How To Predict Your Mile (0r 1500m) Time

Got a 5K or 10K time? Here's how it translates


Posted: 6 May 2002
by Sean Fishpool and Steve Smythe

It’s useful to have a realistic target for your mile, because much of your speedwork will be paced in relation to it. Here are three ways of finding that target, both for speedwork and racing.

1. By feel
The best option for beginners. If you’re running 400m repetitions on a track, for instance (the ideal venue), simply ask yourself ‘could I complete three more laps at this pace?’ Be realistic, and bear in mind that your speed will naturally improve over the next four weeks.

2. By specific sessions
In a mile race, you should be capable of maintaining the best pace you can hold for 4 x 600m with 5-minute recoveries (see Week Three of the ‘Solid Mile’ schedules,). Alternatively, some runners use 3 x 800m with 3-minute recoveries as a predicting session, which gives a more conservative target.

3. By previous race performances
How closely your 5K or 10K times relate to your one-mile or 1500m time depends on your natural inclination towards speed or endurance, but the figures below should be accurate to within 30 seconds.

 

5K 10K 1500m Mile
15:00 31:30 4:00-4:10 4:20-4:30
17:00 36:00 4:30-4:40 4:50-5:00
19:00 40:30 5:00-5:10 5:20-5:30
21:00 45:00 5:30-5:40 5:50-6:00
23:00 49:30 6:00-6:10 6:20-6:30
25:00 54:00 6:30-6:40 6:50-7:00
27:00 58:30 7:00-7:10 7:20-7:30
29:00 1:03:00 7:30-7:40 7:50-8:00
31:00 1:07:30 8:00-8:10 8:20-8:30
33:00 1:12:00 8:30-8:40 8:50-9:00
35:00 1:16:30 9:00-9:10 9:20-9:30

Previous article
Do-It-Yourself 5K Schedule
Next article
The Alternative RW Mile Schedule

mile
TwitterStumbleUponFacebookDiggRedditGoogle

Discuss this article

has anyone used the 4*600 with 5 mins between each as a 1500 pace predictor session as shown in the predict your time article? the 5 mins between each rep seem quite long on the face of it and would lead to an overly ambitious prediction? was going to try this session out tonight to see where i am at the moment!

any other good predictive sessions people are willing to share? One good one I have found useful is one of Frank Horwill's which is 400 (45 secs) 800 (1 min 30) 300. 3 mins recovery then repeat. average of the two times for the sets is a good indicator of current 1500 time
Posted: 01/03/2005 at 13:49

Are you racing indoors?
Posted: 01/03/2005 at 14:04

no, never have done. 3000m track race coming up in March
Posted: 01/03/2005 at 14:27


soc
I don't think it would lead to an overly ambitious prediction - the 600 reps are quite long and you're totalling 2400m anyway which probably would compensate for the long recovery.

200m flat out - x by 2 and add 4 seconds = 400m race predictor

There is a really good 800m one -

run a 400
wait one minute
run a 400

then have 10 mins off and do the same thing again

your 800 time is the average.
Posted: 01/03/2005 at 14:31

a good 800m one for me is 400m, 30secs rest, 200m, 10 min rest do it again and take your average time.
Posted: 01/03/2005 at 14:37

excellent to get these ideas! let's broaden the thread for the 'longer' MD runners amongst us! Any good 3k/5k guides? Doing my first 3k in March and would like to gauge if sub 9 mins is possible
Posted: 01/03/2005 at 14:53

well i did the 4*600 and I would stand by my gut feeling that the time it predicts is a bit ambitious. In the cold and rain I did 1.36,1.39,1.40,1.41 which corresponds to 66 second 400ms and a 4.07.5 1500 which at this stage of the season would be beyond me (PB last year 4.10). Unless I'm in better shape than I think!
Posted: 01/03/2005 at 21:12


soc
hmm, I'm not sure. I think these predictors do have flaws though, and the longer you go, the more inaccurate they are. I don't think I'd have the endurance to hold the 400/400//400/400 session to predict my 800 time right now, but could pull off an 800 race just about. However, in a few months the session might be a better predictor, which makes me think they are quite variable.

Still, you must be pretty pleased, though! How about trying one of the other 1500 predictor sessions, and seeing how it compares?
Posted: 01/03/2005 at 22:23

i did the 400,900,300 * 2 one mentioned above a couple of weeks ago and it predicted a 4.13 which i reckon is more realistic at the mo!
Posted: 02/03/2005 at 09:02

We'd love you to add a comment! Please login or take half a minute to register as a free member

Smart Coach
Free, fully-personalized training plans, designed to suit your racing goals and your lifestyle.