Is it possible to overtrain one discipline (swim)? (Page 2)

Perfect - thanks! So now we have examples of different swim interval strategies. Suzanne's strategy specifically targets a form threshold and aims to work right around that level. Another strategy, probably the dominant one, seems to target other systems while repeatedly pushing through a form threshold. Suzanne's approach sounds well suited to the specific goal of triathletes, which is to swim at a pace beyond which greater effort yields diminishing returns. It would be interesting to hear what others have to offer on the relative merits of these approaches for triathletes.

I'd like to offer a corollary to Suzanne's concept of a 'form threshold.' To start I'll note that in the land-based disciplines, in which conditioning is responsible for 80% or more of improvement and success, and in which there has always been strong correlation between markers like VO2max and performance, pursuit of a 'lactate' (or any other kind of physiological) threshold makes perfect sense.

But in swimming, in which 80% or more of improvement and success are determined by efficiency (figures from Michael Joyner M.D. director of exercise research at Mayo Clinic) and in which there has as often been a negative, as positive, correlation between physiological markers and performance, pursuit of a Form Threshold makes far more sense.

Here's the corollary. If you're pursuing a lactate threshold, then your focus or the quality of your attention makes little difference. Training on autopilot won't compromise your results significantly.

If you're pursuing a form threshold, then the quality of your attention makes ALL the difference. Which should send an alert signal to all those who acknowledge "boredom" as nearly inevitable in their swim training. Clearly you need to rethink what you're doing.

So I'd like to suggest as a requirement to finding your Form Threshold, that you ought to consider pursuing something more basic -- a Focus Threshold.

This morning I was reading on-line an essay by William Deresiewicz reprinting a lecture he gave recently to the plebe class at West Point, where I used to coach. The topic of his lecture (http://www.theamericanscholar.org/solitude-and-leadership/) was how essential independent thinking is to the development of leadership capabilties. He talks about how multitasking hurts our ability to concentrate, to think deeply and critically. That got me thinking about the nature of swim 'workouts' and how so many either include pointless multitasking (buoy, paddles, kickboard, fins, snorkels, resistance tubing) or rote repetitions that lead to autopilot.

Immediately after reading I made these notes, which relate to the concept of Focus Threshold. These are principles I employ in planning my own practices, and which I think likely Suzanne considers in planning hers. They are also principles we teach to TI coaches-in-training. What they address is how the brain process information. Physiologically-oriented training considers how the body metabolizes energy -- appropriate when pursuing a lactate threshold.

>>How long should a practice be?
Usually at least 25 minutes. Seldom longer than an hour. If you continue longer than an hour, be very clear on your reasons and what benefits are produced in the 70th, or 80th or 90th minute. Even if your race is longer than 60 minutes, you can probably prepare more effectively with more frequent practices of 60 minutes, say five per week rather than three – assuming you have the schedule flexibility for that.

How many sets in a practice?
Usually more than one and fewer than four. A single-set practice can be valuable when you have a limited practice time, say 30 minutes or less. It should be designed to focus or challenge you pretty keenly for the entire duration. More than three sets increases the likelihood of too-diffuse thinking, of limiting your ability to stay with one task long enough to attune your brain and nervous system to the task you’re working on, ‘crack the code’ of the task you’re working on and practice your solution long enough to create some adaptation to your nervous system.

What should you include in a practice?
Tasks that force you to think. Avoid tasks you can do on autopilot. As well avoid tasks that overload or cause paralysis by analysis. Avoid multitasking for which you’re not prepared. That inhibits your ability to think deeply.
Avoid mindless, or unexamined, copying of practices planned by others. In that sense, doing all your swimming in group or coached workouts flattens your learning curve. When you do swim in a coached practice, do everything possible to use the assigned sets to sharpen your critical thinking skills. Focal Points? Alternate ways to structure the set? If the coach’s sets encourage autopilot what can you do to sharpen their focus or stimulate your brain?

Plan a main set with a crystal-clear improvement focus.
Base it on a task set that presents a problem to solve or highlights a skill or competence that is right on the edge of your comfort zone. Repeat that task set three to five times. Each time you repeat the basic task set, you should be looking for signs that your nervous system is becoming better attuned and adapted – i.e. that you are ‘optimizing the circuit.’

Keep your practice open-ended
Avoid locking yourself into a predetermined pattern. Continue as long as you feel yourself gaining insight or awareness, deepening perception, honing or imprinting a skill, improving a circuit. Adjust if the task is too easy – to avoid autopilot – or if you’ve set the bar too high – to avoid losing motivation. Stop when you feel you’ve wrung out all the possible improvement, or if physical or mental fatigue are compromising your ability to achieve the set’s objectives – I.E. You’re degrading, rather than optimizing, the circuit. Revisit it tomorrow or next week. The better-designed your task, the more intently you focused during it, the more likely your brain will recognize it as important and continue to encode the solution or pattern when you’re not doing it – even as you sleep.

colin, I think you're overthinking the room a bit here. I hope this doesn't come off snarky as that's not my intent. You know and understand cycling training and you're trying to fit swimming in the cycling box so it makes sense to you. As a swimmer, I don't  look at run training and ask them why they take such long breaks inbetween intervals. It seems a complete waste of time when you're sitting in a pool looking out, but when you get on the track and try some repeat 800's, it makes perfect sense. Likewise, as a cyclist, I don't try and do bench press reps for 15 minutes straight. That's not how that sport works either. To borrow the phrase you hate, "Swimming is just different". There's a high level of technique involved like golf, but there is also a high level of fitness needed as well, like running. Off the top of my head I really can't think of another sport that is so dependant on both. As a beginner, a lot of triathletes take the advice from the Total Immersion folks like Suzanne and Terry. It's a great learn to swim program and will get you going in the right direction. As you develop as a swimmer, you will probably adjust your training so that you train like a swimmer, which is more along the lines of what I discussed.   Good luck and I hope all of this stuff that's been thrown at you helps.

"form threshold" was colin's phrasing, to give credit where credit is due. I like Terry's concept of a "focus threshold" as well.

By introducing the term Form Threshold, Colin points up that the key concept is finding your functional limit and working very close to it, with the goal of incrementally raising that limit.

The reasons Suzanne and I prefer to focus on a Form Threshold are
1) Physiological research has demonstrated that efficiency, rather than physiology, is the more critical component of swimming performance.
2) The Form Threshold is more precisely measurable and allows for a far greater variety of sets aimed at improving it. E.G. I have different Form Thresholds at different distances, Stroke Counts and Tempos.
3) For each of the many combinations of duration, SPL and tempo, your metabolic capacity will adapt to support the specific work being done. More importantly, your nervous system will also be adapting specifically to the chosen task. Adaptation leads to improvement and in a week or two, you're ready to attempt the same combination of SPL and Tempo for a longer duration. Or a higher tempo for the same SPL and duration. Or perhaps save a few strokes over the course of a swim of similar duration and tempo. And when you save strokes, you save seconds.

When you train to improve Lactate, or other physiological thresholds, you're targeting a capacity for which there's no demonstrated correlation with performance, and you have no 'real-time' way of measuring whether you're accurately targeting your threshold.

Thanks, TJ. Overthinking is my favorite of the three extra disciplines of middle-aged triathlon (overthinking, overtraining, and overspending). I didn't mean for this to devolve into another debate about general TI principles, as that's a whole different matter, and it has been played out many times elsewhere. You're right that I'm dissatisfied with the answer "swimming is just different". Not because swimming isn't different. It clearly is. But because that kind of statement could be used to justify a million different approaches. And my hope was to understand why high intensity swim training differs from other endurance sports in very specific ways, in particular the focus on relatively short intervals with very short rest periods.

I'm not pushing on the "why" because it's essential to my own swim training. I have so much to learn right now that anything remotely sensible is helping me to improve. I'm taking lessons from an instructor (who comes from a sprint background), I'm doing what she tells me to do, and it's helping a lot. The reason that I'm pushing on the "why" is because it's interesting to understand the differences in what we're doing to our bodies in the three sports, and because understanding the rationale for the training makes it easier to get out there and do it. I also think that a lot can be learned by comparing ideas from different fields. (That appears to be how interval training methods reached swimming in the mid-1950s, translating ideas developed by Gerschler and Stampfl for running in the 1930s-40s, and brought to great attention by figures like Emil Zatopek and Roger Bannister.)

Getting back to the main question, It's clear that swimming requires high levels of technique and fitness at the same time, and training must target both of these. (Even if LT etc. explain less of the variance in swim performance than they do in cycling, they're still an essential component of training.) In addition to this, TJ and AB both pointed to the importance of learning to maintain good technique under increasing levels of stress (form threshold), and pointed to a couple of different strategies for limiting form breakdown. 

TJ made an interesting remark in his first post, which could provide a clue to a physiologic role for very short rest intervals. He suggested that just a few seconds are sufficient to recover from form breakdown. I'm far from clear what are the causes of loss of form/coordination, but if it's correct that those systems recover much faster than do overloaded aerobic systems (does anybody know the science behind this?), then this could be important. So from what we've heard I could imagine a few different reasons why only swimmers do sets with ultra-short (:10-:20) rests. It depends on what abilities such sets are targeting.

(1) The sets are targeting something akin to form threshold, and the short efforts/rests are an ideal way to do that, for some physiological reason that we're yet to hear about.
(2) The sets are really targeting more basic aspects of conditioning such as LT, and the full sets should be regarded as the 'real' efforts, with the brief rests serving practical/organizational needs (pacing, group dynamics, etc.). This would make the aerobic aspects of the training more similar to other sports than they first appear, e.g., 10 x 100 is really a more structured version of 1 x 1000.
(3) The sets are targeting more basic aspects of conditioning such as LT, but the very short rests serve a special purpose, because they are long enough for the physiologic limiters of form to recover, but too short for the aerobic systems to notice the break.
(4) The goal is not to target an aerobic limiter, but rather is to swim aerobically as fast as possible without crossing the form threshold. The very short breaks enable this to happen. Longer breaks would make the efforts more anaerobic, which would be counterproductive.
(5) We're not quite sure, but if it works for [insert favorite swim star here] then it must be good.
(6) One of many other possibilities that I'm too ignorant to think of.

Thanks, and Happy Christmas to all!

Colin
I've gotten entirely away from all the formulae related to 'energy system training.' I should be clear I used to be as much a part of that 'church' as anyone, having coached competitive swimming at quite a high level - many D2 and D3 national champions, world-ranked swimmers, developed two later Olympic medalists as age groupers - from 1972-89. Since starting TI my work has been mostly with 'adult-onset' swimmers -- and myself.

When I trained myself according to the energy-system formulae I did okay, but not great. Since shifting to setting my intervals organically, rather than via canned formulas, I've gone to an entirely different level, one I never would have guessed possible prior to my 50s.

When I say organic, I mean that if, in a particular set, I've taken on the task of progressing from 50-yd to 100 yd to 200 yd repeats while maintaining 14 SPL at a tempo of 1.1 sec/stroke, I set my rest entirely by feel. Because I do sets like this with such regularity I know - by feel - how 'fresh' I must feel to succeed at the next highest level of the task. That personal sense of readiness - and nothing else - determines my interval.

Here's the difference: If your training is aerobically-oriented, you will probably have read that to develop Lactate Tolerance - and here I'll take a page directly from Ernie Maglischo's book - your sets must be 300 to 1200 yds, repeat distance should be 100 to 200, rest intervals X and repeat speed Y.

I tried to follow those early in my Masters career, mainly because the sense of having a formula, worked out by a scientist brought a degree of comfort. But my nature rebels against the formulaic and I moved gradually to training by feel.

When I became acquainted with Mike Joyner the head of exercise research at Mayo Clinic he told me two things about those formulae
1) In his opinion they were junk science.
2) Because the research was done on treadmills and exercycles - and not in the pool - he further doubted their applicability to swimming.
3) Even in running, where there had been a relatively high degree of correlation between performance and physiological markers (there is no such correlation in swimming) the best coaches paid little attention to formulas, being guided far more by keen intuition and knowledge of the individual athletes.

So now my intervals, my repeat distance, my set distance, etc are guided solely by the combination of SL and Tempo I've tasked myself to work on that day. If the combination is relatively less challenging, I can swim longer repeats on relatively low rest. If the combination is highly challenging -- i.e. at or exceeding a particular Form Threshold -- then it will be shorter repeats and more recovery.

Mike, Jonty Skinner (former director of performance science for USA Swimming and now for British Swimming) and personal experience have convinced me that the quality of neural adaptation, not aerobic adaptation, will determine how well I swim.

Suzanne (AB) spent the pre-TI part of her tri career as an ardent aerobic-conditioner. Her times were stagnant. Since shifting to total focus on neurally-guided training and systematically raising her Form Threshold her times have improved continuously - at all distances - and she now routinely repeats (on relatively short rest) 500-yd practice swims faster than her previous all-out single-effort best.

Thanks, Terry. In an effort to keep track of the different issues that are coming up here, I think we need to distinguish (i) which performance limiters should be training priorities - this is the focus of your most recent post, and (ii) for any specific type of training session, which performance limiters it primarily impacts - this is what's behind the attempt to understand the unique interest of swimmers in sets with very short rest intervals. One may choose to write off a current dominant strategy for swim sets, but I'd prefer to better understand it first.

In order to understand the swimming-specific value of micro-intervals a key question is what changes occur during very short :10 to :20 breaks between efforts (aside from pace checking, group organization, etc., as  already discussed here; those features may be less relevant to triathletes who train alone and have access to increasingly sophisticated monitoring devices). Current understanding about aerobic systems is that it's too short to make much of a difference to them, but there could be all kinds of other abilities that do benefit from such short breaks. The best suggestion that we've heard so far is that the brief rests might selectively benefit some of the limiters of good form, though it remains unclear why. Another possibility that Terry raised is the benefit of a short mental break that could allow refocusing on the goals of the set.

Stepping back from this central issue, I can't help commenting on the main claims of Terry's post, which contrast "neurally-guided training" vs. "energy system training", and makes the claim that there's no correlation between swim performance and physiological markers. I think this runs the risk of diverting the thread (more than I have already done!) to a different issue. But since everybody seems to agree that good form is important, and that it becomes increasingly difficult to maintain good form as physical stress increases, then conditioning is relevant to all. Also, I'm squeamish about notions like "neurally-guided". In my day job I am a bit of a neuroscientist (though definitely not a motor systems expert), and so I've learned to hold onto my wallet when I encounter ideas that are claimed to be based in neuroscience. 

-- Discussions of training "muscle memory" in swimming refer to the development of complex motor skills, i.e., neural adaptations that allow automatic coordination of many different muscle actions. Extensive practice leads to adaptations that reduce the dependence of these skills on attentional control. But even the most ingrained motor skills, such as walking or running, can be impaired by stressors, such as fatigue (or intoxication). So even the most well-practiced neural adaptations require training that reduces the impact of the stressors.

-- I'm skeptical of the claim that physiological markers don't correlate with swim performance. If you take a small sample of elite swimmers then you may find that aerobic abilities do not correlate with swim speed, but it is a statistical error to conclude that aerobic abilities do not contribute to swim performance. This is because another key performance predictor that is not included in the calculation (e.g., form efficiency) can mask the contribution of aerobic abilities.

-- All agree that aerobic abilities are a pretty important contributor to bike performance (although equipment, positioning, etc. are also important). So correlations in triathlon bike/swim times might offer a rough-and-ready measure of the contribution of aerobic conditioning to swim performance. I tried this out with the results from last year's Columbia Triathlon (~1500 sets of splits available). This race has a tougher bike course (only 16% averaged 20+mph, compare 50% at Boulder 70.3), so the contribution of positioning to bike speed is likely smaller than in some other triathlons.

-- Among the top 50 finishers, the correlation of swim/bike times is indeed rather low (0.06). One could take use this as support for the irrelevance of aerobic capacities to swim times, but it probably just reflects the fact that for these elite performers, who all have superior aerobic conditioning with narrow variance, swim technique is a more important predictor of swim times (surprise!).
-- Among all 1500+ finishers, there's a strong correlation between bike times and swim times (0.54). This could show the importance of aerobic conditioning. Or it could merely show that those who train appropriately for biking also train appropriately for swimming.

-- It would be better to focus on the most proficient swimmers, across the full range of fitness, to see if swim/bike correlations remain high. As an approximation to this I took the fastest 5% of swimmers from each band of 100 overall finishers, assuming that the people who swim much faster than the people with similar overall times are most likely to be those with an extensive swim history. For this group, the correlation of swim times to bike times remains high (0.42). An imperfect measure, for sure, but it suggests that aerobic conditioning isn't a waste of time for swimming.

-- Even if somebody doesn't believe that aerobic conditioning is important for swim performance, and so only wants to practice form, a lot, one needs a good amount of conditioning in order to do that form training effectively.

That said, I hope we don't go off on this tangent, as I'm still hoping to learn about the effects of very short rest intervals. 

Not unrelatedly, here's an interesting article (well, for training history geeks at least) that I found on the history of Australian swim training over the past 120 years, by somebody who has been around for most of that time. The author makes an interesting claim (p. 20) that failure to emphasize short-rest intervals may be responsible for a stagnation in performance. 
http://www.h2oecon.com/AUSSWIM.pdf

I have never, nor would I ever, assert that aerobic capacity is unimportant to swim performance. Rather I assert, with support from eminent exercise scientists, that the quality of one's neural programming is the more critical factor on race day.
The essential point I make is that the heavy - indeed virtually exclusive -- reliance on planning training according to 'canned' energy system formulas is misguided. I understand entirely why this is the case -- that research was all that was available when swim coaching began to shift from pastime to profession. Neural adaptation doesn't receive a single mention in either Counsilman's Science of Swimming, nor Maglischo's Swimming Faster.

My point is that training sets designed based on insights into how to 'optimize circuits' have strikingly different parameters and the key metrics are very different, compared to training sets designed according to the much older research on how the body metabolizes energy. Which I further note correlates reasonably well with land performance and very poorly with aquatic performance.

As well, when one is task-oriented in training design (examples in both AB and my earlier posts) aerobic conditioning still occurs, but it's task-specific, rather than general. In mid-set, I distinctly sense that when I'm at my Form Threshold, that my LT is also under stress. The difference is that I can precisely -- indeed mathematically -- measure and adjust my FT. Doing the same for LT can never be other than guesswork.

I'm 3 chapters into a book titled "The Grownup's Guide to Swimming Faster" which I immodestly feel will prove eye-opening reading for the majority of triathletes. I'm posting chapters for advance viewing and critical comment here http://www.totalimmersion.net/index.php?option=com_jfusion&wrap=sho...

If you have professional experience in neuroscience I would certainly welcome any critical comment you may care to offer.

Thanks again, Terry. I'm still trying to keep this thread on-topic, focusing on what might make the design of standard swim sets effective for triathlon swimming but not for, say, triathlon biking. In your post and your book excerpts you question the value of trying to understand it, arguing for an approach that you view as quite different ('neuro-motor training'). There's more consensus than this implies. We can identify specific points of contention, and I have a couple of thoughts on this (based on my neuroscience background; I don't have any swimming credibility, hence my questioning).

Everybody agrees that training aerobic systems is important to all, including triathlon swimmers. Even Terry concurs, at the start of his last post. (He continues to write things that could could mislead people to think that they're unimportant: from the new draft "Increasing power won't make us faster", "Those same metrics [aerobic markers] have never correlated with performance in swimming". But we can set these aside. Everybody really does agree.) I continue to wonder about the physiological effects of :10-:20s breaks on aerobic training, but we're yet to hear a specifically aerobic motivation for it.

Everybody agrees that good technique is important for swimming, and far more so than in biking and running. Everybody recognizes the value of practicing technique at low fatigue levels.

The third ingredient, which is our focus here, is the importance of learning to maintain good technique at increased levels of physical stress. Everybody agrees that this is important, and Terry and Suzanne pointed out some straightforward ways of measuring loss of form as fatigue increases. What's not yet clear is whether there's a training level ('form threshold training') that is different from the standard aerobic and anaerobic regimes that so many swimmers use. Secondary to this is Terry's notion that there's a distinctive notion of 'hardening neuro-motor circuits' that is essential in this ability. What's the justification for such a training level? It probably depends on the swimmer's expertise level.

Novices first. Novel complex motor skills require a great deal of attentional control, and so they're vulnerable even under quite mild stress levels, presumably well below standard aerobic thresholds. So novices could hit a form threshold very easily. Gradually automating the various aspects of good technique should remove this limiter. We could guess that this automatization should be most effective when developed at mild to medium fatigue levels. This could be the level that Terry focuses on when he talks of developing 'muscle memory'. Conditioning at higher levels of intensity could make it easier to train at the lower intensities that support automatization. Also, the science tells us that it would be a great idea to get some REM sleep shortly after sessions that focus on technique training (5am swim workouts not so ideal for that).

Once a complex skill is automatized, it's far less dependent on attentional resources. In fact, conscious attention can be harmful (there's interesting recent research on how this causes 'choking' in sports). So mild fatigue should no longer impact good form. For swimmers whose technique is automated, it's certainly the case that standard aerobic limiters (V02max, LT, etc.) can undermine good form, so raising those thresholds should correspondingly raise the form threshold for those swimmers. So standard approaches seem well justified. *If* there are lower levels of fatigue that also impair automatized motor skills, then there would be good reason to focus on training at lower levels. My reading of the science hasn't yet turned up any evidence of a lower fatigue threshold for that affects coordination of automatized complex skills. 

Motor neuroscience tells us that there's another important level of skill development, which goes beyond 'muscle memory'. You could call is 'flexible proficiency'. It involves the ability to achieve the same motor goal using different combinations of joint and muscle actions, and to be able to rapidly correct motor plans in the face of errors and disturbances. This ability could be very relevant when fatigued, as it allows the swimmer to maintain good form even when some muscles are becoming more fatigued. It could allow swimmers to maintain good form when at least some muscle systems cross aerobic thresholds. It could be particularly relevant to the varied environment of open water swimming. Flexible proficiency requires a shift from automatized motor actions to automatized motor goals. My read of the science is that experts are currently beginning to understand how to measure flexible proficiency, and that it's not yet well understood how it develops. It's obviously a key part of what happens with the most overlearned motor skills, such as walking, reaching, and talking. The best guess would be that it develops better when people (i) know what they're trying to do, and can recognize success/failure, and (ii) are exposed to varied circumstances. (i) is presumably what experienced swimmers refer to as 'feel for the water', and (ii) is probably handled pretty well by standard training. 

Based on this, then, it seems that training the form threshold should be pretty well served by the goals of standard swim training, with the proviso that the threshold is likely far lower for novices, which is why Terry's methods can serve them so well at early stages. And why they're less relevant for training the form threshold for an automatized stroke. And with the additional caveat that we're a little bit in the dark about how flexible proficiency develops.

Stepping back from this for a moment, I should comment on a couple of the claims about neural adaptation etc. that are being raised in Terry's remarks. "Neural adaptation" is a fashionable 21st century term for "learning", and so it's not such a novel idea. The most important learning/adaptations are likely different at different skill levels. I'd also take issue with the suggestion in Terry's new draft that recent brain research justifies the notion that "we can hone skills decades longer than we can improve aerobically". It's certainly true that there are some striking recent findings about brain plasticity (i.e., adaptation, learning) in adult mammals, but this doesn't change the long-known fact that skill learning in adults is far more difficult than in children. New neurons can develop in adults, but they're small in number, and may be concentrated in a small number of brain areas. The most impressive demonstrations of adult neural plasticity occur following brain damage or periods of total sensory deprivation, neither of which are obvious models for swim training. Adults of any age are likely below their full aerobic potential and below their full skill potential, so there are gains to be made across the board. Where adults more likely have the edge over children is their ability to learn from patient, explicit instruction, i.e., the first step on the path to automatization.

(I'm guessing that most people have stopped following the discussion at this point. But I'm finding it helpful for myself, and figured that there might be the odd nerd or two who cares about this.)