉
Take Out
I trust Load Focus more than some of the other Garmin metrics. I kinda understand what’s happening behind the scenes and so Garmin should reliably get this right and, in any case, I’ve used it for years and it also feels right in the context of my experiences training.
Load Focus is a good metric to help self-training athletes tweak their training on any given day but it’s also a sense-check for anyone else following a plan. If Garmin is showing a significant discrepancy in aspects of your training it’s good to understand why your plan or coach is telling you to do something different.
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.
Interesting, I’m still in two minds with there algorithms, especially the Training Status. I’ve was in bed last week with a chronic sinus infection, and ventured out yesterday for a very easy 30 min run. As there is no way to record illness within the device, it saw my bed ridden days as tapering, so yesterday my TS changed from ‘recovery’ to ‘peaking’.. apparently I am ‘race ready’… mmmm
Any advice for getting anaerobic results? Both my low and high aerobic numbers are way over, but I struggle to get anaerobic. I’m mostly running, and I can’t sprint (old injury), so I don’t know how I can get anaerobic. Maybe get back on the bike? IDK. Thanks 🙂
I am primarily a runner, but I find it much easier to get Anaerobic scores on the bike, but only with a power meter. The Garmin algorithm seems to be looking for super high values (in watts or bpm) then a drastic drop during the recovery period, which needs to be ample relative to the work period. For running, I need to go VERY HARD and then take a long time fully recovering (i.e. walking or standing around).
def the bike
so you can’t even do short hill reps, they can be a good way to minimise injuries when anaerobic
Weight lifting, strength training, HIIT will help
HIIT or CrossFit will get you there. I am doing the latter twice a week (training load for 1 hour is between 250 and 350) and will get above the anaerobic target if I mix in any short interval running.
What Garmin activity are you using for these workouts?
tfk – My load focus is always underestimated because Garmin continues to vastly under-rate my cycling events where I don’t use a power meter. We’ve chatted on this before, but even though my HR zones are correctly set – I’ll get half-or-less on Load values when I have the power meter vs when I don’t. It’s frustrating but I’ve just come to expect it.
conversely maybe garmin is over estimating my endurance abilities. see my image at the end. i wrote about this ‘Elite’ endurance score a while back, nearly all comes from cycling WITH a power meter as I’m not running much right now
Hilarious take for me, hear me out: “my heart is too efficient”
Whenever I do a threshold, tempo or VO2max workouts, my overall avg HR of the workout is too low to get a high aerobic load, it always fall down to Base label (low aerobic load added).
After some investigation it’s not because my hrmax or my LTHR zones are wrong, it’s because my warm-up and cooldown are “too long” compared to the workout itself. My low HR in these two periods make it that my avg HR of the workout is under the condition for Garmin to consider it as something higher than base.
I have come to the conclusion to stop activity after the warm up, do a new one for the workout + cool down.
I can reproduce this easily on a threshold run or cycling workout (I use yellow HRM Pro and a power meter).
For context, I don’t have other choice for warm-up and cooldown distance, it’s the distance from my apartment to the location where I can do the actual workout (~15min each).
Correct,
Garmin’s largest weakness in Load Focus is its lack of granularity. It wants to put all the points from one workout in one bucket. As you describe, breaking a single activity into multiple records will effectively break those points into multiple (more appropriate IMHO) buckets.
Yes it’s a little fastidious having to stop/start new activities to slip warmup / workout / cool-down while being in the workout but as I have read on other forums, it’s apparently the way to do it if you care about having correct workout labels and correct load assigned.
“To further complicate matters, Garmin incorporates HRV (Heart Rate Variability) in its LTHR estimate.”
Nitpicky me says Garmin isn’t /incorporating/ HRV in its LTHR estimate, it’s /using/ your HRV inflection point as a detection method. It doesn’t matter what your baseline is or where your resting HRV was relative to that baseline, Garmin is looking for the change in HRV indicative of crossing the L2 point.
Yes I think that is correct.
Although, to nitpick, that means Garmin IS using HRV for LTHR 😉
I’m not sure about that (low average hr for entire workout)
If you do only record the high intensity part and it supports your theory then something is wrong with how garmin does this.
15 minutes IS a perfectly sensible and normal warmup time.
I can’t attach screenshots but trust me, if you do a very intense workout where you expect to have high aerobic load (obviously, with either tempo, threshold or vo2max label) but your warmup and cooldown make your overall avg HR of the workout go below 79% HR MAX (for me 174*79%=137.46) then you have Base label with low aerobic load.
And yes the workout part in the middle was avg >165. But cooldown and warmup, depending of how long it is, at least for me ~15/20min with HR 110/115 reduce the avg HR of the workout to the point it goes below 137.46. That’s why I now, when I don’t forget, start/stop new activity between the 3 parts.
Of course it’s not intended by Garmin. But I am pretty sure it’s intended by Firstbeat, the way they speak about their labels and stuff.
HRV from Garmin (FR 955) is for me close to useless, it doesn’t correlate well with my (perceived) load/recovery feeling. RHR works for me much better and is usually spot on.
For tracking my load situation, i prefer a simple model like the one from Trainingpeaks, based on TSS. By simple i mean it’s fully transparent / all formulas are known and issues can easily be traced back and (at least partially) fixed. For example when your power meter is through the roof because of a wrong configuration, you can dial down TSS for this session retroactively.
These Garmin metrics, on the other hand, are intransparent. You don’t know how they are calculated and if/when changes i the algorithm happen.
generally true
but TL and HRV are different. TL measures the work done whereas HRV measure how your body coped with it.
I do correlations most morning of various tech versus a 2 minute Polar H9/H10 HRV reading with HRV4training. I’ve not come to any definitve conclusions after a couple of years but I go through periods when the HRVs correlate and other periods when they dont. Is it the kind of training I’m doing? dunno?!?!? best correlation figures i had were for whoop many months ago, now it has similar correlations to the best of other tech
I find the garmin load focus too pyramidal for me. I prefer more of 80% z1-z2, 10% z3-z4 10% z5. I don’t respond well to the amount of high aerobic Garmin wants me to do. Not sure if i can change the training scheme. Surely base, polarized and 80/20 should be an option?