I went down a bit of a rabbit hole today trying to understand how Apple Watch heart-rate sampling actually works during a workout.

I did a fartlek run this morning with my Apple Watch Ultra 4 and exported the raw Health data afterward.

The workout was about 40 minutes long. When I looked at the heart-rate timestamps, I expected to see something simple like:

10:00:00
10:00:05
10:00:10
10:00:15

But that isn’t what happens.

Instead I saw intervals like:

4 sec
8 sec
3 sec
6 sec
5 sec
7 sec
2 sec

Across the workout, though, it averaged almost exactly one stored HR value every ~5 seconds. The median was also 5 seconds.

There were nearly 500 HR samples, and even during a fartlek with constantly changing intensity, I didn’t find any meaningful sensor dropout. The longest interval was only around 9 seconds.

That made me think about what “continuous heart-rate monitoring” actually means on a wrist-based optical sensor.

Wrist HR is always an estimate

A chest strap has a relatively easy job.

It detects the electrical signal from each heartbeat and measures the time between beats.

A watch has a much harder problem.

The Apple Watch shines green LEDs into the wrist and measures tiny changes in reflected light as blood volume changes with each heartbeat.

But while running, that signal is mixed with:

• arm movement
• cadence
• changes in wrist contact
• vibration
• skin characteristics
• blood perfusion
• ambient conditions

So the watch isn’t directly “seeing 165 BPM.”

It’s seeing a very dense, noisy optical waveform.

Conceptually, it is more like:

Hundreds of optical measurements per second

- detect the pulse waveform

- remove motion/noise using other sensors such as the accelerometer

- identify reliable heartbeats

- estimate the underlying heart rate

- eventually output something like 165 BPM

That final 165 BPM is therefore a processed estimate, not an instantaneous raw measurement.

Why approximately five seconds might actually make sense

Initially I wondered why Apple doesn’t simply give us a new HR number every second.

But maybe that is looking at the problem backward.

The goal isn’t:

“Give me a number as frequently as possible.”

The goal is:

“Give me the most reliable estimate of my actual HR.”

At 165 BPM, your heart is beating around 2.75 times per second. Over roughly five seconds, that’s almost 14 heartbeats.

That gives the algorithm much more information to work with than trying to calculate an independent heart-rate value from a very short window.

So instead of chasing every one-second fluctuation, Apple can use several beats plus a huge number of optical sensor readings to produce a more stable HR estimate.

I suspect it is probably a rolling process

One thing I don’t think is happening is:

“Collect exactly five seconds → calculate HR → discard everything → start another five-second block.”

Apple doesn’t publicly document the exact algorithm.

A rolling or overlapping processing window seems more plausible.

Something like:

previous pulse information

• newest optical samples
• motion information
• signal confidence = updated HR estimate

That could explain why the timestamps aren’t perfectly five seconds apart.

Sometimes the next reliable value arrives after 3 seconds.

Sometimes 6 seconds.

Sometimes 8 seconds.

But over a long workout, the cadence averages around five seconds.

Interestingly, my data also showed that Apple wasn’t simply storing a new HR value only when HR changed. There were periods where it repeatedly stored essentially the same heart rate.

So it doesn’t appear to be:

“HR changed enough → save a sample.”

It looks much more like a continuous signal-processing system periodically producing a confident HR estimate.

And this may explain why Apple Watch can track a chest strap surprisingly well

A chest strap is still measuring something fundamentally easier: the electrical activity of the heart.

Apple has to infer the same underlying heart rate from blood-flow changes at the wrist.

But Apple also has a huge amount of raw sensor information available.

So you can think of it this way:

Chest strap

Electrical heartbeat
- beat interval
- HR

Apple Watch

Thousands of optical observations

• motion information
• filtering
• multiple heartbeats → best estimate of HR

And Apple’s approach seems to prioritize getting a good estimate every few seconds rather than producing a noisier number every second just because it can.

So when we say Apple Watch is “continuously measuring HR,” I think the important distinction is:

It is continuously observing the pulse signal.

That does not necessarily mean it is producing a brand-new BPM value every second.

The sensor may be working constantly, while the useful output of all that processing is approximately one high-confidence heart-rate estimate every five seconds.

That distinction finally made the Apple Watch HR behavior make much more sense to me.

I went down a rabbit hole trying to understand one thing about the new Apple Watch Ultra 4 and Series 12.

Apple, says the Ultra 4 and Series 12 measures background heart rate every 5 seconds. That sounds like WHOOP (Whoop does 26x/sec) is sampling your heart 130x more frequently. But after reading Apple's own documentation, I don't think that's the right comparison at all.

Apple is sampling the optical signal much faster than every 5 seconds

Apple's own support document says:

“By flashing its LED lights hundreds of times per second, Apple Watch can calculate the number of times the heart beats each minute.”

Apple also says the optical sensor can compensate for weak signals by increasing both LED brightness and sampling rate.

There is a much higher-frequency optical signal underneath.

Apple basically in their newer sensor samples the data 10x more than Whoop as per their support document. Probably 256hx - 500hz (256 to 500 times per second)

Then where does Apple's “every 5 seconds” number come from?

This is a different layer. Here is the full picture on how it works

RAW / SENSOR LAYER
Apple:
LEDs flash hundreds of times/sec sampling at more than 250hz

        ↓

ALGORITHM
Build pulse waveform
Identify beats
Remove noise
Estimate HR

        ↓

OUTPUT / HEALTH DATA LAYER
Ultra 4:
background HR measurement every ~5 sec

And that's an important distinction.

A higher sampling rate gives the watch a higher-resolution picture of the pulse waveform. During exercise, where arm movement and watch movement create a lot of noise, having more samples gives the algorithm more information to separate the real heartbeat signal from motion artifacts.

For running specifically, motion-artifact rejection is probably more important than the raw sampling-rate number once the sampling rate is already reasonably high. that is a big part of why Apple often does well at rejecting bad HR data during exercise

Apple is not just looking at the raw optical pulse signal. It combines a dense PPG signal with motion information from the watch’s accelerometers/gyroscope, then tries to decide:

>“Does this pattern look like a real heartbeat, or does it look like wrist movement?”

During running, that matters a lot because your arm swing can create a very strong repeating pattern close to your cadence. That is the classic cadence-lock problem.

This is also why Apple sometimes appears to drop HR briefly or show a gap instead of confidently displaying a wrong number. In many cases, rejecting a low-confidence signal is better than locking onto your running cadence and telling you your HR is 170 when it is actually 145.

But I also would NOT say: Apple samples PPG at 200 Hz/300 Hz/etc. Apple doesn't publish a fixed number like that, at least not in the documentation I've found. But it does say hundreds of time per second and it can increase sampling if signal is weak due to motion.

Ultra 4 and Series 12 massively increases HRV measurements too

Apple says Ultra 4 can measure HRV: as often as every five minutes which Apple says is 24x more frequently than Ultra 3.

Apple also specifically says HRV is sampled while you're still, so the actual number of readings will vary depending on your activity.

That part makes sense.

Trying to calculate recovery HRV while your wrist is bouncing around during normal activity would introduce a lot of noise.

So the watch waits for suitable conditions.

Apple now has two concepts of HRV

This is new with the Health Sensing System.

Apple distinguishes between:

Recovery HRV

Designed more around short-term changes related to stress and recovery.

Overall HRV

More useful for broader health trends.

Apple also now puts Recovery HRV into overnight Vitals and compares it against your personal baseline.

And Ultra 4 now has Readiness, which uses things including:

• recent activity
• training load
• vitals
• sleep score

Apple says Readiness updates during the day as new physiological/activity information comes in.

So Apple is clearly moving much deeper into the WHOOP/Garmin recovery territory.

Apple Watch has changed quite a lot with Ultra 4 and Series 12

This is probably the part I find most interesting.

Historically we thought of Apple Watch like this:

excellent workout HR sensor + relatively sparse background physiological measurements

while WHOOP/Garmin were more: continuous physiology + recovery tracking

Ultra 4 and Series 12 changes that equation significantly.

Apple now advertises:

HR: every 5 seconds throughout the day
HRV: as often as every 5 minutes
Raw optical operation: LEDs flashing hundreds of times per second
Vitals: daytime + overnight
Recovery HRV: personal baseline
Readiness: activity + load + vitals + sleep

That starts looking much more like a proper 24/7 physiological monitoring system.

Official sources if anyone wants to verify this rather than taking my word for it:

Apple Support: Monitor your heart rate with Apple Watch
Apple: Apple Watch Ultra 4 announcement
Apple: Health Sensing System details
Apple Watch Ultra 4 product page
WHOOP: WHOOP 5.0 and MG announcement

The new Apple Watch Ultra and Series 12 are really very good. The 5-second heart rate recording and 5-minute HRV are really great, especially with the readiness app and the redesigned health app. In my opinion, it's a very big feature upgrade.

Adding HRV as part of vitals, this was a very great welcome change.

It supports a charge up to 50 hours and, I think, continuous GPS activity for up to 18 hours. Really great! This is what I was waiting for. Given the history of how reliable Apple Watch's sleep tracking and heart rate tracking are, the new sensors just make it more and more reliable.

I'm jumping ship from Garmin.

Edit: reasons

Per 5 sec heart rate sampling

5 minutes hrv sampling

Hrv is now part of vitals app

Redesigned health app

There is a new readiness metrics (which is similar to recovery metrics) the readiness updates realtime whole day and notifies you if you pushed hard and should take rest

The new reengineered sensor also samples hr and hrv during movements which whoop/garmin and other fitness wearable cannot do.

The new sensor makes the HR reading very reliable and accurate very close to a ecg type chest strap

If I configure a custom running workout in the Google Fit app, does it notify me whenever I start that workout during my run?

Specifically, will I get notifications showing things like whether my heart rate and pace are within their target ranges?

Hello,

I have garmin watch and planning to switch to Coros Pace 4, / Pace Pro.

Willing to upgrade to Pace 4 because it has newer hardware and was recently released.

Not sure if i should wait for Pacr pro 2 or something.

Overall i have below questions.

1. HR accuracy, the gen5 in Garmin was really awesome, in my experience it was almost beat to beat with HR strap,. How is the newer sensor in Pace 4 handles it here?

2. Sleep... I dont care about sleep stages, i just want the start and end times to be accurate or near accurate.

3. Do Coros have alarms?

4. I do a lot of medidation and breath work using native garmin app in the watch itself and they are really good, do coros have similar ?

5. Can i create a structured workouts like runs, strength training etc?

6. How good is GPS as compared to garmin?

7. What metrices are captured during sleep?, HRV, RHR, breathing, tempreature?

8. Do coros have an equivalent of garmin's health status?

9. Do coros allow changing the HR zones and does it calculate LTHR?

10. What is coros equivalent of Garmin's Training Readiness

11. What is coros equivalent of Garmin's Body Battery

12. What is coros equivalent of Garmin's Training Load

13. What is coros equivalent of Garmin's Load Focus?, Like garmin gives a nice UI in the app which tells how much anaerobic, aerobic (low and high) i am doing.

14. Do Coros calculate intensity minutes? Like in garmin i can customize what is moderate intensity equivalent to (Z1, 2, 3, 4 etc of HR), and i can set weekly goal as well,

15. Can i customize daily steps goal, and does it calculate the floors climbed? can it be set as goal like garmin?

16. How is the weather app?

17. Breadcrumb nav is fine for me, and if there are built in maps its just good to have, not a need.

18. Does it have a Health snapshot like feature and what health metrics it captures as part of health snapshot?

19. Race predictor?

20. Can i simply add events in calendar, Do coros have a calendar like garmin?

I know its a big list, some items are deal breaker for me, my garmin is getting old, and planning to upgrade, if Coros is match, then i can jump into coros. Let me know what do you think or have any suggestions.

CIRQA is more comfortable than my FR570, but the sensor design still feels poorly thought out
I have been wearing the CIRQA for 3 days now.
It is definitely more comfortable than my FR570 47mm, but it is not 100% comfortable.

The protruding Elevate Gen 4 sensor is not as comfortable as something like Fitbit Air, which has a flatter sensor profile. After wearing CIRQA for the whole day, the protruding sensor leaves a visible mark and a deep indent on my wrist, even when the band is not very tight.
This is not just because it uses the Gen 4 sensor. The FR265 also uses Elevate Gen 4, but the shape of the rear housing is different and feels better on the wrist in my experience. With CIRQA, because the band and body are narrow and slightly curved, the sensor bump feels more concentrated than it does on a watch.

Did the CIRQA product team compare the side profiles and comfort of bands from competitors?
Looking at the side profile of CIRQA, I don’t understand why the body could not have been designed so the back rests more evenly against the skin. Yes, making the bottom flatter might have required making the main body slightly thicker, but perhaps that could also have created space for a larger battery or a shorter body.
Obviously I don’t know their internal engineering limitations, but the current shape feels like a missed opportunity. The device is already 27 x 47 x 9 mm, so this is not an extremely tiny tracker where every millimeter was removed.

IMO Garmin has one of the strongest software feature sets, but the CIRQA hardware feels like Garmin repurposed an existing sensor design instead of designing a flatter sensor package specifically for a screenless band.
Garmin Elevate is Garmin’s proprietary optical HR technology, including its signal-processing algorithms, so it would be unfair to say Garmin does not innovate at all in sensors. My criticism is mainly around the packaging and overall product design.

Things like processors, accelerometers, gyroscopes, GPS chipsets and other components may come from third-party suppliers, but building a wearable is also about antenna design, battery management, waterproofing, sensor placement and how everything feels on the wrist. CIRQA does not feel like Garmin gave enough importance to that last part.
Even from the software side, many of Garmin’s most useful physiological and training metrics came from or were strengthened by Firstbeat Analytics, which Garmin acquired in 2020. There is nothing wrong with acquiring good technology. But after getting that strong analytics foundation, Garmin still had to integrate it properly and present it through a clean app.

Garmin redesigned Connect in 2024 and described it as a simpler and more customized experience. Personally, I still find the app cluttered and less polished than it should be.

Overall, CIRQA is not a bad product. The heart-rate accuracy may be good, the Garmin integration is useful and there is no subscription. But for a $199 screenless band that is specifically meant for 24/7 wear, comfort and industrial design should have been the first priority.
For me, the protruding sensor makes it hard to justify keeping it.

EDIT: final thoughts

What I have learned from using CIRQA is that Garmin’s experience making smartwatches did not automatically translate into a perfectly designed screenless wrist band.
CIRQA is Garmin’s first screen-free smart band, although

Garmin previously released the Index Sleep Monitor armband. It appears Garmin created CIRQA by combining existing Garmin sensors, software and health metrics into a new screenless form factor, rather than designing every component from scratch.

That approach probably reduced development risk and helped Garmin release a reliable product, but the result also feels like a conservative first-generation design.
I can already see several areas that could improve in the next version. The module could be shorter, the optical sensor could have a flatter profile, and the band attachment could potentially use a more compact removable-pebble design.

I would not necessarily call CIRQA a bad product. The software integration, battery life and lack of subscription are valuable. But from a comfort and industrial-design perspective, it still feels like a first iteration.
My takeaway is not that nobody should buy a first-generation product. It is that you should only keep one when the current design already works for you, rather than buying based on what the second generation might eventually become.

today I did a run with FR570 and I wear CIRQA during sleep and daily wear.

I was not wearing CIRQA during activity

When I checked my garmin connect app, it said 0 hours recovery, and my Training Readiness was unaffected.

When I checked with Garmin support, this is what they said exactly.

>Training readiness is affected by both the activities and rest. Your data may not be completely accurate if you are recording across multiple devices.

This doesn't make any sense, especially when I updated FR570 as my training device and CIRQA as Primary device in Connect app

Have anyone else got the same issue?

So today I did a threshold run. the total session was about 40mins.

CIRQA on the left wrist and FR570+HRM200 on right wrist.

I was really impressed by the HR accuracy here... (TIP: Just make sure when you are starting an activity in the CIRQA, ensure you make it a bit right (not too tight), so that the sensor has some nice contact with skin.)

I didnt expect this kind of accuracy, but then it would be easier for any optical sensor because it was a simple run and not rapidly changing HR.

I will continue testing it more and share results.

https://preview.redd.it/hszci3d5ikgh1.png?width=2360&format=png&auto=webp&s=647aafdd84ade788ec5bea88600f9e9120ce2552

https://preview.redd.it/n8ay6k0bikgh1.png?width=2360&format=png&auto=webp&s=293092bf736d2071d2eb2a7458627cb20b30184d

I recieved the CIRQA today and updated it to be primary wearable and FR570 secondary.

I climbed several times floors today, normally the FR570 would record the floors climbed but CIRQA doesn't.

I guess this is the missing sensor in CIRQA?

This is a follow-up to my previous Garmin Connect HAR analysis.

In Part 1, I found production code for a distinct Garmin “screenless device,” including device capability fields, primary-wearable logic and screenless workout support. However, the word CIRQA never appeared in the HAR.

This time, I went back to the original CIRQA leak image. Remember this image?

https://preview.redd.it/5cehlxgql3eh1.png?width=1080&format=png&auto=webp&s=fd77d11a4d4219677d38db4dab829a411a8b48f0

The important part of that image is not just the product name. It contains a specific Garmin part number:

CIRQA Smart Band
Black, L/XL
PART NUMBER 010-04675-00

It also shows:

Sizes: S/M and L/XL
Colors: Black and French Gray

A screenshot can obviously be edited, so I wanted to answer a simpler question:

Does Garmin’s own backend recognize 010-04675-00**, or was the number invented for the leak?**

Garmin’s pricing system recognizes it

This public Garmin endpoint can be opened without logging in:

https://buy.garmin.com/pricing-proxy-services/countries/US/skus/010-04675-00/price

{
  "listPrice": {
    "formattedPrice": "$199.99 USD",
    "price": 199.99
  },
  "partNumber": "010-04675-00",
  "salePrice": null
}

That response is coming directly from buy.garmin.com. It is not a Google cache, retailer listing or archived screenshot.

The same SKU also has localized prices in Garmin’s backend:

United States: $199.99 USD
Canada: $279.99 CAD
Mexico: $4,099 MXN

I also tested a made-up Garmin-style part number as a control:

010-99999-99

The pricing endpoint returned no product data for the fake number. So this is not an endpoint that supplies a generic price for anything that looks like a Garmin SKU.

There is also a second recognized part number:

010-04675-10

Garmin returns the same $199.99 US price for that SKU. A deleted Garmin support link from the original leak also used 010-04675-10.

This strongly suggests multiple CIRQA variants, although I cannot determine which size or color -10 represents.

Garmin also has an accessory record for it

This second official endpoint accepts the leaked part number:

https://buy.garmin.com/ecommerceServices/productCompatibility/en-US/US/p/010-04675-00/accessories

It currently returns nine associated items, including:

Garmin Connect App
Garmin Connect+ Monthly Plan
Garmin Connect+ Annual Plan
USB-C charging/data cables
USB-A charging/data cables
Garmin power adapters

The same request using my fake 010-99999-99 part number returns an empty list.

That means Garmin has a product-specific accessory/compatibility record attached to 010-04675-00.

One important caveat: the presence of Garmin Connect+ in this list does not prove that CIRQA will require a subscription. Garmin lists Connect+ alongside several existing products as well. It only proves that Garmin associated it with this product record.

The deleted product page had a real Garmin product ID

The original Garmin product URL was:

https://www.garmin.com/en-US/p/1989182/

It now returns 404, but the URL gives us the internal retail product ID:

1989182

Garmin still returns product-association data when that ID is used here:

https://www.garmin.com/en-US/p/api/1989182/compatibleProducts

A random product ID returns an empty list. I would not read too much into the particular products returned because this could be cross-selling data rather than functional compatibility. The relevant point is that Garmin still has backend data associated with the deleted product ID.

Combining this with Part 1 of my post which analyzed the HAR file gives us two independent sides of the same story:

Garmin Connect contains production support for a screenless wearable
+
The CIRQA leak contains a part number that is still active in Garmin’s backend

The remaining missing link is a Garmin response explicitly connecting the two, such as:

010-04675-00 → screenlessDeviceCapable: true

We do not have that response yet, so I would not claim that every CIRQA specification or launch detail is confirmed.

However, the leaked screenshot using a real, unreleased Garmin SKU—with matching pricing, accessory records, support records and a Garmin product ID—makes it very difficult to dismiss CIRQA as a fake.

My balanced conclusion is:

CIRQA is, or at least was, a real Garmin product under development.

I used GPT-5.6 Sol High to help trace the endpoint patterns, but I manually opened and verified the responses above. The Garmin URLs are public, so anyone can check the results directly.

Note: If you use any findings from this analysis in an article, blog post or video, please credit the research and include a link to this Reddit post.

This is a follow-up to my previous Garmin Connect HAR analysis.

In Part 1, I found production code for a distinct Garmin “screenless device,” including device capability fields, primary-wearable logic and screenless workout support. However, the word CIRQA never appeared in the HAR.

This time, I went back to the original CIRQA leak image. Remember this image?

https://preview.redd.it/b1skbwwyh3eh1.png?width=1920&format=png&auto=webp&s=e2a9c23cf25c3cb82e74befd945be18a86a5d8fd

The important part of that image is not just the product name. It contains a specific Garmin part number:

CIRQA Smart Band
Black, L/XL
PART NUMBER 010-04675-00

It also shows:

Sizes: S/M and L/XL
Colors: Black and French Gray

A screenshot can obviously be edited, so I wanted to answer a simpler question:

Does Garmin’s own backend recognize 010-04675-00, or was the number invented for the leak?

Garmin’s pricing system recognizes it

This public Garmin endpoint can be opened without logging in:

https://buy.garmin.com/pricing-proxy-services/countries/US/skus/010-04675-00/price

{
  "listPrice": {
    "formattedPrice": "$199.99 USD",
    "price": 199.99
  },
  "partNumber": "010-04675-00",
  "salePrice": null
}

That response is coming directly from buy.garmin.com. It is not a Google cache, retailer listing or archived screenshot.

The same SKU also has localized prices in Garmin’s backend:

United States: $199.99 USD
Canada: $279.99 CAD
Mexico: $4,099 MXN

I also tested a made-up Garmin-style part number as a control:

010-99999-99

The pricing endpoint returned no product data for the fake number. So this is not an endpoint that supplies a generic price for anything that looks like a Garmin SKU.

There is also a second recognized part number:

010-04675-10

Garmin returns the same $199.99 US price for that SKU. A deleted Garmin support link from the original leak also used 010-04675-10.

This strongly suggests multiple CIRQA variants, although I cannot determine which size or color -10 represents.

Garmin also has an accessory record for it

This second official endpoint accepts the leaked part number:

https://buy.garmin.com/ecommerceServices/productCompatibility/en-US/US/p/010-04675-00/accessories

It currently returns nine associated items, including:

Garmin Connect App
Garmin Connect+ Monthly Plan
Garmin Connect+ Annual Plan
USB-C charging/data cables
USB-A charging/data cables
Garmin power adapters

The same request using my fake 010-99999-99 part number returns an empty list.

That means Garmin has a product-specific accessory/compatibility record attached to 010-04675-00.

One important caveat: the presence of Garmin Connect+ in this list does not prove that CIRQA will require a subscription. Garmin lists Connect+ alongside several existing products as well. It only proves that Garmin associated it with this product record.

The deleted product page had a real Garmin product ID

The original Garmin product URL was:

https://www.garmin.com/en-US/p/1989182/

It now returns 404, but the URL gives us the internal retail product ID:

1989182

Garmin still returns product-association data when that ID is used here:

https://www.garmin.com/en-US/p/api/1989182/compatibleProducts

A random product ID returns an empty list. I would not read too much into the particular products returned because this could be cross-selling data rather than functional compatibility. The relevant point is that Garmin still has backend data associated with the deleted product ID.

Combining this with Part 1 of my post which analyzed the HAR file gives us two independent sides of the same story:

Garmin Connect contains production support for a screenless wearable
+
The CIRQA leak contains a part number that is still active in Garmin’s backend

The remaining missing link is a Garmin response explicitly connecting the two, such as:

010-04675-00 → screenlessDeviceCapable: true

We do not have that response yet, so I would not claim that every CIRQA specification or launch detail is confirmed.

However, the leaked screenshot using a real, unreleased Garmin SKU—with matching pricing, accessory records, support records and a Garmin product ID—makes it very difficult to dismiss CIRQA as a fake.

My balanced conclusion is:

CIRQA is, or at least was, a real Garmin product under development.

I used GPT-5.6 Sol High to help trace the endpoint patterns, but I manually opened and verified the responses above. The Garmin URLs are public, so anyone can check the results directly.

>
Note: If you use any findings from this analysis in an article, blog post or video, please credit the research and include a link to this Reddit post.

I know this is not monitored by Garmin, I have suggested my idea to them, but just incase their PO or a dev sees this post then maybe they can ask their product and engg team to review it and fix or implement below suggestions.

1. Make Sleep Coach a dedicated glance

Sleep Coach already tells you two useful things:

• How much sleep you need
• What time you should go to bed to hit that target

But right now this is buried inside the Sleep widget, it's not always visible. Please make Sleep Coach its own dedicated glance in Garmin Connect, so users can check their sleep recommendation without digging through the app.

2. Training Readiness needs to be more sensitive to recent changes

Training Readiness is useful, but it is training-focused, not health-focused, and it reacts too slowly.

Problem: HR-V Status is based on a 7-day rolling average. So if your HR-V dropped sharply last night, the 7-day average can still look "normal" and Training Readiness stays high, even though you feel fatigued. This could be result of a post race or intense training fatique, maybe less sleep because your baby woke you up at night or you were travelling and you didnt sleep in last 24 hours.

Result: the metric says "you're ready to train," but your body is actually telling a different story. The algorithm should give more weight to recent (last 1–2 nights) deviation from your personal baseline, not just the smoothed average.

3. New metric: a "Recovery Strain" score (like Whoop's Strain)

Being "in range" isn't the same as being "fine." Suggest a new metric that:

• Looks at how much your overnight vitals (HR V, resting HR, respiration, etc.) dropped compared to your own baseline not just whether they're within a generic healthy range
• Converts that drop into a single combined score or graph, similar to Whoop's Strain
• Could be shown as a meter/gauge so it's easy to read at a glance

Purpose: catch the situation where every individual metric is technically "in range," but the combined overnight drop is large enough that you're actually fatigued — something Training Readiness currently misses.

These are honestly small wins, its pretty easy to use latest Anthropic models to fix these metrics..

Many elite athletes wear whoop, i am seeing many instagram, fitness influencers, YT tech reviewers who are not even 10% athlete are wearing Whoop.

In my company i am seeing many people who do a 9-5 desk Job, who cannot complete a 5K run.... wears the Whoop.

Before pulling the trigger i have below thoughts based on my research with Claude Opus4.8/Fable5.

1. Stress is physiological, you see a tiger you are stressed,.... flight or fight response you got from evolution. its normal, you dont need a tracker to tell you you are stressed so you can start running from tiger. Well, IF YOU TRAIN LIKE ATHLETE, your body really goes into stress, your HRV and HR gets messed up, but that is physiological stress and not mental stress. A good sleep will bump up recovery.
2. I was about to get sick and my Whoop told me by showing that my temp, hrv, hr was falling apart days before... So what you would have done?, your body is gonna get sick eventually. I dont get what people are trying to claim by saying this...

So to summarize, there are some niche groups who needs those levels of strain and recovery. you would know it by now and i dont want to call out each and every category of training/exercise.

if you remove those niche group out of it, a cheap amazfit helio strap is more than sufficient for you folks. these who are using whoop have never used a fitness tracker before.

Its your money and i dont care if you like to burn it.

Note : this is how i make decisions, put up my raw thoughts and i like people to reflect and prove me false and it lets me know if that idea is right or not, this is how i come to know real feedback.

I GOT MY ANSWER:
I expected some mature response but looks like i will abandon my plan to buy whoop for now. All the whoop engagement i see online via some athletes or others seems like paid and sponsorship and has very less to do with features that whoop provides.