TRAINING: Rosa vs. Lydiard

I decided to write up this article as an attempt to answer Barry Minnock's question on my take on the training program pioneered by Gabriele Rosa (as described in this Running Times article from 2006 written by Lydiard Foundation member Greg McMillan).

First, I should start by saying there are always many ways to skin a cat that, if the overall principles are sound, will all lead to good results. Rosa’s system seems largely similar to Lydiard’s and is probably based on sound principles, so it may not always be a question of either/or but rather where and when and for whom? (using Rudyard Kipling’s “Fighting Men”).

Since one core tenet of Lydiard system is that “everyone knows what to do, but few know when to do it”, we should not underestimate the importance of timing of your workouts leading into a race peak. It is not just a consideration: It is the consideration, so why does Rosa move it?


I’m compelled to mention that what I am really trying to assess here is Rosa’s system as presented by Greg McMillan after spending some time with him, not his system as presented by Rosa himself (and not how it necessarily really plays out).

How many have misinterpreted Lydiard or other coaches’ system after spending some time with them? (many!) Its likely there are many finer details we are missing on exactly how Rosa implements his system (making it more or less different potentially) and these finer details are all-important. So this article should really be “Greg’s Rosa Training description versus Lydiard’s system).

Rosa Pyramid versus Lydiard Pyramid

The Running Times article presents a comparison between the “Classic Marathon” and the “New Marathon” pyramid (Rosa’s being the latter). In Rosa’s pyramid, you start with aerobic conditioning (as usual), then insert speed training and then strength (hills, stamina) before the “peak”.

The “classic” pyramid starts out with aerobic conditioning, strength, then speed and finally the peak. So the subtle change is the speed and strength phases switching places.

Lydiard’s pyramid, however, is not similar to the “classic pyramid” rather the classic marathon pyramid represents how the Lydiard principles are generally misinterpreted today.

The Lydiard pyramid starts with “marathon conditioning”, moves on to strength (hills), then anaerobic conditioning (not speed!), then coordination and then peaking before a prolonged phase of racing can follow.

Before we look at how the phases differ between the approaches, let’s just look into Rosa’s reason for changing around the order of his phases.


In the article Greg McMillan writes: Rosa’s experience showed that it was first important to develop the athlete’s speed so this would not be a limiting factor in the marathon training phase. The goal in marathon training is to fatigue the athlete with the duration of the workouts and not the speed, so speed needed to be developed first.

I should clarify the usage of “speed” and “speedwork” as the article uses a different, and to my mind more imprecise, terminology than the Lydiard Foundation (LF).

Speed is your raw pace, basically what you see when you run a 200m all-out. It expresses your current biomechanical and muscular maximal velocity. Most runners can run 200m at a pace far faster than any speed they can sustain over any meaningful endurance distance (e.g. 800m to marathon and beyond). Unless, your cruising pace is very close to the maximal pace you can sprint, then speed will simply not be a constraint for any training you do during the aerobic phase.

Speed work as understood by the LF are very short anaerobic bursts of speed (in reality “alactic” since the workout is too short to trigger the lactic energy system but this term is not commonly adopted yet). These training modalities include striders, sprints (60s, 100s, 200s and for very fast athletes 300s) and certain fartleks.

The speed work talked about in the article is actually anaerobic conditioning (reps of 200m to mile repeats and so on) which is not an ideal way to develop speed, to quote Arthur:

“Repetitions or interval training can improve your speed to a certain degree simply because of anaerobic development and improved mechanics. This gives the false idea that you are actually improving speed.”

“The actual fact is repetitions or intervals are used for anaerobic development. You are creating huge oxygen debt to develop a buffer against this type of fatigue. Problem is that when you do that, you invariably tighten up. You can not develop fine speed when you tighten up. The best way to develop speed is to use some of the American sprint drills.”

So if it was “speed” Rosa was primarily interested in developing, he would not use anaerobic training as he does which tells me that his interest is slightly different – rather he wants to ensure his athletes are strong enough to keep a reasonable pace throughout and is achieving that by having them do faster paced running (e.g. anaerobic paces) before beginning stamina work.

In a Lydiard Pyramid, there are two reasons why this is not necessary which I’ll cover below. But first, I should deal with Greg’s description that the goal of switching phases is to ensure you raise your speed so you can better sustain it. This assumption is entirely unfounded: Read any physiology book (or any of of Lydiard’s own books) and its clear that its not your lack of speed that stops you from running at a strong pace in your marathon workouts. It’s a lack of endurance, stamina and strength. Anaerobic training, temporary as its effect are, can perhaps provide a quick boost to comfortable pace, but it does not provide any of the long-term effects a marathoner is looking for to improve their maximum steady-state.

“A Plodding State”

As McMillan rightly states, the Lydiard pyramid was originally conceived to serve runners peaking for track (but applied perfectly well to Barry Magee’s marathon Bronze medal even then). Since track was in summer, marathon conditioning would begin in late winter or early spring, and Lydiard employed the cross-country season to ensure his runners did not enter the main training cycle in a “plodding state”. Nobby Hashizume explained this to me in a discussion we had on cross-country on the LF forums:

I do really like the original "Run to the Top" idea of starting out with some cross country running, get some racings down without too much pressure, and then all along prepare your body to a full cycle of Lydiard pyramid for track. In other words, the whole XC season is just a preliminary preparation for track cycle. In other words, by the time his runners stepped off and start doing 100-miles a week, they were already semi-sharp; not a plodding state. That's already been done when they started out on time-based training.

So if you employ the Lydiard Pyramid, you should not have this problem as long as you do a cross-country season as described above (or in “Running with Lydiard”, p. 57-61) or something similar (if far enough from an autumn marathon a “controlled” hill running season could be used in the same fashion).

Also, the aerobic conditioning done in a “Lydiard Pyramid” is markedly different from how the “classical approach” is often construed, e.g. “long slow distance”. While always targeting aerobic development, the key facet of a Lydiard program is that this is not long slow distance (and having finally done it right, I can vouch for the marked difference, but that’s another article). If done properly, stamina and faster running is done from very early on. At the end of their aerobic phase, Snell, Magee and Halberg were running the 22-mile hilly Waiatarua circuit in just around 6-minute mile pace (starting at 2:30 for the circuit, they’d work their way down to 2:10 to 2:15). It’s safe to say that with 200m times of between 22.5 (Snell) and 26 seconds (Magee), they had both speed, stamina and endurance at this stage, but in order to be their best, they had to put it all together in the right order before race day and polish each element off as they went along as well (each in their own way with Snell focusing on the 800m, Halberg on the 5k and Magee on the marathon).

Strength Phase

Another comparison is the Strength phase which Rosa places after his anaerobic training where Lydiard puts it before: The pyramid shown by Greg in the article points out that stamina is one of the primary objectives of the strength phase. This is also the case for the programs on the Go2Lydiard website which includes a “Progress Calibration Run” which is basically a very strong aerobic to somewhat anaerobic run (eventually well over the hour for experienced athletes) in addition to two medium-long and one long aerobic run which are still run at a steady pace.

The two main workouts of this period, however, are two sessions of Lydiard’s hill circuits, which primary objective is to develop strength, sharpen up speed and develop flexibility and power. These weeks are crucial in ensuring a safer transition from the aerobic to the anaerobic phase. Rosa’s pyramid does not seem to offer this, but he may of course have designed the specific weeks in his program so that they introduced certain elements gradually (in fact, given his success, I’m sure he did, this just does not shine through in the article).


I should also say a few quick words on coordination as this phase is not mentioned in Greg’s article but its a crucial part of a Lydiard build-up. Not only does it freshen you up before the race, it also puts together all the training elements you have developed so far and provides another reason why I cannot see the reason to have to move to Rosa’s approach:

A standard marathon coordination week looks like this:

Sun: Long Jog, Mon: Windsprints, Tue: Easy fartlek, Wed: Out and Back, Thu: Aerobic, Fri: Cut-Downs, Sat: Out and Back or Long Race

The Long Jog and Aerobic run serve to maintain the aerobic base you have built. The windsprints maintain the anaerobic capacity you have built (the little you need for the marathon) while the fartlek and cut-downs sharpen and maintain your speed, power and strength. Finally, the Out and Back and “Long Races” (e.g. 10 miles and up) is where you do your “dress rehearsal”, practice your pacing, run at aerobic paces faster than marathon pace and run any others tests you need before the big day.

With this phase in mind you would be unlikely to face the burn-out described by Greg in the article nor leave any element of your training untouched, underdeveloped or neglected by race day or by the time its needed in your training (e.g. when you need speed, it’ll be there).

A Final word on Kenyans

As a final comment, and not to belittle Rosa’s efforts, I see some truth in the assertion that “anyone can train a Kenyan” for the reason that most Kenyans have run thousands of aerobic miles more than the average modern Westerner by the time they reach age Eighteen.

The larger your aerobic foundation the more quality work you can benefit from and there is a lot to suggest that the reason some elite Kenyans can benefit from higher workloads of “quality” work is derived from this advantage.

The Kenyans are also very good at maintaining their aerobic conditioning despite the quality workouts, many training us much as thrice a day (one or two of the workouts being relatively pedestrian by their own standards). Their ability to do this, deducing from my own readings, seems derived from a mixture of a disciplined professional lifestyle, huge aerobic capacity from childhood and reasonable gene pool (the latter being the least important as far as I am concerned).

The training examples – proof in the pudding?

As the absolutely “final word”, I was not impressed by the examples of runners presented, not only would I say most should have done better (that’s for another article) but as a scientist Greg is well aware that anecdotal evidence such as this has no weight in assessing anything and I can only imagine he included it because of constraints of the format with the aim of writing a thought-inducing article rather than a scientific thesis. But for this reason, its worthwhile approaching the content of the article with a healthy does of scepticism.


Ian Spleen said…
Enjoyable post Rene. I think I follow something similar to Rosa. I try not to get technical, the mainstay is always of course the aerobic base.

Start like I said, with the base; then it's faster paced runs with windsprints and moving to the hills if not already on them; I then introduce fartlek and longer races/early season races; then I just build the length/time of the fartlek and do a good many drills (kinda like the lydiard-style hill reps but on the flat - RPE my coach called it). I just build it up then and then I'm into race season.

I found that inserting swimming and cycling as well as 'brick' sessions a la triathletes is beneficial here, and in the base period. It keeps the mind/body fresh and aerobic capacity can be topped up nicely. Also, if cycling on hills you can really hammer away and it's very mountain-race specific.
Renny said…
Thanks for your comment Ian, your approach sounds well-balanced and moves from one phase to the other in a consistent manner. This all reminds us that there are a few ways to skin the same cat as long as the basic principles are upheld and followed.

I will be doing a similar "modification" for my own runners coming into the cross-country season - e.g. the "transition/strength" phase will be slightly different from the hill phase used for the preparation for the summer season. I plan to seek inspiration in Lydiard's original cross-country schedules which Nobby was kind enough to forward a PDF of. I will be sure to write an article about what they look like so people can assess it for their own training.

I believe a runner doing extensive sprints drills and other drills aimed at strengthening core and lower legs can get away with less of the hill drills (as they aim at many, if not all, of the same adaptations).