those are the values shown yesterday within the buckets data field (Edge 830, latest DF version).
Screenshot from Garmin connect, where those values also get reported:
In contrast, xert online shows those values:
So basically they all (high, low, peak xss) slightly differ: 92 vs 90, 14.8 vs 15.2, 2.5 vs 2.6
This is a consistent observation during my last rides and it makes training planning and execution somehow unprecise.
Is that a bug in one of the two (bucket DF, xert)?
I have seen similar discrepancies. I suspect that XSS buckets looks at your adapted forecast, even if you haven’t pressed “Adapt Forecast” on the website. But I am just speculating.
Second by second data analysis is more robust on XO when the activity is uploaded to the server.
I don’t think a Connect IQ data field can match the same level.
Another factor is signature adjustment. A BT during the ride would significantly change the result compared to original target values. A minor daily adjustment can also occur especially after a long and/or hard ride. A point or two would make a difference in the final analysis.
I don’t think the numbers need to be precise for training purposes especially when riding outdoors.
I think that is the reason, signature is updated before xss are acounted for the activity, while the bucket field bases all calculations on the initial signature from before the ride.
@xertedbrain Is that the reason? Anything you could do about that discrepancy?
I agree that the effect is minor in the grand scheme of things, but that´s not how I see it while returning from a 1…2h interval session where I just summoned all remaining will power to crank out the last 0.5 high-xss, dreaming about kicking that bike in the ditch and switching to bowling for the rest of my existance, just to learn that xert substracts 0.5 xss from what I suffered for! 
Today the discrepancy was more extrem.
Buckets/Garmin connect:
Xert:
1.2 difference in high-xss, quite annoying 
For the workout portion of the ride do you roughly follow XATA or XFAI advice with Target Watts in mind along with Focus and Specificity at the suggested Difficulty level?
Then concentrate on filling any bucket showing a point gap before ending the ride?
Not really following advises as that´s too static for my daily life.
But well, as said, it´s more a motivational thing than something which really impacts my training or results.
I too have seen a difference between post-ride numbers and the bucket numbers. For me it was obvious because I proudly met my High goal of 12.7 (at 13.0) on my bucket only to see it as 12.5 later on-line. Yes, not very important (4% difference), but a bit annoying.
Were the Moderate Mixed Pursuiter and Difficult Pure Breakaway Specialist activities strictly random results based how you rode the route that day or did you have specific HIT workouts in mind?
I, too, have noticed a difference between XSS Buckets and what is calculated online. It started roughly 3 weeks ago for me. I don’t recall much of a difference when doing my workouts on the road. When doing workouts off road, it is a different story.
I have gotten breakthroughs on my Garmin, but it hasn’t registered so much as a near-breakthrough online. Yesterday I was over 100% on all my buckets while still maintaining green checkmarks, but I was under the XSS when I loaded the workout.
I am making sure the app loads my most current signature when starting the workout. I don’t know. My Peak XSS was high enough I thought it was going to give me the red X, but online it shows under the forecast. That’s a pretty big difference. The same was true for High XSS, although is was a lower percentage than the Peak XSS. I’m not worrying about the Low XSS really, but it was a tad over 100%, but shows roughly 10% under the forecast online. (I wish I had taken a screen shot of the Garmin).
To be honest, it is close enough. I just have to over perform on the XSS Buckets app in order to hit close to targets when I am doing CX training. As mentioned before, it has been fine for me out on the road.
BTW, I love XSS Buckets. It is awesome and one of the best on-bike training tools I’ve encountered. Thank you for all the enhancements this year.
-g
PS: The workout I referred to from yesterday wasn’t structured. It was performed with friends doing hot laps at a local CX friendly park. The workout the prior Thursday, Sept. 12th was the same. That workout also showed over 100% on all but low XSS and showed a breakthrough on my Garmin. I had a Red X on both Peak and High XSS from over shooting the targets. Online it is a good bit lower. Have a wonderful weekend.
Agreed. Doesn’t need to be exact to meet the goal for the day.
I think CX and MTB activities have a similar strain score issue as discussed here –
XSS for MTB vs Road - General - Xert Community Forum (xertonline.com)
BTW I am just a fellow Xerter who contributed to the XSS data field challenge.
Others more knowledgeable than me made it happen. 