More on the Marathon – getting to 3:40

I’ve probably over-analysed all this data that I have thanks to my Garmin 305 Forerunner, my record keeping and SportTracks (get this software if you’ve got a GPS watch), but I’m fascinated by what it takes to be a “good” marathon runner. I’m never going to be Kenyan of course, but pound for pound, I’ve run faster this year than I’ve ever run. But, I ran less than half the number of miles per week that the most widely-followed training plans would have had me running and, at most, ran 3 times a week (often only twice) when the programmes have up to 6 days training (with one day cross-training).

That said, I did run far more than I managed last year and way more than I ran in 1999 (the only other year I have data for). Here’s a graphic that shows year to year comparisons by month.

The additional miles I put in are plain from this graph, especially looking at March – the month before the marathon in all cases. I put in 185km this year, 105 last year and 135 in 1999 (3:51, 4:44 and 4:13 respectively). The 1999 marathon was very early in April – maybe April 2nd or something – which is why the distance in April is only 43km.

Taking the date of the marathon out of the picture and looking at the 30 days before the marathon (excluding the race itself), the distances are 142, 95 and 135. The extra 7km between 1999 and 2006 doesn’t account for 22 minutes off my time I’m sure.

Another way of looking at the data shows the step changes between months in each year:

From this picture it’s pretty clear where the faster time came from. I stepped up the training significantly in 2006 versus previous runs. 655km total in 2006 against 466km and 370km. So 41% more km in 2006 than 2005 (which was itself 26% up on 1999 – even though my 1999 time was faster than 2005).

But the distance isn’t enough. A big chunk of this is being able to run fast. My training times at 10km and 20km weren’t nearly as quick as I used to run, but my race times were faster than I’m used to – so there’s something about the competition that makes it happen. In training I broke 50 minutes for a 10k only once (and that was in the Nike 10k last year which is hardly training really); yet in both the half marathons and this year’s big race, I was inside 50 minutes. The half way point in the marathon came in at 1h 47m which was only a little slower than I’d run Liverpool. The second half, though, took 2h 3 1/2m, which shows that I need to do more longer runs to help learn to maintain my speed.

I suspect the weather also helped. Last year, it was hot and sunny. Every 5 miles we ran across a patch of near-velcro as squeezed out lucozade bottles littered the floor making for a sticky mess everywhere. This year it was a good 10C colder and there was a pretty persistent drizzle for much of the race.

I’m thinking of running New York in the first week of November. If I do that, I’m going to be looking for a time of around 3h 35m to 3h 40m, which would get me somewhere in the top 7,000 or so in London (NY is a bigger field so it might put me in the top 12,000 there, I’m not sure of the profile of time yet). I’m definitely not ready to step up to Brad Feld’s commitment to run a marathon in every state before he’s 50, but I admire him enormously for even trying that.

That would mean stepping up the miles a little more, but at least it would be summer with long evenings and some warmth. I think 3h 40m is achievable – for much of Sunday I was tracking to 3h 42m and it wasn’t until the last 4-5 miles that it became clear I wasn’t going to make that.