I have been using XMB to drive most of my training for … a long time now, it seems.
It only just occurred to me that I have not been doing any of the sorts of intervals I like the least: Threshold (Z4) and VO2Max (Z5).
XMB always suggests about ten ~30 second Anaerobic (Z6) intervals an hour. Recovery is usually Endurance (Z2). Steady state is Tempo (Z3). I occasionally push the intervals up to Neuromuscular (Z7).
Shouldn’t I be doing some Sweet Spot training? Or is that old school?
Check this out:
I rarely glance at the Z1-7 charts on Garmin Connect but I understand why others monitor those numbers especially if that’s what they were brought up on. 
I’ve kicked the 5-7 Zone habit some time ago. Polarized or Pyramidal work for me. YMMV
What program type are you on?
XMB doesn’t prevent you from riding intervals however you want.
The default behavior is shorter intervals to get the job done in an efficient manner, but you can ride longer intervals at lower watts than the target or vice versa.
You can also manage the Difficulty level achieved by leveraging the Challenge Level gauge and Recovery timer.
For example, yesterday my goal was 180 XSS in a ratio to target GC Specialist.
XMB estimate was 53 intervals in 2-1/2 hours. I whittled that down to 40 intervals in just over 2 hours by riding over-target for many of the intervals while pushing CL up to 6 towards the end.
Then I filled remaining Low pts plus added another 30 pts of Low by mulling about town riding on the greenway.
Result was Difficult Mixed Rouleur at 3.5 diamonds (100 Difficulty score) which is what I was going for – a shorter Focus than GC Specialist at a similar Difficulty level.
I’ve also run XMB in parallel with a “normal” workout to see how it reacts. The buckets get filled either way. Seems reasonable to me that strain is strain regardless how you implement it (within reason).
I avoid old-school SS blocks but if you are time-crunched or compete and find that of value for riding in the peloton, you can certainly add that into the mix.
I also avoid traditional VO2max 3x5s and much prefer Ronnestad-style efforts – something else I can do while monitoring XMB to fill the buckets.
OTOH with traffic and terrain outdoors it’s much easier to stick with XMB guidelines to fill the buckets.
I only ride longer blocks indoors or on particular routes where I’m chasing down a segment PR.
Something else you can do with XMB is mix and match. 
I may start out one way and finish up another.
Here’s a relevant excerpt from that article:
“This principle has come into dispute recently with research published by Dr. Stephen Seiler described here by our very own Dr. Stephen Cheung: Polarized Training and Stephen Seiler. What the polarized method advocates is to avoid sweetspot training rather than intentionally train there, the complete opposite of what has been advocated! It is argued that SST doesn’t provide the training value and in fact can compromise your training. It is termed the black hole of training where athletes make the mistake of training too hard during easy sessions and training too easy during hard sessions .”
There’s an issue with the concept of XSS (and Magic Buckets) if one considers Seiler’s studies to be valid: In Xert, it doesn’t matter which intensity zone you train in below TP—as long as you accumulate enough XSS. Training below LTP just takes longer; training between LTP and TP accumulates XSS faster, but it all goes into the same bucket.
However, research seems to suggest that it does not have the same physiological effect whether you train X minutes < LTP or e × X minutes > LTP but < TP (with a suitable e < 1). Xert, on the other hand—unless I’ve misunderstood—treats these as having an identical impact.
Or is the subtle difference perhaps reflected in the different recovery times that result from this?
Since I’m still new to Xert (but a total data nerd), I’d really appreciate any corrections to my understanding—thanks!
Thank you. That article is really helpful.
I am on continuous maintenance, climber focus. I do love the freedom of XMB, particularly as I find it harder to successfully complete a structured workout these days. Garmin often tells me I’m unproductive, because I don’t do enough high intensity aerobic work, but I mainly use Garmin just for L/R balance. b
You are free to apply your own nuance to the 3 XSS values, in the same way that a coach will interpret the benefits of TSS from different rides in their own way. The key point to consider is that with TSS, you have to treat rides with the same TSS differently in order to use it correctly. It requires intervention and experience to make it work properly.
WIth XSS, you can add some nuance if you prefer doing Low XSS differently and you think it will provide different results. But the differences are really only incremental, if indeed they exist. They certainly don’t show up readily in terms of performance. Low TL (calculated from Low XSS) and TP are colinear, hence using Low XSS without any nuance is what will correspond with your TP. If you feel you respond differently between to some forms of Low XSS vs. others, then apply it indirectly to your training. Nothing is stopping you from doing this! 
There are 2 other thoughts that crossed my mind when thinking about this thread, which I’ll share below:
If your goal is to accumulate High & Peak Strain (XSS), it makes sense to train approximately where they’re the largest contributing energy systems, which is ~10s to ~3min power (or Zone 6/7, if you follow that methodology). But since %FTP scales so poorly for athletes in this part of the power-duration curve, other systems typically don’t bother with prescribing it much. A good coach will likely include efforts in this region, if you’re able to work with one.
Now you can accumulate High & even some Peak XSS by doing intervals just over threshold, but the rate is much, much slower. In fact, if you have relatively small Peak XSS target (<1.0) and a relatively small dose of High XSS (<5), you’ll likely see an interval target power not too far above threshold. But if you need to be accumulating 20 High XSS, you need to be doing these at a very high intensity, else you’ll be massively over-shooting your Low target just to achieve the High XSS needed.
I also speculate that many athletes overlook the “recovery” interval aspect. After all, it’s a recovery interval, right? But Magic Bucket recovery intervals do 2 things: 1) they do allow for some recovery of your MPA & keep the overall difficulty manageable, but 2) they are also a great way to accumulate Low strain (XSS). So you could do 2x20 or 4x8 intervals near threshold & those would accumulate lots of Low XSS, but you can also accumulate a hefty dose of Low XSS by doing some hard surges that accumulate High/Peak XSS and pull MPA down and then hold an endurance pace effort with that fatigue. In this way, doing Z2/Z3 efforts under fatigue can be nearly as potent as doing plain Z4/5 intervals without that initial fatigue.
I was looking for such post because I have a similar feeling. I would add that when you have a time trialist profile, you don’t have much high xss to do in the proposed planning. It is quite easy for me to fill the high xss buckets with these short 30 s efforts, I tried to do the same with just above ftp efforts, but as it as been said, the filling rate is so slow, I am exhausting myself for 5 min and more just to get what I can easily obtain in 30 seconds.
I understand that high effort are beneficial even for time trialist profiles, but I cannot help thinking that doing long intervals at just above ftp are needed for climber / time trialist profiles. Not every day, but regularly.To be brief: why don’t we have medium buckets?
I was waiting magic buckets for long in ebc, but now I am quite disappointed as it cannot allow me to know how much strain I have done in these medium efforts I would like to train.
I’m not a sports scientist or Xert expert so I’m just repeating something I I think I’ve read. I think that you can adjust your Polarization settings to get more of what you’re looking for in the zones you’re talking about. It’s also possible that magic buckets aren’t the best tool for training time trialing…