10% "rule" with regard to weekly speedwork -- Who says?

Coach B wrote:

Though I've often heard that a runner's weekly speedwork sessions should not exceed more than 10% of his/her weekly mileage, where are the studies that document this? Is this rule truly grounded in scientific studies that point to a detrimental effect when the speed work exceeds 10% of weekly mileage, or is this just another example of the "conventional wisdom" that might be more anecdotal than truly accurate? I coach runners,, some of who are running roughly 30 miles a week, so their speed sessions "should" not exceed three miles, but I will routinely have them do, say, 2 sets of [ 600 ; 1000 ; 1600 ] for a total of four miles, and they don't seem to be the worse for wear. Does anyone have any insights into where the 10% rule came from and what, if any, science actually backs it up? Thanks, Coach B

But what KIND of speedwork? Anaerobic Capacity? Aerobic Capacity? Tempo even? It's a very "cookie cutter" RW type formula. It cannot be the same for 800m / Marathon, or for individuals. I think, generally, 75% at "aerobic conditioning" pace is typical for me, with the other 25% being tempo or higher.

Do you think Jeremy Wariner follows this rule? Webb--with speed session volume of 8-10K twice a week and a weekly mileage of 50-70 mpw this year--doesn't seem to be following it. Coe, Spivey, etc? I've seen a bunch of epidemeological studies on running showing previous injury history and high mileage to be primary causes of injuries in runners: I usually don't see speedwork mentioned, except anecdotally for some injuries like achilles tendonitis.

My guess is this rule came from beginners, runners with biomechanical problems, and maybe those not in shape to handle the work. The 10% rule is probably one way of making sure the athlete is in some kind of shape, but it's probably more oriented toward higher mileage types, rather than someone doing 2-3 speed sessions per week off low mileage. It needs to be adjusted for the athlete.

These kinds of numbers just don't make much sense at all to me. So many factors come into play that you can't just pick one number and say that it is the right percentage of speed work to do. If I'm training for the 800 and in my taper before my last meet, I'm really not doing much other than a few speed workouts. If It's the summer before xc season I'm really not doing any at all. If you have a training philosophy where you are doing your speed work at a higher intensity than the average runner, you'll be doing less. If most of your stuff is at race pace or below, you'll probably be doing more. If you have a low milage program you'll probably run a higher percentage of that milage fast. Too many factors to pick one number.

Daniels' Running Formula has more specific guidelines on %age of weekly mileage at various paces. Basically LT and MP work can be higher, and mile-5km pace running a little less. But it all depends on conditioning, intervals between hard efforts, and what period of the season you're at.

I'm just curious, but how do you have your athletes run that workout, the 2 x (600m, 1000m, 1600m)?

for 16:00 3 mile runners, pace = 5:20/mile = 320/1.6 = 3:20/km pace, and = 80 sec./400m = 2:00/600m

If I were to make this more of an VO2 max workout, I would do this:

2 x (600m @ 3k effort(2:00-2:30), 2 min. rest, 1000m @ 5k effort, 3 min. rest, 1600m @ 8k effort) rest 5-7 min.

If I were looking at this as a speed workout, I would do it like this:

2 x (1600m @ 3 mile pace, 4-5 min. rest, 1000m @ 2 mile pace pace, 3-4 min. rest, 600m @ 1500m-3000m pace) rest 5-10 min.

I imagine there are a number of different ways to organize this workout to get something different out of it for your athletes.

Typically, for a 1600, we run it in an 8:00 or 8:30 time block. Say it's an 8:00 block, the runner who does it in 5:30 has a 2:30 recovery; the runner who runs it in 7:15 has a 45 second recovery. 1000s are done in a 5:00 minute time block; 600s are in a 3:00 time block. This was, all the groups start each interval together. The faster runners get more of a recovery - - - Tonight we did repeat miles in an 8:30 time block. 3 X 1600 for our slowest runners; one runner who is getting ready for a marathon did 6 X 1600 @ 7:15 pace (his goal marathon pace is 8:15-8:20) . . . In general, my runners are encouraged to use the McMillan pace calculator to determine how fast they should be doing their intervals. E.g., we have a woman who is a new runner in roughly 41:00 10K pace shape, and she does her mile repeats in 6:30-6:40. It seems to work out! We get anywhere from 9 - 20 runners per week, and finish up at the local pint afterwards.

No studies. Your runners will be fine. As training rules of thumb go in coaching, I think "nothing etched in stone" is a pretty good one. Percentages this or that and one -size -fits -all mileage suggestions don't make nearly as much sense as that.

There are some pretty useful guidelines using perceived effort,

mood, and complaint index but those work because they are global and flexible. The same workout can be useful today and too much tomorrow.

I should have registered this name long ago.

I think when you are talking about real fast stuff (800-mile pace or faster), then 10% is a good guidline. You can't count tempo stuff in there though.

Of couse the same rules don't apply for the mid distance or sprint types.

there is another '10% rule' i've heard of if you care to listen, because maybe you were thinking about this one also, as it is related to speed.

canova and some others(hudson, etc.) practice the philosophy/science/whatever rule that when trying to train for specificity, as in training a 5000m runner to run their best 5000m race, they don't do reps at paces 10% faster than their goal pace for whatever distance theyre training for. for example, if teg was trying to train himself to be comfortable running at 62-second per 400m pace, 10% faster than 62 seconds would be 62 - .1(62) = 62 - 6.2 = 55.8, so running 56's is the fastest that would be useful for training his 'systems' to get into the kind of shape he is aiming at. i think the belief is that training at a pace faster than 10% doesn't really have much positive feedback, as in helping as much. maybe it just builds up more lactate, i dont know. so these specificity specialists, coaches like hudson, canova, many others, etc. will use the zones 10% slower and 10% faster than goal race pace. if someone is going to be doing longer intervals, something a little slower than goal race pace is used, if something shorter and meant more for speed, then a little faster than goal race pace. obviously you don't want to train people by having them try to do 2000m or 3000m repeats at 5k race pace, and you also don't want them running 200m or 300m repeats at long run tempo pace because those wouldn't help much either.

That's a helpful tip -- I like it! So if a 5k runner wanted to average 6:00s per mile (360 seocnds) for their race, 10% of that is 36 seconds, so they could do their mile repeats in 6:00 minus 36 = 5:24s? Or at least no faster than 5:24?

But WHY is it a good guideline (beyond your intuitive sense that it is)?

if you do reps at a pace 10% faster than goal pace, it's probably dipping too much into the anaerobic zone and maybe the athlete will build up more lactate and not develop the aerobic system for that specific distance enough. i don't know the exact science behind it, but canova might say something about it in his locked thread.

if you wanted to do repeat miles with them, my coach in high school tried using an idea that coach/dr.rosa uses where the mile repeat is a little slower than goal race pace, but the recovery is shorter. so instead of doing 5 x mile @ 6:00 w/ 5:00 rest, rosa's technique would imply doing 5 x mile @ 6:20-6:35 w/ 3:00 rests.

i think the longer the distance, the harder it is to train an athlete at faster than race pace, so if i had an athlete who's goal was to finish a 3 mile race in 18:00 min., I wouldn't have them try doing mile repeats faster than 6:00. But I might have them do the workout on 2 seperate occasions. the first time i would have them do it like 4-5 x 6:20 w/ 4:00 rest, and the next time(2-5 weeks later) i might have them try doing it again but like 3-4 x mile @ 6:00 w/ 3:00 rest.

having the athlete do reps at 10% faster than goal race pace is usually done when they are doing shorter reps, as in distances no longer than 20% of the race distance. for example, if you had an athlete try doing 2 x 3000m @ 5k goal pace, it might not be do-able(that would be like asking k.bekele to run two back to back 3k's with some sort of ample rest in 7:40- he might be able to do it because he's part superman, but you get the point- it could be to much straining).

Coach B wrote:

But WHY is it a good guideline (beyond your intuitive sense that it is)?

This is the important question that deserves an answer. I feel that too many people say, "do x because these famous coaches/runners do it."

One reason I'm guessing it's hard to get an answer is because there are so many possibilities, that it would be hard to answer the question, "Is running 10% of your weekly mileage for speedwork optimal?"

What do we think we know?

The higher the percentage, the more fatigued you will be. Will this effect future workouts? Recovery? Likelihood of injury?

What is the physical response to a speed workout (tons of variable just defining what a speed workout is) and what does the benefit curve look like?

What opportunity costs are there involved with the speed workouts? Are there other workouts that we could put in there that would be just as beneficial or more?

Even if you know all the answers you think you could throw at a single workout and how it effects an athlete, then you have to work like a chess player, with all the workouts, to maximize the benefits...and then to add to the complexity, figure out how to peak for a race or a season and deal with injury interruptions, etc, etc.

Coaching is ridiculously complex, and so that's why you get all these general rules that have come from successful coaches/athletes, that may or may not have any research backing them up.

You´ve got it wrong. Training methods are judged by the results they produce, not if they have any research backing them up. American distance running sucked in the 90s because runners and coaches started to listen to the exercise physiology siren song.

experience wrote:

You´ve got it wrong. Training methods are judged by the results they produce, not if they have any research backing them up. American distance running sucked in the 90s because runners and coaches started to listen to the exercise physiology siren song.

Well, I guess that was the point I tried to make at the end of my post. Perhaps the problem is just too complex to research from the ground up. It doesn't mean we shouldn't try. I think exercise physiology has given us a good understanding of why the training has worked. But trial and error with millions of athletes usually produces success before we can understand how it works.

The problem with trial and error is that the results can be flat out wrong. If an athlete starts wearing Newton shoes and then wins the Ironman, then you have a million triathletes wearing Newton shoes that probably make no difference in performance. So, the original question of "why" is still a good one to ask, and a good coach should have an answer.

Coach B wrote:

But WHY is it a good guideline (beyond your intuitive sense that it is)?

This doesn't apply to some Africans and high level NCAA runners where you might have guys with 3:31 1500 times or so running 10000. But generally I think long distance runners are lesser athletes in terms of their anaerobic systems, compared to mid-distance and faster types. The long distance types thus have less ability to buffer lactate/H+ ions and less ability ro recruit additional IIa/IIx muscle fibers to replace those damaged by acidosis, compared to, say, a 400 sprinter. Thus, they have to depend more on their aerobic systems to clear lactate and to avoid severe oxygen depletion during interval sessions. In these cases, someone running, say, 90 miles per week is likely to have a more productive aerobic system than someone running 30 miles per week, so some relation between mileage and interval load makes sense. But keep in mind, what matters is the abioity to tolerate and recover from high lactate loads, and 10X400 with short rest and 3X1200 with long rest are not necessarily the same in terms of load on the system, even though the volume is similiar.

That's a good point. At the beginning of our "speedwork" season, we do more shorter stuff -- 400s, 600s, 800s -- this is done to get the runners to recruit more of their fast muscle twitch fibres, and to get them used to the feeling of turning their legs over more quickly. Now that we are entering the cross country / marathon season, we are progressing more to mile repeats, obviously at a somewhat slower pace. These are more designed to build strength as well as mental toughness. We mostly play around with the "rest period' to make the workouts tougher. That's where I came up with the "8 minute time blocks" for mile repeats. Any runner who can run 4 miles in 32:00 can finish the workout in the proper time (4 miles at back to back 8:00 minute pace). The faster you go, the more "rest time" you build-up prior to the next interval. So the guys running their miles in 5:20 have 2:40 rest period before the next mile. One benefit is it keeps the group "together," in that everyone starts off each mile at the same time, builds a little cohesiveness into the workout. But these workouts were intuitively developed, I've never seen any science that validates what I'm doing (but the runners see to like it and they are improving, so I guess that's what counts).

Coach B wrote:

Any runner who can run 4 miles in 32:00 can finish the workout in the proper time (4 miles at back to back 8:00 minute pace). The faster you go, the more "rest time" you build-up prior to the next interval. So the guys running their miles in 5:20 have 2:40 rest period before the next mile.

This approach assumes that all your runners are grouped fairly closely in current ability. If you have a range -- say some people running 5:20 with 2:40 rest, while others are running 7:15 with 45s rest -- then different runners are doing completely different workouts. Some are doing essentially tempo pace with minimal rest, other have enough rest to do hard mile repeats. If the workouts are part of a coherent training plan where each workout has a purpose, then you'd expect that all the runners would be working on similar energy systems in a given workout. Otherwise it seems to suggest that the details of the workout don't actually matter at all.

Of course, if the group is mostly beginning runners, then it's probably true that the details aren't crucial -- just getting people out there to run hard will lead to significant improvement.