The training method I advocate is unusual. Firstly, it doesn't involve any traditional "sessions"; it's just running. Forget about your 800m reps and go for a run. The best way to train for a marathon is simply to run. Lots.
Secondly, I advocate a slow and steady training style. If you're running so hard you can't hold a conversation then slow down. Have a chat with whoever you're running with, enjoy the scenery around you; there's no need to be out of breath.
When you put the two of these elements together you get a training style of regular runs (often more than once per day), which are usually with friends for a chat, for some fresh air during lunchtime, or to get from A to B e.g. as a commute.
As an example of how this might actually be applied to a runner's training, imagine an athlete aiming for a 3h marathon. While many people aim for a sub3 time each year (and don't make it) I claim this result is achievable for any runner. On way of getting into sub3 shape is to train as described in the table below. Other than Saturday, all of the running is done at 5:00min/km.
|7.5km before work
10km at lunchtime
|5km in 19min at Parkrun||17.5km (anytime of day)|
The science says that by averaging this every week, you'll be in 3h shape. Of course, this is just one example of how a runner might train, there's an infinite number of other ways to structure this training load. The point is that training shouldn't be complicated - it's just running.
However, there’s more than one way to skin a cat, and there’s more than one way to train for a marathon. So, if you want train in a more traditional way, that’s fine. JSET is a great tool nonetheless. It’ll tell you exactly what training load your body has experienced each week, and how much training you need to achieve your goal.
How far do I have to run? How accurate is JSET? How fast do I need to run? Find out below.
The original equation that JSET is based on claimed to be able to predict marathon performance to within 4min, only by knowing average weekly distance and average pace over the eight weeks before a race. However, this used data from a relatively small group of runners. In reality I've found the original equation has a consistent offset. For me it's -14min (I can always run 14min faster than the original equation predicts). For another JSET runner it's +10min (He's always 10min slower than the original equation predicts). Once you know your offset then you can use JSET to predict with exceptionally high accuracy. Most the time your offset can be calculated by looking at your training data over the last few years.
That depends on your goal. If you're aiming for a 3h30min marathon you might only need to run 4 times per week, but if you're aiming for 2h30min you'll need to run every day. The points system makes it easy to know how much you should be running. Once you've met your points goal each week, put your feet up. Until then, go out for a run.
No. You've probably heard a lot about how important things like "threshold runs" are. That style of running isn't bad, but it's by no means better than running at a steady pace for a few extra miles.
Absolutely! There are numerous way to fit running into a busy life. If it's possible you might want to consider a run-commute. Many people go for a few miles at lunchtime. These are just two example of easy ways to rack up the miles without using too much time from your non-running life.
Yes! There's plenty of data and science to support the theory as well as lots of examples of runners who've improved their times dramatically.
The equation was originally designed for people completing the marathon between 2h45min and 3h30min, but I have found it works just as well up to 2h28min. If you're looking to run slower than 3h30min, the equation becomes slightly less accurate (because lifestyle factors play a bigger role in your training) but it still works.