The RW Race-Time Predictor

[Deleted User]

Interestingly (for me anyway), by entering the PBs of Steve Jones into the calculator I got this:

5k best of 13:18 predicts 13.1miles in 1:01:07 and 26.2miles in 2:07:26.

10k best of 27:39 predicts a 13.1miles in 1:00:56 26.2miles in 2:07:04.

Jones' actual PBs for the half marathon and marathon are 1:00:59 and 2:07:13 - so quite accurate then.

[Deleted User]

I took the Lake Vyrnwy half marathon and fiull marathon results for 2004 and found 14 runners that had done both. The full marathon is twice round the lake and a fairly flat route - good for PB's ! If I input each half time to get a 26.2 time then 13 of them had predictions about 10-15 minutesw faster than the actual. Going in the reverse direction, I input the 26.2 and got 13.1 predictions. 13 of the 14 gave predictions that were longer than the actual, the average factor was 0.91, so the average error was about 12 minutes. On this basis I'd say that the predicor needs a bit of tweaking to get it more accurate. If anyone wants the comparitive times to try the maths, I can happily email them. I'd be interested if anyne could do the comparative results based on other real races. smileplas@aol.com

[Deleted User]

my 10K PB is 45:25 so it reckons I should be able to do a marathon in 3:29...I'd planned next years FLM with a target of beating 3:45 (my time last yr was 4:07) - should I changing my training to be expecting to do a 3:29?

[Deleted User]

When I try to use the run pace thing this log in box comes up >help!!!!!!!

[Deleted User]

Reading through this thread it is clear that it is the marathon times that are most commonly out of sync. Of course it is a tough event, people tend to run them less frequently than shorter distances and there is more that can go wrong on the day, but I'd still argue that most people do not train enough to fuffill their potential at the marathon. Colin's analysis bears this out - rather than proving the predictions inaccurate.

[Deleted User]

The general consesus seems to be that if you train properely for marathon distances then you can achieve what the race predictor suggests. However, most people don't do enough endurance training to reach this level and so take longer over a marathon.

I am doing a marathon in April and my previous 10K and Half Marathon times of 44:09 and 1:43:13 suggest a Marathon of between 3:23 and 3:35.

If I'm starting from a base of 25miles a week and aiming to peak at about 50 miles/week, do you think I could reach that level or should I be more conservative and aim for more like 3:45 ?

[Deleted User]

erm well this aint right now is it..10k time 46 mins... marathon prediction time 209 hrs 41 mins 26 seconds!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!

[Deleted User]

Perhaps if you put 46 in the minutes field instead of the hours field, it might predict something a little bit quicker. By the way, there are 26.2 miles in a marathon, not 26

BrynRunning

lol, Bazza, maybe in 10 years or so, for now I think I'd have a major struggle breaking 3 hours, even with some marathon only training.

Steve Zx, at least you'd probably pick up a record for longest time continously run, all you'd have to do would be finish.

Actually that's not a bad idea, run a marathon in the slowest time ever and when someone asks you what you did you can say you broke a world record

[Deleted User]

Picking up on the exercist, the predictor uses a factor of 6% to allow for th greater distance, but it seems to be that this shouldnt be constant. For example it may be right to use a constant factor to go from 10k to 10 miles, but to go from half marathon to marathon brings in more factors, i.e. hitting the wall and at that point times deteriorate far quicker due to all the biological reasons that marathon runners know about.

So if you imagine a graph the line should be a curve not a straight line for performance against distance.

Am I being an anorak here???

[Deleted User]

Nowt wrong with being an anorak RH3! I think that a curve, more logarithmic, than a linear relationship is valid. It needs to include factors such as body size/type and metabolic rate i guess, but they are difficult thigns to factor in i guess.
3 weeks before i ran FLM i did a half in 1.53 and did FLM in 4.14.43. Not that i hit the wall or anything, maybe could have run faster a bit faster but not 20mins faster. But it also works the other way, and i geuss this revolves around the types of muscle fibers you have (fast vs slow twitch) i cannot run fast for a 10 or 5 k. Cant get under 55mins for 10 k or 25 for 5k...where technically i should be able to/.

[Deleted User]

I took the Vyrnwy results again (see above) and found that if you use a power of 1.15 instead of 1.06 then you get a result which is far closer to the actuals when going to marathon from half and vice versa. Sean reckons that most athletes don't train hard enough for the marathon which is why the 1.06 factor gives a much faster marathon time than most people seem to get. If you apply the 1.06 factor to some of the worlds best long distance runners (i.e. from IAAF stats )then the predicted marathon time is maybe a minute faster than their best. eg Tergat 2.03.36 v 2.04.55 actual. This bears out Sean's suggestion, so if we all trained as hard as Paul Tergat, then we would hit the predicted marathon times !! I wonder if anyone of us normal average (and older) club runners were to train this hard whether we could hit the figures, given that our half marathon times would also drop - any takers? I am sure Sean would sponsor someone with a suitable HRM/distance meter to prove they'd done the training ?

[Deleted User]

Just used this with my recent 10K and 10 mile times. Last 10K was in 1:09 which predicted a 10 mile time of 1:54:17. I came in at around 1:54:37 so this was pretty accurate! Doubly pleased as this 10 mile time predicts a 20 mile time of just under 4 hours which means I should make the cut-off point - just!!!

:-)

[Deleted User]

The predicted marathon time is a bit pessimistic for me based on my 5K and 10K times (15:36 and 32:20 give me 2H29 and 2H27, whereas my pb is 2H24). However my half-marathon time (69:36) gave a prediction which was spot on (2H25'08", within 20 secs of my pb).
By no means do I run big mileage (80 miles a week av.; up to 100M for the marathon), so I'd say it's pretty accurate.

[Deleted User]

I ran 5K about 2 weeks ago, then 10K last sunday. I put in my 5K time and it predicted I would do the 10K 3.5 minutes faster than i actually managed. Am I slow or what??

[Deleted User]

Muddy, I'm confused, you say you can't get under 55mins for 10k but ran a half in 1:53? Am I missing something here?!

Only interested as my half and subsequent marathon times for last year a practically identical to yours! Got my 10k down to 48m50s this year and aiming for a sub 1:50 half in March.

[Deleted User]

Yeah,
i don't know why. I do do speed work, but i just don't seem to be able to judge it right for shorter stuff. I think i start too slow, forgetting that i'm not actually going to have to run an extra 7 miles. But at the moment my leg turn over is shockingly slow and can only seem to plod.
I did run 27 mins for a 5k (with a dodgy knee) but this would give me a 10k time of 56 mins, rather than the predicted 51.15 or 24.35 for 5k.
Basically i think i#'m just a bit odd!

[Deleted User]

Kerry, remember the predictor doesn't take into effect tough courses or having an off day. And you could look at it the other way round - that compared to your 10k time you're really speedy at 5k!

On a serious note maybe you should look into more stamina work e.g. running slower for longer to get used to the extra distance?

Good luck :-)

[Deleted User]

Muddy, even if you kept the same half mara pace for the 10k you'd get in at 53:30.

Why and how do you run a slower pace for a shorter distance?!

[Deleted User]

Well I am totally odd.

I ran a 1hr 51 10 miler and according to the calculator I should have got a 67mins 10k - I got 61 mins, according to the calculator if I can do a 61min 10K I should have got round 10miles in 1hr 41mins. hmmmmmmm

[Deleted User]

Seems not too bad for me:

10K actual time 44:09 -> predicted Half Marathon 1:37:19

Half Mar. actual time 1:38:15 -> predicted 10K time 44:34

(both races reasonably flat when fitness was similar)

Interestingly, the McMillan race time predictor:

http://www.mcmillanrunning.com/Running University/Article 1/calculator3.htm

-gets it spot on for me. i.e. 1:38:15 HM -> 44:09 10K

I'm doing a marathon in a few weeks and this predicts 3:27:13. I don't think I'm capable of that, but might try for 3:40 ish.

It'd be interesting if when people write down how well their times compare with the predictions if they said how much training they did for that distance.

e.g. at the moment I'm doing 40mpw, with long runs up to 20miles and some tempo work.
Is that enough endurance work to make my marathon predicted time? I expect probably not, but for people who get actual marathon times equal to predicted times, how much training did you do?

[Deleted User]

Blimey, it says I can do a marathon in 3.44.

I haven't done one yet as first is in 5 weeks, but there is no way I would come in in that time!

I think 4.15 is more realistic, but maybe I'm being conservative then?

Anyway I'm not going to worry about a time, but was just curious!

SamMurphyRuns

I think the formula gives predictions that are a bit overambitious. It predicts me a sub 3.15 marathon but I've not ever run sub 3.30 (though am hoping to this year). And if I put my marathon PB in, it predicts I should run a half marathon 10 mins slower than I actually do. Like they say, running is an art AND a science...

Pammie*

Well - i wonder

My best performance to date (in this running lifetime that is) is a 10km - 51:40 (flat course but coastal)

I have a 15 miler on Sunday (4 sizeable hills, the race blurb tells us, otherwise a fast course- we'll see)

Anyhow
RW says 2:11:28 (yes i've put in long runs but nothing over 11 miles in the last 5 weeks)
Mcmillian reckons 2:12:36. I guess i'll know if any of this is true around 12:43 on Sunday afternoon

Wayne Kerr

entering my recent 10 mile time 0f 65.10 I was predicted a half marathon time of 1.26.30;actually ran 1.28, Predicted a marathon time of 3.03 which does seem a little ambitious

Final Mile

i'm training for Halstead Marathon in May, and am (fairly) faithfully following the Runner's World 3.30 schedule to hopefully get round in 3.45. My intended time is based on my previous marathon history (I've only done one previous marathon last sept, with 4.00 pace group - we got round in just under 4h). The thing is, i did a 15.2m run this morning in 1h48m30s, knocking 6m off previous best! Now to the point... the predictor reckons I can get round the marathon in under 3.15... would this be wildly over optimistic or do I really stand a chance of this? Be interested to hear comments from others, thanks

Sam.

put in 10k of 46:44 to predict half mara:

predicted time = 1:43:00

actual time = 1:44:50

:0(

[Deleted User]

I keyed in my half marathon time and it says I should do 20 miles in 2.43.02.
I recently ran Finchley 20 in 2.51.59 so it's way too optimistic!

My marathon time is predicted 3.48.58 but hopefully I will at least sub 4hr.

[Deleted User]

This calculator was recetnly more accurate than my own predictions. I did Fleet half recently and used the time i got to predict the time i would get on the Maidenhead Easter 10m. I had predicted i would get 1hr 10min, the calculator said 1hr 8min 47 secs. I was surprised to get 1hr 8min 3sec. The course was very flat though. It has predicted i will get 3hr 11min for FLM 2005, so we shall see, I am hoping to better 3hr 30min. If the calculator is right for this one i will be ecstatic!

[Deleted User]

Calculator was a tad optimistic for me in the FLM, based on a half done a month earlier:

Bath half: 1:23:54 PB
RW Calc : 2:54:55
Macmillian Calc: 2:56:57
Actual: 2:57:24 PB

So the Mcmillian calc was the closest being only 30 secs out and you could put that down to not always following the blue racing line.