Tag Archives: age-graded times

What’s the point of a personal best?


“Anyone knows; if you’re on for a PB, you’ll run a bit harder.” The words are Ben’s, chatting before our last Parkrun. They reflect a runners’ common sense, and they are the sort of thing you hear often at club races or on Marathon Talk.

Beneath them, is an assumption about what the typical runner’s “career” will look like. Take a friend of mine, Jane, who is a typical club runner in her early to mid-40s. She began running about 7 years ago, when her best time was about 26 minutes for 5k. Over seven years, she has reduced it to within a few seconds of 22 minutes. Her improvement has been pretty steady: 40 seconds in year 1 of running, a full minute in year two, with no noticeable trailing off for age. Every year, more or less, she has achieved a personal best for the distance and there is no particular reason for anyone to assume that she is going to reach her best ever time – yet.

But take another runner, Charles: he is older than Jane (mid-50s), has been running for longer, and runs further (she runs a couple of marathons a year; he runs 40 miles a week plus, every week, without fail). Over the past seven years he has reduced his 5k time to a best of just over 21 minutes. But as well as a “PB” in one of his last four races, his other three 5k times were a lot slower – between 23 and 24 minutes. Just as his best times have improved a little, his average speed has dipped – because of the time he has been running, and with age.

I know that computer gamers sometimes talk of an “arc of discovery”, by which I understand them to mean that when you start playing the game, there will be lots of skills you acquire. With a relatively small investment of time you will feel that your interest in the game is being rewarded. But a clever programmer will build into the game further accomplishments, so that someone who plays for a long time will continue to have a good reason to maintain their interest. If the initial arc is too shallow, no-one will get in to the game. If it is too steep, no-one will stick with it.

Running can have its own ark of discovery, but if this about beating your own best time, it should be pretty obvious that someone like Jane, even though she has been running a while, is likely to keep on hitting “pbs” in future (if, perhaps, not quite as often as before) while for someone like Charles, simply because of the amount of time they’ve been running and their age, there will probably be fewer pbs in future.

Moreover, I suspect that Jane is unusual in that she has had seven years of pretty consistent improvement, and maintains further room to improve. Compare Jim who seven years after taking up the sport is a sub 2:45 Marathon runner, and whose age graded times I’ve listed below:

(NB: the Y axis shows races, listed by month and year. Each dot is a race. They are a range of events, from 5k to a Marathon. I only have access to this information as age-graded, not absolute times. The effect of age-grading is equivalent to about 0.5% p/a. I.E. if a runner’s annual pb for a distance remains static over a 5 year period age-grading would lead to an “automatic” improvement of 2.5% or just a little more than one of the bands 76-78%, 78-80%. I. E. of Jim’s improvement, only around a third is down to age grading, the rest is sheer improvement in Jim’s actual times).

What I see here is about 4 years of dramatic improvement followed by around a year of consolidation. And I suspect this arc is pretty¬† commonplace: that people who take up running seriously (at any stage) “should” improve naturally for about three to four years. Whether they carry on improving beyond that stage will depend largely on their age – if they’re still under 25, they should have some age-related improvement ahead of them – if not, then they will probably “plateau”.

This in turn begs the question of why someone like Charles keeps on running. I don’t know how long he has been running, but my best guess is that it has been for more than 20 years. My sense is that for many lifelong runners their motivation is maintained by a much wider set of factors than pure time. It might be that they enjoy running with friends. It might be that they want their club to win certain kinds of races. It might be that they experiment with running a wider variety of distances (including simply, longer distances: much as I could never participate in the ultra running boom, I know it has to come from somewhere).

I don’t think that even someone in their mid-50s who has been running for 40 years or more ever quite loses the search for “the perfect race” (to which the personal best is of course an approximation), but I think that their notion of perfection alters. It might be that self-transcendence is about a passage of five minutes or an hour in which just everything “clicks” – rhythm, pace, environment.

A runner is in much the same position as a studio musician. We spend our lives traveling from race to race compelling our friends and running companions to endure endless “jams” or mediocre albums in the hope of a unique moment of musical / athletic transition.

A decade running; in a graph


The above picture shows what my 30s have looked like in terms of running times. The dates at the bottom are slightly misleading (the last red dot in the left hand bunch should be for a race in March 2005); and this is is not a complete picture anyway (eg it excludes my most recent time in the July 2012 summer league 5 mile, which I’m rather proud of – more on that shortly). The big long gap requires explaining: it begins when I was injured, just two months before my eldest son was born.

These are age-graded times: if my race time for say 5k had remained constant over this period, you would see a line sloping slightly upwards (which in a sense this is): a record of maintaining similar times against the negative effects of ageing.

Many of the consistent club runners I know have a very different patter; with more races and a much clearer sense of progression (the greater the progression, the greater the incentive to keep running); the following is relatively typical:

(What I’d give for five years that looked like that…)

Finally, this a chance to reflect on my last competitive race – a 5 miler two weeks ago for the Mornington Chasers as part of the Summer League. I finished in about 38 minutes 20 (age graded = 58%) with splits as follows: 7.28, 7.32, 7.42, 7.37, 7.59. The race was rather more dramatic than it sounds. Over the last mile, I accelerated a lot, taking myself down to about 6 minutes 30 pace or maybe faster. It was a lot of fun, except that I tore a muscle in my right calf and had to walk the last 400 metres.¬† I’m desperately hoping to run a similar distance (10k) today – slower, but without injuring myself. If you look back to the top graph, imagine an extra red dot, somewhere further to the southeast of the last one…