2 Autoregulation methods to improve your training progress

Autoregulation, periodization, deloading, recovery, fatigue… these concepts are thrown around a lot without specification of what they mean, often because the person using these terms only has a vague notion him- or herself and is mainly using the term to sound sophisticated. To cut through the confusion, in this article I’ll share 2 concrete and well defined autoregulation methods you can use to improve your training progress.


Before we get to the actual methods, we first have to define ‘autoregulation’.


Note: This article consists of excerpts of the online Bayesian PT Course Certification Program. While it refers to other sections of the course, you should be able to understand the article fully on its own. Beware that this article, being about training programming, is relatively dense reading specifically for serious trainees that engage in structured strength training. If you work out doing more or less random stuff every time you’re in the gym, this article is probably too advanced for you.


What is autoregulation?

True autoregulation is a form of programming that automatically regulates a certain process, like fatigue. You can think of autoregulation as a system or a ruleset instead of a fixed prescription.


For example, a non-autoregulated program might have the following workout: ‘Perform 5 sets of 8 reps with 75% of your 1 RM in the squat with a 2 minute rest interval’. An autoregulated variant of this workout is: ‘Perform sets of 8 reps with 75% of your 1 RM in the squat with a 2 minute rest interval until your repetition speed decreases to the point that you experience the sticking point’. The amount of sets and thereby the training volume is hereby autoregulated by the person’s work capacity.

(See the course topic on repetition tempo for the usefulness of using bar speed to monitor proximity to failure.)


Other examples of true autoregulation have already been discussed in this course, such as the muscle-specific hypertrophy method (autoregulates training program volume on a muscle-specific basis) and autoregulated rest intervals (autoregulates inter-set rest intervals).


Autoregulation is an incredibly useful programming method, because it automatically individualizes the training program. As you’ve seen in the course topic on individualized program design, there is huge interindividual variability in many aspects of fitness. People differ significantly in how many reps they can do at a certain training intensity, how much rest they need for full recovery, how fast they can gain strength, etc. Autoregulating these factors in the program is thus preferable to the traditional fixed program prescriptions, which require making arbitrary decisions, relying on averages or predicting future performance with improbable accuracy.


For example, many one-size-fits-all Powerlifting programs base the training of the next month on the person’s current 1 RMs. On a certain day, the program may call for 6 reps at 85% of 1 RM. Most people can’t do that with a true 85% intensity, but because this is based on last month’s strength, this is the program’s way of planning progression. Now, what if on this particular day you can’t nearly reach 6 reps because your diet has been less than ideal? And what if yesterday you could have done it but the program called for a light workout?

Many coaches that have tested these programs are aware of these problems, so they often plan for little progression to ensure almost everyone can achieve the program’s planned rate of progress. However, this inherently also requires that almost everyone will make less progress than they could have on a more individualized, autoregulated program.


Autoregulation vs. cybernetic periodization

A related concept to autoregulation is cybernetic periodization. Although the definition of cybernetic periodization has become blurred over time, many people use this term interchangeably with autoregulation in meaning that it’s a form of flexible periodization that allows program modifications based on how you feel during any workout. For example, when feeling poorly going into a workout, you may opt to make this a light training day even though it was planned as a heavy workout.


This form of cybernetic periodization is, however, not true autoregulation by definition, since nothing is being autoregulated and in fact, very conscious decisions are required.


Distinguishing between true autoregulation and cybernetic periodization and making ad hoc, on the fly programming decisions based on how you feel is important. True autoregulation is a highly useful programming concept. If you can autoregulate any process, that is almost always preferable over trying to plan it in advance or making arbitrary decisions because it automatically individualizes the program. However, the usefulness of cybernetic periodization is far more debatable. As you’ve learned in the course topics on the different kinds of fatigue and willpower, the fundamental idea that your subjective feelings truly represent your physical readiness to perform or level of recovernedness is flawed in several ways.


Now that we’ve defined autoregulation and why it’s useful, we can get to the actual methods.


Autoregulatory volume training (AVT)

AVT is a programming method in which you only plan the weight and number of repetitions of the first set, e.g. 260 pounds for 8 reps, for a multiple set exercise. This first set is your ‘benchmark set’. The subsequent sets are ‘volume sets’. They’re performed with the same weight and proximity to failure, but you do not plan how many repetitions you’re going to do in advance. In fact, you don’t even need to count your reps in these sets.


The purpose of AVT is to autoregulate training volume based on the difficulty of the first set. The more neuromuscular fatigue the first set induces, the lower the total volume for that session will be. If the first set had you bust out your tomato face to grind through the sticking point in the squat and feeling so lightheaded you wondered how you even managed to rack the bar afterwards, then you will naturally perform fewer reps in the subsequent sets. In contrast, if the first set had you progress as planned with more left in the tank, then you will naturally perform more reps in the subsequent sets.


The result is that AVT normalizes the training stress over time to prevent overreaching yet also ensure a sufficient training stimulus.


Importantly, AVT is implemented at an exercise-specific level, which is highly preferable to the more common regulation of training volume at the whole body level. As you’ve learned in the course topics on the physiology of strength training adaptations, structural balance theory and what neuromuscular fatigue really is, muscle fatigue is largely a local process and it makes little sense to take it easy on your biceps curls today because your quads haven’t recovered yet.


Here’s a simple example of AVT. Let’s say we’ve got a novice lifter who can still put on 5 pounds on the bar every squat session and perform 8 reps with that. He squats on Monday (because screw national bench press day) and Friday (because screw ‘early weekend’) with a volume of 3 sets. Last Friday he performed 200 pounds x 8, 7, 6 reps. Next Monday he was feeling frisky because he got laid for the first time in months and had slept like a baby that weekend. So he was very well recovered, he hit 205 pounds x 8 reps with ease and he managed to do 8 reps in the 2 subsequent sets as well. On Friday that week he was mildly sleep deprived and more stressed from the work week, so he hadn’t recovered that well and it took everything he had to hit 210 pounds x 8 reps. In the subsequent sets he only managed 3 reps, which is fine, because the near-failure squat set induced a lot of fatigue already.


Another benefit of AVT is that individuals learn to mentally break free from the constraints of performing each set while counting the repetitions, mentally as well as physically. This allows you to focus more on your exercise technique and tends to greatly reduce the mental stress of feeling like you “have to perform X reps or I’ll have failed today” every set.


Physically, AVT increases force production throughout the set. During exercise with a defined endpoint, like a certain number of repetitions that you want to achieve, you will subconsciously pace yourself by holding back during the early part of the exercise. This is good for endurance and the achievement of the goal you have in mind in terms of quantity, but when you don’t need to hit any specific performance goal, it can detract from the quality of the exercise in terms of muscle activation, exercise technique and force production.


One obvious caveat to the use of AVT is that it only works for serious, motivated strength trainees that have no problem pushing themselves and won’t see not counting their reps as an excuse to slack off. If your idea of training intensely is experiencing difficulty reading Shape magazine during your leg extension set, AVT is not for you.


In sum, AVT normalizes the training stress of your program over time by autoregulating the training volume of your subsequent ‘volume sets’ based on the difficulty of your first ‘benchmark set’. Not having rep targets during your volume sets allows you to focus on your exercise technique, improves force production and can help reduce performance anxiety to make your workouts more enjoyable.


Reactive deloading

Sometimes, AVT is not enough to prevent overreaching from occurring. This is when deloads are applicable. A deload is a reduction in weight to reduce the training stress. Sometimes the word deload is also used to refer to a more general reduction in training stress, like a reduction in training volume by reducing the amount of training sets.


Traditional deloading

A traditional implementation of deloading is to take each 4th week of your program off from the gym or to reduce the weight in this week. This is commonly seen in Powerlifting programs employing block periodization. Traditional deloading methods vary in the degree of deloading, ranging from no training at all to only a reduction in the number of sets, and in the duration of deloading, ranging from single workouts to a full week. But by and large, they have 2 aspects in common: the deloads are arbitrary and proactive.


The arbitrariness and proactivity of this type of deloading is that it’s planned in advance at a set date or time in the program. At best, it’s an educated guess about when overreaching is likely to occur, but in practice it often just comes down to one-size-fits-all programming in an attempt to make the program look fancy and sophisticated without more than abstract theory as its rationale. You can’t accuratley predict in advance when an individual will experience higher stress in their life, sleep less well or deviate from their diet, so overreaching in any program can occur at many different time points for different people, if it occurs at all.


Autoregulated deloading: reactive deloading

Autoregulation can solve the probem of the interindividual variability in when deloads are needed. The Bayesian Bodybuilding method employs an autoregulated form of deloading called reactive deloading. As the name suggests, a reactive deload is not scheduled ahead of time. Just like AVT, reactive deloading is only applied to the affected exercise(s) in a single training session.


Specifically, a reactive deload is implemented whenever a trainee does not progress as planned for that exercise and it is deemed likely that this was due overreaching, not an external factor that reduces performance but not recovery status, like simple lack of focus during the set or a circadian rhythm disturbance.


Several factors determine whether reactive deloading is warranted for an exercise in your training program. (See the course topic on exercise selection for more detailed definitions of these criteria).

  • Compoundedness: The more muscle mass and the more joints involved in an exercise, the higher the total and central neural stress and thus the higher the potential need for deloads. Isolation exercises generally don’t require much deloading.
  • Terminal consistency: Exercises that score poorly on this principle have an inherently higher variability in performance. As such, there is a decreased probability that lack of progress is due to overreaching. Therefore, reactive deloads are not needed as much.
  • Microloadability: The larger the increment, the less likely the need for reactive deloading is. If the increment is too large a jump in weight to realistically achieve on a regular basis, reactive deloading will cause you to enter a cycle of perpetual deloading and result in undertraining.
  • Recovery capacity variability: The more variable someone’s rate of recovery, the greater the need for reactive deloading. If someone has a very irregular recovery capacity due to, for example, a variable sleep pattern or circadian rhythm, due to poor diet adherence or due to concurrent sport-specific training, there is a greater need for reactive deloading. Reactive deloading will then take care of underrecovery. For someone with a very stable lifestyle and good recovery capacity, it is more likely that a plateau signals the need for more fundamental program change than a mere reactive deload.


Based on the above factors, bilateral deadlift exercises usually benefit from reactive deloading, since they have excellent microloadability and terminal consistency and they’re very compounded and neurally taxing. An example of an exercise that rarely requires a reactive deload is the delt lateral raise. It generally has abysmal microloadability, it’s an isolation exercise (sort of), it induces very little muscle damage or central fatigue and it has mediocre terminal consistency.


When you’ve established that reactive deloading is useful for an exercise in your program and a plateau occurs in any workout, you implement a reactive deload as follows. You replace your remaining sets with low rep, explosive technique work: 1 to 5 reps per set at 60 – 70% of 1 RM (roughly equal to a weight you could do 12-20 reps with). This type of speed work allows you to reach high muscle activation levels, perform a decent amount of work and work on your technique while only inducing a minimal amount of neuromuscular fatigue. It’s important to realize that to reap these benefits without the cost of high further fatigue, it has to stay speed work. If your movement velocity decreases noticeably at all during any set, you are going far too heavy and only digging yourself deeper into your recovery hole.


In the presence of more severe fatigue, it is better to deload reactively by 100%, i.e. by skipping all subsequent sets altogether. Speed work is appropriate if you only júst missed the last rep needed to progress as planned or you feel like you could have hit your planned number of reps if your technique was a bit better. When you didn’t come close to your planned performance, when your trained muscles were still extremely sore or when you experienced pain, speed work may still be too much and you’re better off just moving on to the next exercise.


Let’s look at 2 examples. Say an advanced bodybuilder is planned to perform 350 pounds for 5 reps in the squat this workout, which would be a new best in the program. He’s already implementing AVT and undulating periodization with a training volume of 4 sets and progress is generally consistent, so reactive deloading is implemented for the squat. First work set, he manages only 4 instead of 5 reps. Since he likely hadn’t recovered sufficiently yet, he reactively deloads to prevent overreaching: for his remaining sets he drops the weight to 250 pounds and performs 3 more sets of 3 as speed work.


Example 2: Say that same bodybuilder does Bayesian flys later in the same training session. Here too he doesn’t progress as planned, but since flys have an inherently higher variance in their performance and they’re not a very centrally taxing exercise, he doesn’t implement a reactive deload and performs his remaining sets as planned (possibly with AVT).


In sum, reactive deloading allows you to program deloads in your program in a systematic but individualized manner so that you reduce your training stress only when and if needed and only for the muscle groups that actually need it, in contrast to proactively and arbitrarily scheduling a whole-body deload in your program when you guesstimate it may be needed.



Interested in more information like this? This article is an excerpt of the online Bayesian PT Course Certification Program. The course contains similar evidence-based information about everything related to muscle growth and fat loss.




  1. Hey Menno, excellent article.

    How about functional overreaching? From my understanding, that’s utilized on an exercise-level during the AVT and Reactive Deloading methods, instead of on a “full-body fatigue” level. You don’t call for a full-body deload ever??

    Also, if you could ‘time’ almost all of your big compound lifts to functionally overreach in week 5 (by using the AVR-method, and consistently adding sets per exercise on a weekly basis), wouldn’t a full-body 1-week deload then be warranted for overall joint-health and central nervous system fatigue? Or is there no such thing as those, and do joint problems and fatigue happen at the local, exercise-level? That would seem strange, as some weeks you just feel so hammered and are obviously in need for a sabbatical from training to come back refreshed afterwards.

    ?Looking forward to your response.

    • Thanks, Stijn.

      Full-body deloading generally makes no sense, because it ignores the fundamental principle that muscle fatigue is largely a local process. So it’s only during serious overtraining that this is required.

      There’s zero evidence that ‘functional overreaching’ results in more growth than continued progression and in my experience, more consistent progress almost always wins out over time.

      Much of the confusion surrounding deloading is based on the lack of differentiation between mental and physical fatigue. It is commonly assumed these 2 correlate strongly, but that is not the case at all in many contexts.

  2. DAVID W VAN HORN says:

    “Physically, AVT increases force production throughout the set. During exercise with a defined endpoint, like a certain number of repetitions that you want to achieve, you will subconsciously pace yourself by holding back during the early part of the exercise.”

    The linked study states the opposite of what is being said here. They found that knowing how many repetitions subjects were expected to perform decreased force output, while having an unknown amount of repetitions to perform increased force production.

    To quote: “[E]xpecting to perform fewer repetitions than they actually will, may lead to a greater force production, whereas withholding an endpoint in terms of repetition number will result in under performance.”

    Am I missing something here?

    • DAVID W VAN HORN says:

      Or perhaps I’m misunderstanding the application. Is the point to get maximal force production during the first set (“deceiving” yourself as to how many repetitions to perform), then not worrying about how many repetitions you’ll perform in the next sets? So the first set’s force output is prioritized over the force output for the subsequent sets?

      • No, what I say and the study finds are the same. During AVT you don’t know how many reps you’re going to perform in your volume sets, so you can focus on quality over quantity and force production will be higher.

  3. Wade says:

    A resounding “YES” to all of this! I love the rationality inherent to AVT. I can personally attest that it frees up your mind to focus on the repetition at hand and not stress about completing X amount of reps.

    I look forward to more articles like this in the future! Thank you for producing such rich content, Menno. It’s beyond appreciated…

  4. Rikard Dahl says:

    Really interesting read!

    Can you give some indication of how to assess terminal consistency for an exercise? And what does terminal mean in this context; is the terminus the outcome of the performance of the exercise? Or is the meaning that each term of training – every instance of performing the exercise – is consistent with other terms?

  5. mo says:

    Menno: I find the idea that whole body deloads are not necessary interesting, as it paints the old time Reeves era bodybuilders as not so ignorant, as they never did deloads (AFAIK). Yet, they obviously progressed.

    The trouble I have is that this goes against what I’ve learned about periodization in the last few years, mainly about inducing a functional over-reaching to create a larger benefit from super-compensation (thus the prefix “super”). And I’ve heard Schoenfeld and Israetel speak to this in one way or another (Israetel suggesting a deload every 4-6 weeks, Schoenfeld every 4th week, if I recall). Still, most periodization comes out of sports training and because of this may have little relevance to hypertrophy, and it’s differentiating what is sports specific from the non-sports centric which is difficult.

    In reference to Israetel, he’d stated that deloads are necessary because, no matter how you manipulate variables, tissues don’t heal 100% from session to session, necessitating an eventual deload, although I can see this not necessarily being the whole body.

    Your recommendation reminds me of Izzy Narvaez that suggests resetting stalled exercises on an exercise case by case scenario.

    I can see your suggestion as an alternative to the crazy practice in bodybuilding (and Westside barbell powerlifting) of dumping and rotating exercises once they stall. Personally, I’ve just changed the workload intensity, after doing an active deload followed by a passive deload if I really feel drained. You’d think I’ve wasted time “resting”, but that’s not what some research suggests that shows the same rate of progress comparing groups that deloaded and those that didn’t (if memory serves).

    Anyway, you’ve given me something to think about.

  6. Victor says:

    HI Menno,

    Concerning AVT I don’t understand something. Maybe you could explain it further.
    For instance, you say that your first set is your benchmark and the rest of the sets are volume sets. Ok, but I don’t understand how many of this sets need to be done.

    For example. I incline press 225×8
    My first set is 225×8 and I feel fine, the volume sets are done in a specific rep range? If not, I don’t understand when you stop.
    Let’s say i feel like crap and I do 225×8 and was a grinder. The next set I do 225×5 and I stop at 1 rep close to failure. How many sets I do? Another and I get 3 reps. I keep going? I really don’t understand how many sets I need to do.

    Another question, how do you progress in your benchmark set?

    Thanks Menno. Hope you can shed some light here.

  7. Pantherhare says:

    Great article, Menno. I remember you first talking about your autoregulation approach in an interview with Armi Legge, but this article definitely paints a more detailed picture, especially with your view on reactive deloads.

    What is the advantage to not autoregulating the first set? Don’t you risk (if you’re having a bad day) grinding/missing reps and thus negatively affecting volume on subsequent sets or (if you’re having a good day) leaving too many reps on the table? Why not, in your example, just program the first set at 260 lbs at RPE 9? It seems that overall volume will still be less if you’re having a bad day and more if you’re having a good one.

    What do you think of the central governor theory? That while hypertrophy is local, performance is global? For example, being overly fatigued with squats and romanian deadlifts may diminish your performance with bench press.

    • Grinders or failure during the first set shouldn’t interfere with your progress much if you’re already skipping all remaining sets after that, so that’s generally not a concern for higher volume training. Likewise, it’s very rare that someone on a good progression model can hit the rep target (however that was determined) for all sets if that wasn’t the plan. That suggests a more fundamental problem with the program’s progression model.

      Central fatigue exists, but it’s not nearly as common as local fatigue and people very often erroneously attribute the former to the latter. That’s a separate topic though.

      • mo says:

        It’s my understanding that failure training early on does affect the volume for that exercise in that session. That’s why it’s avoided like the plague in powerlifting, especially with test or top sets (what you’re calling AVT, which is really DAPRE, if memory serves [and your volumes sets work sets sometimes drop sets but not in the bodybuilding manner]).

        In fact, I’ve observed from my own training, fatigue accumulates within a session (no news here) and affects subsequent exercises. That is, when I’m done with my primary lifts and onto my secondary ones, I lose about 10%; if I mean to hit the same amount of reps in my secondary lifts as in my primary ones, I need to lower the 1RM about 10%. I can only imagine how much more fatigue I’d accumulate and would affect the following exercises if I were to hit failure on the primary lifts.

        I’m not sure it’s clear what you by “it being rare to hit target reps, if it wasn’t the plan” … “You usually hit target reps, when you plan for it”.

        I’m wondering where the terminology used in this article comes from, such as AVT, considering there already is existent terms in current and popular use, such as those already mentioned top set, test set, work sets, etc?

        • I wouldn’t say there are well and consensually defined terms of this at all and I’ve never seen anyone use AVT like I do, which is why I designed it.

          • mo says:

            True, some terminology is confusing such as the term “drop set” which means different things in the powerlifting environment than in bodybuilding; it’s really two different things.

            And so I can see the need for a “unifying language” to encompass seemingly contradictory and confusing terminology, but I would argue that there is consensus on terms such as top set, test set, and work set.

            Still, I appreciate your work and given me a lot to ponder.

  8. Dear Mr. Henselmans,

    Would you ever consider allowing individuals to purchase select topics from the PT course. This would be an option for those of us not interested in become a PT/ possess less interest in certain topics?

    Thank You
    Zack Kaplan

    • I’ve done that, yes. It’s possible for certain topics at certain points (limited by the course structure and topic interrelation). Send me an email if you’re interested ( MennoHenselmans @ BayesianBodybuilding.com ).

      • Zack Kaplan says:

        Dear Mr. Henselmans,

        I emailed you as stated above but have yet to receive a reply. I realize that you most likely receive a high volume of emails and have many other priorities. Thank you for taking the time to read and answer my posts. I look forward to your reply.


  9. Saif says:

    Hi Menno,

    Your articles changed my life! I signed up for the course and can’t wait for it to start.

    Prior to discovering your work, I used the Hypertrophy Spesific Training (HST) method. After reading this, I’ve been using AVT and I’ve increased my frequency after reading your Norwegian Frequency Project article. I’m not sure how to keep doing that while applying the Progressive Overload and Strategic Deconditioning principles of the HST method, or if I should continue to adhere to these principles. What are your thoughts on them? Since you can’t increase weights/Progressively Overload forever, is there an alternative solution besides Strategic Deconditioning?


  10. Arthur says:

    what do you mean by terminal consistency ?

  11. Vincent says:

    “Microloadability: The larger the increment, the less likely the need for reactive deloading is.” Shouldn’t it be “the smaller the increment…” ?

  12. Patrik says:

    Hey Menno, great article!

    Might be a silly question, but I was wondering if it is okay to utilize a “double progression” method for the first benchmark set?

    For example, I set a rep range of 8-12 for my benchmark set.
    Session 1: I hit 10 reps with the base load in the first set, followed by 2 volume sets.
    Session 2: I hit 12 reps with the same load in the first set, followed by 2 volume sets.
    Session 3: I increase the base load and hit 8 reps in the first set, followed by 2 volume sets.
    AND IF I happen to get below 8 reps in the benchmark set, I will utilize reactive deloading.

    Thank you! Appreciate the information you share!

  13. Tjark says:

    Hey Menno!
    First of all, it’s awesome, that you share such a detailed article with your readership for free, so thank you for that!
    My question is, what is your reasoning for keeping the weight constant on the “volume sets”, instead of dropping some weight from set to set, aiming to roughly reproduce the RPE and reps of the “benchmark set”?

  14. Tom says:

    Hey, I’m doing something like Texas Method, but autoregulated. I’ve been doing 5% fatigue for press and 5% fatigue for incline bench on a monday-volume day for 1,5 months. My press is growing but incline bench is stuck.

    If this is the same movement pattern (I count it as press pattern with 10% altogether), shouldn’t both progress simultaneously?

  15. Tom says:

    “The result is that AVT normalizes the training stress over time to prevent overreaching yet also ensure a sufficient training stimulus.”

    How can this be true? How do you know, how much fatigue % have you induced by performing sets of 8,5,5? Magnitude of a training stimulus is determined by volume (intensity is training effect) which subsequently cause neuromuscular fatigue (if we’re talking about strength training of a particular lift). Here we only know, the lifter performed less reps in subsequent sets, nothing more.

    Ex. lift 100kg for 5 reps rpe 9. Then substract 10% and perform multiply sets until your rpe catches up with the initial, which is 9. This way, we know we’ve just caused 10% fatigue on this particular lift.

    Or did you mean – in above examples – ‘rep drop’ method which is also a way to measure induced fatigue %?

    • I’m not fond at all of using RPEs, since they’re highly confounded by many factors, but that’s another topic.

      The subsequent set volume autoregulates training fatigue. E.g. if you recovered very well and the first set was easy, you’ll automatically perform a higher training volume.

      • Tom says:

        Could you expand what flaws does RPE-using training has? As far as I know, many powerlifters and strength training athletes are using this system to measure exercise fatigue, ex. Mike Tuchscherer. So it’s a serious accusation.

        This could be idea for an interesting article.

  16. Sam says:

    Hi Menno

    Thank you for the article.

    What happens the next week/session for the exercise you had to deload?

    Do you just keep progressing as if you never took the deload and try to hit the same weight you were meant to the week before (week of the reactive deload) or do you start over, say in a linear periodised fashion?

    Also if you have more exercises for the same muscle group do you skip these?

Leave a Comment