The RW Race-Time Predictor

2456710

Comments

  • jenny, have you measured the speed of the treadmill at your 5k and 10k pace? I wonder if the speed is more accurate at one than the other? Though thinking about it, the lack of cooling effect/wind resistence on the treadmill should have different consequences at different speeds, so maybe that could account for it?
  • Foo Bar, no I haven't measured the speed, I've just taken the readings as correct. I think there is too little difference in speed for factors like lack of cooling (it's in the conservatory, windows and doors open plus fan!). I think it could be more to do with stride length. Faster speeds favour longer strides and it's much easier to do that outside. The treadmill automatically shortens your stride which would favour longer distances.

    I'd be interested if someone else did this experiment - compare 5 and 10k times road to treadmill on 2%. Does RW have an opinion on this?
  • Don't suppose you've recorded your average heart rate for road versus treadmill 5k and 10k runs?
  • Using my 10k time as I've run a few (well 4) over a few years and they are all pretty close in time (within 7 seconds of each other).

    My 5k time is about 10 seconds down on what it should be according to my 10k time, my half marathon time is 5 minutes 30 down on what it should be by my 10k time, my marathon predicted time is 2.52 ! - I've only run one which I did as a training run but I reckon about 3 hours is more realistic even when I was in form.

    I do have a better 5 mile time than 10k which would throw it even further out but there is some doubt in my mind about the accuracy of the course for that one because it is so much faster than any other race I've done - though I was training more at that time.
  • It does seem to suggest that I need to put in a bit more endurance work. The 10k/half marathon comparison is really close, I'm only about 30 seconds off. My marathon time was way off, but that's not surprising as I have only run one and almost pulled out after losing almost 6 weeks training to injury and illness and crawled home in about 3:20. Encouraging to think that with the right endurance training before my next marathon, I should be able to knock at least 12 minutes off my last time though!
  • Foo Bar, no sorry! What are you suggesting? I'd like to think I work just as hard for both though!
  • completely off for me, I put in my 1500m time for last season and it reckons I should be running sub 34 for the 10k and I'm down to about 35/36, put in my most recent training time for 1500 and it reckons I should be doing almost 32 minutes.

    About right with my 5k, 10k and 10 mile times though, my 10 mile time is a little slower comparatively but I've only done one. I'd say the calcs a little on the fast side though.
  • It better do - I will let you know on Sunday!
  • My 5k time predicts my marathon target time which is nice (but 16 minutes faster than current time)and is faster than my current times at all distances. Putting in my marathon time isn't too far out down to 10k but is significantly slower than actual 5k and 5m. I think I probably am biased towards the shorter distances though.
  • Works for me, comparing 10km, 10mile, 1/2 marathon and marathon times. 10mile from 10k was bob on. marathon was within 3 mins of actual. I think I might pinch that formula! Cheers
  • Sub 3:30 marathon - aye, that'll be right(!)
    Nice if it was though.
    The rest seems to work quite nicely though.
  • Seems to work pretty well for me. 48 min 10k race one weekend followed by 1 hr 21 min 10 milers the next - spot on by the calculator. Marathon at around 3 hr 45 though - I wish! What it really means is that I didn't train well enough when I did FLM in 4 hrs 44, must work harder for next year.
  • pb 1:49 1/2m, says 3:47 for marathon. My marathon pb is 4:47. The formula should take into account the significant slowing down most of use do towards the end of a marathon. That doesn't happen at < 15 miles.
    If I get in the FLM, I'll try for sub 4 hours, but 3:47 ? Dream on !!
  • works for me too from 2m to 1/2m, just trying to decide what pace to aim for at London and got a fright as it predicts nearly 5mins faster than I planned (2:32!) not sure Im feeling that do or die!
  • agree with Bryn. If I put my 1500 time in it comes up with some ridiculously fast times for 2 miles, 5k,10k. Or am I just underperforming?!

    seems to work better for my 5k,10k and 10 mile times when i put those in though
  • Well, based on my 10k PB my actual marathon time was about 45 mins shy of my potential!

    I know that already though - not enough endurance.
  • WrintyWrinty ✭✭✭
    Well that is the best thing that I have seen for a long time, when I put in my best Ten time for this year it gives me my best Half time for this year, within a second.

    Both times give me a sub-three hour marathon time, that I shall hold RW too when I attempt one in January 2006.
  • My Manchester 10K time of 49.27 predicted 57mins less than I did the FLM.
    If you extended the calculator to collect data into a database and include the average length and time of training runs then it may be possible to guestimate the optimum training distance for each race length.
    Most of my training was been 5 and 10 miles over the winter, no surprise with hindsight, that from the Tower to the finish I averaged just less than 14min miles.
  • I find the Reigel formula works well for using a 5k to predict a 1/2 marathon or a 10k for a 3k. it is generous when predicting marathons unless you are a highly trained injury free mentally tough runner with less than 10% body fat! http://www.runningforfitness.org/ uses 5 different formulae and gives an average prediction too. I ran 10k 2 weeks before Berlin this year and used the 10k to predict my marathon time. my marathon was 90 secs slower than the average prediction but I had to go to the toilet during the race which took.........90 secs!! spooky
  • I thought this was the same formula that is on the MacMillan site, but it gave me an odd result.

    Putting in my 10 mile time of 1:08:30 it gave me a time for 13.1 miles of 1:20:32!

    This implies that have done 10 miles at about 6:50 pace, I'm suddenly going to do the next 3.1 in well under 5 min/mile!

    Wish that I could...
  • http://www.runningforfitness.org/ says you'd take anywhere from 1:30:58 to 1:31:32 (had to guess your age from the picture though!)
  • It doesn't seem to handle 08 or 09 in the minutes field. If you ask it to calculate your 10 mile time from your 10 mile time, you'll see what it does.
  • Ah, I see what it's doing.
    According to the documentation of javascript's parseInt() function, "If the string starts with a 0 it will be parsed as an octal number." which is fine for 0-7 and 10-99, but fails on 08 and 09.

    To get it to work correctly (before RW remove the bogus call to parseInt), just put 8 in instead of 08.
  • thanks Foo Bar - we're on to that now
  • No! It doesn't work for me, at all.

    Even predicting times from races that are one step away in distance are over 10% out! All the discrepancies are in the same direction, though.

    However, it does predict my 10K time from my 5K time quite well, which is funny as I've only run 10K once so thought that would be the least accurate!

    Then again, it does predict my 10K time as 31:50 (WHAT?!?!) from another time, so I just don't get it.

    Laura
  • It seems pretty accurate for all distances apart from Marathon - where it's too generous (of course there is the possibility of slightly inadequate marathon training!). Although, I have experimented with many different calculators (including runningforfitness.co.uk, marathonguide.com and runnersworld.com) and find that this is a recurring problem, both for me and my other half - predicting an unrealistic marathon time. The most accurate was the one published in the RW Race Guide handed out in the FLM pack this year - which was disappointingly spot on! I ran a 1:39 half and 2:38 for 20 miles, was therefore hoping for sub 3:30 (based on predictors) at FLM and did 3:40. The RW guide predicted 3:39! whereas all others (including this one) are around 3:20.
  • Unfortunately I don't have enough race info at similar fitness levels to evaluate these predictors. However, I came across runpaces some time ago which performs predictions based upon your individual performance curve. The link in the page doesn't work, but you can download it here. I'd be interested in people's feedback of that one.
  • Clearly the closer the race distance the more accurate it will be. i.e. predicts 10k - 1/2m more closely than 5k to marathon. But based on my 20 mile time it gives me a marathon time 10 mins faster than i did only 8 weeks later.
Sign In or Register to comment.