Saturday, October 04, 2008

ARTICLE: Using Target Pace Levels

Running guru Matt Fitzgerald, of Runner's World fame, introduces a new concept in his ground-breaking book "Brain Training for Runners" (not to be mistaken for "mental training"), called Target Pace Levels (henceforth referred to as TPLs)

Pace Tables
This is not a completely novel concept and essentially look the same as the "Pace Tables" or "Pace Charts" you can today find in almost every running publication today. I'll reject the definition of Pace Chart as it is semantically incorrect; charts are maps or graphs, while a Pace Table comes in a matrix format (a rectangular array of elements arranged in rows and columns).

A "Pace Table" bases itself on the scientific finding that running performance over a shorter distance is the most reliable predictor of racing time at longer distances (e.g. your 5k time is a good predictor of your marathon time). This appears to hold true for distances of 800m or up, but certainly from 1500m and up (sprinting times have proven less reliable as the important performance factor of fatigue resistance does not come into play).

Timothy Noakes has presented some evidence that even sprinting time can predict performance in endurance events, but only in endurance athletes not in trained sprinters (as their hearts and muscles are not adapted to endurance running despite their great speed on 100-400m distances) (Noakes 2002).

Because of their purpose, Noakes prefers the use of the term "Prediction Tables" and presents the 5 models that have been developed so far in "Lore of Running": Davies-Thompson, Daniels-Gilbert, Osler, Mercer-Leger-Desjardins and Gardner-Purdy. The tables are similar but not identical, and this highlights the necessity to approach the Prediction Tables not as literary truth, but rather as solid rules of thumb.

Target Pace Levels - The Scientific Reasoning
Matt Fitzgerald modified Daniel's original Prediction Table into his own Prediction Table and dubbed it "Target Pace Levels". Matt worked with his employer TrainingPeaks ( to adapt the extensive findings of Daniels to the changes mode of thinking resulting from the Central Governor Theory (e.g. that fatigue is not caused by any hard physiological limit but is instead a sensory manifestation of a pacing strategy of the brain, known as teleoanticipation (Fitzgerald 2007)).

This changes the goal of your training sessions (but not necessarily the format): Instead of attempting to push back physiological causes of fatigue you simply attempt to gradually increase the speed you can sustain over race distance as well as the total duration you can sustain race speed or faster (Fitzgerald 2007).

So in Fitzgerald's TPL system the intensity-based pace levels prescribed by Daniels are replaced with performance-based levels (e.g. instead of doing your tempo at "anaerobic threshold" training, these would be done at "half-marathon" or "10k pace"). Incidentally, these pace levels are not radically different (Fitzgerald shows that the AnT lies between most people's 10k and 1/2-marathon speeds).

The key advantage, Fitzgerald argues, is therefore in the meaning of these targets. Instead of training at some threshold (and firm thresholds don't actually exist, as Noakes has shown conclusively in "Lore of Running" (Noakes 2003)), you're training at the exact pace you plan to use in your races. This trains the brain to run precise simulations of what it will need to do on race day (or the Central Governor, which may, in fact, not be located exclusively in the brain. Modern science has shown that the brain may not be the sole "intelligence" in our body, but rather the whole body is more likely one big interconnected information system).

Now the scientific reasoning aside, let's look at some practical benefits of TPL vs. Daniel's older system.

Target Pace Levels - Practical Advantages
Fitzgerald's system is extremely easy to use as it divides runners into 50 categories, TPLs ranging from 50 (back of the pack) to 1 (world class). He unfortunately only provides 5k, 10k, 1/2 marathon, and marathon times as a means to pick your TPL.

Going by my own 5k time last year (18:48), I would be a TPL of 26, while my marathon time of 3:18:48 suggests a TPL of 32. If I correct my marathon time for the distance I ran (42.5km rather than 42.1km), my real time is 3:16:56 which is still 32.

My only "10k" is the 10.5km intermediate cross-country, not a great predictor, but if I translate my time loosely, it would put my TPL at 30. My half-marathon time from Connemara is 01:34:06, which translates to 32 as well. I choose to disregard this result, however, as I was running as a pacer, unprepared, and carrying water and gels for the runner I supported and was therefore not expending 100% effort. You should apply similar criteria.

Why do they come up different, you may ask? The simple answer is probably that I was in better shape for some races than others. In general, you should choose the most recent race you ran that you trained specifically for. If you run a 5k in about 18:40, you should be capable of a 2:58 marathon according to the charts. However, while this is your current fitness level, it will take specific marathon training to achieve this target.

This, interestingly, explains the conundrum that theoretically, someone who is faster than you over 800m, should also beat you on any other distance. When this is not true, it is simply because the 800m runner is not training for the longer distances and is thus not adapted for endurance at the time. Should he train exactly like the endurance runner, he would be faster on all comparable distances (to reiterate: this seems to hold true at least up to the marathon, but in many cases also up to the 100km distance, as exemplified by Rafael Salazar's stunning win in the Comrades Marathon in the early 90s. Beyond 100km a new set of physiological factors come into play that cannot be determined from performance in shorter races).

As an anecdote, when running my first half-marathon (01:38:28) more than a year earlier, after having trained in a structured manner for 3 months was 34. My "untrained" level is therefore probably somewhere in the range of 38-40. Hopefully, tests will confirm that even in my current "detrained" state (detraining is periods of inactivity rather than recovery following exercise resulting in an athlete losing training adaptations ("Essentials of Strenght Training and Conditioning")). Detraining is little understood, and while athletes seemingly return to their untrained state, some adaptations, such as muscle mass, do not return to pre-training levels. Science has much still to explain for the phenomenon of better athletes returning quicker to shape (this could be an effect of favourable genes, better training attitude, and the new Brain Training paradigm's theory that the brain will remember previously completed tasks indefinitely).

Picking Your TPL
As a hill runner, I only have the few times above to base myself on, and as I've been out of training for almost 3 months, injured, I chose the lowest TPL: 32. I then looked at the times it would predict for me on the 5k, 10k, half-marathon and marathon distance and tried to gauge from my first runs if they sounded realistic (common sense should not be ignored when picking TPL, runners have a strong intuitive sense of their own performance limits).

20:39 (4:08min/km). Marathon pace of 4:40 is hard for me at the moment, so even this relatively modest time would be a stretch.
10k: 42:50 (4:17min/km). Would have been a very comfortable pace 3 months ago, but a slight stretch over 10k at the moment as my endurance over distance beyond 5k is almost completely eroded
Half-marathon: 01:34:53 (4:30min/km). I've run faster than this even with a backpack supporting Aoife in the hilly Connemara Half-Marathon, so at first a very easy target, and it wouldn't take much base work to bring me back to a comfortable 4:30 running level.
Marathon: 3:17:29 (4:41min/km). I managed 30min at this pace in my 5th run back, I should be 4-8 weeks away from this level.

Using these predictions, I decided to adjust my TPL to 33. Why? Here' s why!

When to Change TPL
Once you have your TPL, Matt Fitzgerald prescribes a training speed for Recovery Pace, Base Pace, Marathon Pace, Half-Marathon Pace, 10k pace, 5k pace, 3k pace, and 1 mile pace.

When you consistenly start to outperform your pace targets in your workouts, you should adjust your TPL downwards one level and use the new (and faster) times. You should expect to change it down once every 4-8weeks but should only do so when you exceed your targets consistently.

My end-target is running a 10k in sub-33min which corresponds to a TPL of 15 (someone like Leinster League champion Barry Minnock would be a TPL of 11 or 12 for comparison purposes). If you're interested it takes a 13:26 5k, 27:59 10k, 1:01:34 half-marathon or 2:09:02 marathon to "earn" TPL 1. (Haile is beyond this system clearly!).

That's a reduction of 18 TPL levels that I am looking at or equivalent to 72-144 weeks of training. Given that at least 4 weeks each year is recovery, it means 80-152 weeks from now, translating into 1.5-3 years. In reality, it's probably more, as running follows a law of dimishing returns.

The closer you get to your maximum training volume, the less speed you will gain in return (e.g. you may improve by 20% by moving from 40 to 60km per week, but only 5% by moving from 60km to 80km, and only 2.5% from 80km and 100km. Conversively, if you move beyond your genetic training capacity, you will start to regress, e.g. moving from 100 to 120km could cause a performance reduction. The best athletes are believed to be those who have the genes most suitable to adapt to the highest levels of training.

Next season, the best I can therefore hope for is a TPL of 21 or 22 translating into: 17:17 5k, 35:52 10k, 1:18 half-marathon and 2:45 marathon. Now obviously, I won't be physically ready for a 2:45 marathon, even if the genetic capacity is proven to be there (by reaching the target speeds above), and indeed, I don't plan to do any marathons next year. as my body hasn't adapted to competitive running at that level yet and will need another 3-5 years to build resistance.

The 10k target is therefore my main one, and the 35:52 suits well with the 35min point I had hoped to hit by 2009 when I laid out my 5-year plan in 2007. An intermediate target of mine is to return to the 26 TPL level that I had reached in late 2007, and after that start to focus on TPL15.

It's my gut feeling that my genetic limit lies at that level. I feel, given my 5k debut, that I've got a 16min, 5k in me (TPL 15 is 15:54), but this is, of course, more guesswork than actual deduction.

It's my hope that this walkthrough has given you enough of an overview of the TPL system to go out and buy "Brain Training for Running" to refine your programme. Don't be left behind in the old paradigm. You may already be doing most things right, so let this new visionary programme cut the final edges of your existing training regime.

Interestingly, a TPL of 15 is equivalent to a marathon time of 2:32:35,Would that be realistic later? No, I don't think it is. My fortitude towards injury is not sufficient for the training required. I do think that given my proven genetic speed, that with a more modest training amount and 5-10 years of consistent running, 2:45 is a target that I must set myself. Anything else would lack of ambition and given how important a factor belief is in running, not recommended.

No comments:

Search This Blog