Wednesday, 2 April 2014

A Helpful Marathon Pacing Calculator


Yes, back again, so soon, but tonight's post is quite timely, as in the UK this weekend there are the Brighton, Manchester, and Blackpool road marathons.  Then next weekend there is the London marathon and also the Jersey marathon.

My plan tonight is to simply introduce  a web page which has a marathon pacing calculator which may help many runners who are racing a road marathon in the next two weeks.  

When it comes to running a road marathon, there are probably two things that cause the most discussion amongst marathon runners:  nutrition, and pacing.  Out of the 125 posts I have published on my UltraStu blog, by far the posts that have generated the most comments and discussion have been to do with race pacing.  From  my third ever post that introduced some of my ideas on pacing for ultra trail races it created a little bit of a stir.  But my road marathon pacing analysis based on the first 25,000 finishers at last year's London Marathon really created quite an animated response, with some readers feeling sorry for me in that I was so confused "Sorry Stu, this article shouts out that you don't understand how to go about analysing pacing strategies."  and other readers just simply coming to the point with "OMG, what utter BS", which I presume means "Oh My God! What Utter Bull Shit!"  Although maybe I am confused and it should have been interpreted as "Oh My God! What Utter Brilliance. Superb!

Well over the last four years of writing on UltraStu, I have learnt that even if many people disagree with my ideas, it doesn't actually mean that my ideas are wrong, as there are usually other people who agree with them, and most important, at the time of writing I obviously thought my ideas were good.  Now nearly a year later, re-reading my two blog posts on road marathon pacing strategy, (post I, and post II), I am still TOTALLY  convinced that what I wrote back then makes total sense.

It was therefore quite pleasing that when chatting to Tim, a follower of my blog, when he commented that he really liked my London marathon pacing analysis and that he also thought that it could be hugely beneficial to thousands of runners to help them achieve their marathon target time.  He then offered to turn my ideas from my blog post into a marathon pacing formula which would calculate mile split times for every one of the twenty six miles, to ensure that runners remained on target to achieve their target time, with the formula based on real life data from tens of thousands of runners.

Yes, I know that it is only a few days to this weekend's marathon, but I would expect even at this late stage, there will be thousands of runners just not sure what pace to start out at, in order to achieve their goal finish time.  I guess most marathon runners would have heard of the negative split, i.e. running the second half of the race slightly quicker than the first half of the race.  And many may like that idea, as it is frequently promoted as the best strategy within running magazines or on running podcasts.  Or perhaps if the negative split strategy seems a bit too ambitious, then many runners may favour the even split approach.  Where the runner maintains the same running pace throughout every one of the twenty six miles.

Now I won't repeat my views on this totally flawed idea, as you can read these within my two blog posts from May last year.  But it does still absolutely amaze me how seemingly clever people think that one should still be able to run at exactly the same running pace after 20 - 26 miles, as the pace they would run at when totally fresh, as if fatigue doesn't exist!  It seems so obvious, that the only way this even pacing can be achieved, is if the pace run at the start of the race is just far too easy.  It has to be so much slower than one is really capable of, to be able to allow for the fatiguing affects of running 20 - 26 miles to equally balance the amount of time the runner at the start of the race is deliberately going slower per mile.  To put it simple.  If runners didn't fatigue during a marathon race, then an even paced strategy would be the result.  But fatigue does occur during a marathon.  Ask pretty well every marathon runner.  Therefore to achieve an even paced strategy, one has to run the miles at the start of the race at the pace that they feel that they are able to achieve when they are near exhaustion at say around the 25 mile mark.  One has to deliberately slow themselves down at the start when not fatigued, just to ensure that they aren't running too quickly at the start, otherwise there would be no chance at all at running at anywhere near the same pace when near exhaustion 25 miles later!

Sorry if the above paragraph is a bit jumbled.  I just found it quite hard trying to explain something which seems just so obvious that it shouldn't really need explaining!  Anyway, I did say that this post was going to be short, so lets finally go to the marathon formula calculator, available on the following website:

The web page calculator is titled the ReSUltS Marathon Formula.  Yes, implementing the pacing strategy that the formula calculates, runners are MUCH, MUCH more likely to achieve the result they want, i.e. their target marathon time.  Adopting the even split, or even worse the negative split pacing strategy, then as the data from last year's London Marathon demonstrates, runners are therefore 96% likely to FAIL!  Yes, I will repeat it in words, just to make it that much clearer.  Adopting the even split pacing strategy, has a NINETY SIX PERCENT chance of FAILING.

To those of you running a road marathon in the next two weeks,  I strongly ask you to consider is it wise to follow a strategy with only a 4%, yes ONLY a FOUR PERCENT chance of succeeding.  Weigh up the odds.  Look at the data. Fatigue does occur!  Slowing down during a marathon IS A REALITY, due to the gradually developing fatigue as the race progresses, especially during the final third of the race.  Talk to any marathon runner, that is when the marathon starts really getting challenging.  But somehow, one isn't meant to slow down, even when massively fatigued!  Is that reality?  No not for 96% of the runners!

The  ReSUltS Marathon Formula. therefore gives the marathon runner clear guidance on how much one should expect to slow down as the runner fatigues during a road marathon.  The amount of slowdown isn't a guess, nor a dream (i.e. not slowing down), but based on the data of over 10,000 runners, with the amount of slowdown being dependent upon the runner's target marathon finish time.  Just use the arrows at the top of the web page to enter your target marathon time, and with one click, your pacing strategy is calculated.  As simple as that!  All the guess work removed.

No doubt there will be many people who will still think that I write "Utter Bull Shit".  But everyone is entitled to their views.  However, I find it so disappointing, that over the next two weeks, so many marathon runners  will not achieve their target marathon time.  Not because they were not capable of achieving their goal time, but for many, simply because they followed an unrealistic pacing strategy and simply ran the first half of the marathon too slow, i.e. running at a pace slower than they could have actually ran at.  Hopefully the ReSUltS Marathon Formula calculator will help increase the number of marathon runners achieving their goal time.

If you know some runners that are running a marathon in the next two weeks, then forward them the link to the ReSUltS Marathon Formula web page, especially if they are thinking that they can achieve an even split marathon, whilst at the same time, running a marathon time that truly represents what they are capable of achieving.

I will sign off with a quote that perhaps helps illustrate the logic some people apply in justifying that the even paced split is obviously the correct pacing strategy.

"No offense to hundreds of thousands of runners, but quite frankly I don't care that 95% of them run positive splits. The real question should be could they have run a faster time if they would have run more even splits? And again, who cares if it was even a PR - could that PR have been a better PR?"
Brett, 2011 ( A comment left on a previous post)

All the best to all the runners running a marathon in the next two weeks, especially those at Brighton.  I will be cheering you on loudly on Sunday morning, hopefully in the usual Brighton marathon sunshine.



  1. Yay, I got quoted!

    The point I was trying to make was to look at all the current records in distance running - if they all are even to slightly negative splits, that has to mean something.

    For example, the current marathon world record is around 2:03 and was an even split to nearly the second.

  2. Hi Brett

    I hope you didn't mind being quoted. In relation to your new comment this morning, to paraphrase, no offense to you, but using the argument that the World marathon records are achieved by running an even split or slightly negative split is exactly the misjudged logic I am referring to. When in any other situation does the club level athlete, the 'average' person (although I dislike the word average but I think it makes the point clear) try to mimic what the World's best can do. Whatever activity; e.g. scoring a maximum 147 break in snooker, cycle racing at top-end pace for hours every day for near 22 consecutive days around and over the alps of France, or managing to descend to an ocean depth of 214 metres on one single breath. Yes, regardless of the activity, 'average' people do not expect to be able to replicate these amazing feats. So please explain to me, why is it that when it comes to running road marathons, that it is assumed that the 'average' person, who is not a full-time athlete, who does not have the same opportunities to prepare, the same resources, the same environment, and dare I say, the same genes, that this 'average' person can then achieve the same as the World's best, I just don't understand! Could someone please explain this logic.



    1. Love the comments. Agree with Brett. Also most people running London or most marathons poorly execute, and run to fast to begin with due to various reasons, then tire later on hence positive splits. This doesn't mean for a second it proves they are delivering the best performance on the day and would in my opinion run better, if they ran a strategy thT meant they were stronger later. For me it is mentally very positive to be overtaking in the last 6 miles and feel good than struggling and seeing a time slip away. Fascinating conversation though!

  3. Hi Stuart, no I don't mind one bit. And I welcome such a debate as this. I am frequently wrong, just ask my wife.

    I'm not saying average people should target the same times, I'm saying if you want to run the fastest time you can on any given day, you should look to what good people are doing and to any applicable scientific studies. This is not to say you should ignore the masses either though.

    In the previous comments that you referenced, I referenced various scientific studies that show even splits was the route towards achieving the best time. This pace calculator (in my opinion) gives you a higher chance of running a certain time...but that time would not be the fastest you could run on that day. For example, I am roughly a 3 hour marathoner, which is ironically the initial time populated in that pace calculator. It tells me to run a 5 minute positive split. I cannot fathom starting out running so fast such that I positive split 5 minutes in the last several miles. That is a borderline blowup. And I've been there and felt that way :)

    So that's the point I'm trying to make - if you want to guarantee you a better chance of hitting a certain time in a marathon, then yes you can allow for a several minute positive split. But if you want to guarantee the fastest possible time on that day, you would not set up your pacing the same way.

    In the USA, we also had the 100 mile track record broken a couple times in the last few months. The first time it was broken (Jon Olsen), he ran two 50 mile splits within 2 minutes of each other. The latest time the record was broken a few weeks ago (Zach Bitter), he ran a slight negative split of a few minutes in the second 50 mile section. So as best I can tell, this same behavior is seen across ultramarathons and down to half marathons and 10ks as well.

  4. Paul the builder3 April 2014 at 16:29

    Brett's quoted comment is the best part of this blog entry.

    Stu - correlation is not causation. How many runners are under-trained for the marathon, yet you happily include their stats? How many runners make *obvious* pacing or executional mistakes, yet you include their stats too?

    I asked on a (admittedly quite recent) comment on one of your original posts - "to simply say that most people slow down, therefore positive splitting is the most efficient strategy is illogical. You don't justify it, why should we accept it?". Actually - perhaps "illogical" is not the best word there. what I really mean is - it doesn't *automatically* follow. Most people drive their cars on a motorway at 70-80mph - can I *automatically* draw a conclusion about safety, fuel efficiency, time-to-reach destination from that? Of course not.

    I've thought of a simpler example. *Most* people who run the London Marathon where you draw your data and conclusions from, do it off low training mileage. Why should you, Brett, I, or anyone else who aspires to run a better marathon than average (or other ambition - even just better than one's own previous performance) be limited to low training mileage? We aren't. We're ambitious. We try to do better. So why would you suggest we limit ourselves to the pacing strategy of the 'majority'?

  5. Thank you Stu!
    This makes perfect sense to me. A strategy which I use in my maras.

  6. Stu - I believe it would be useful to modify your online calculator to allow runners to specify the split % then want to follow rather than have it prescribed based on averages for each pace band. Basing this % on the slow downs measured from all comers in marathon is unlikely to lead to optimal pacing, on vast majority of of runners don't pace well in marathons for a range of reasons, codifying this bad practice in splits don't seem helpful.

    On a more general note I'm doubtful that pacing by splits is good thing to do, regardless of whether you intend to positive, even or negative split. The problem with using pre-computed splits is that can't account for conditions on the day like temperature and wind, course terrain and how the runner feels on the day. If it's a hot day, or windy then what were reasonable splits become impossible, also if their are hills on route trying to keep to a particular split for each mile will lead to pushing on too hard on ascents and too slow on descents. Finally keeping anything other than even splits in your head or person is simply impractical.

    What is a far easier way for people to pace by is to simply pace using a heart rate monitor. The simplest model is to target a roughly constant HR through the race, this will lead to a steady slowing through the race due to HR drift, which is obviously means a positive split that is well established in sport science as being sub-optimal, but as long as HR drift isn't crazy then pacing balance shouldn't be too bad. The only trick with aiming for a constant HR through a race is knowing what HR to target, here experience in training and average HR from other races is invaluable. It's crucial to make sure this target HR is tailored to the person and for the distance though. Too low a target HR will lead to overly slow race, too high a HR and the runner won't be able to maintain pace and will likely crash and burn in the second half.

    A refinement of using the HR monitor as a guide is to account for HR drift and target different HR values at different stages, the optimal split for performance is a even or slight negative split, so this will require the HR progressing slowly higher through the race. The MARCO online HR/pace calculator attempts to provide a guide for HR profile required for what the authors of it deemed to be ideal pacing:

    Personally I found the suggested HR drift too pessimistic for the Loch Katrine marathon than I did 10 days ago. So again I believe one should only use this as guide and tailor it to ones on HR drift. Yesterday I had a bash at writing my own thoughts on my blog:

  7. Stu, How many of your races in the last year would you have judge to have been a success?

    I ask this as answering it might just set you on a journey where you have an epiphany about where you could improve your own running. To me your race success has been really hit and miss over the last few years. Often you cite your lack of TOTAL preparation as the cause of lack lustre performances. However, if your approach to TOTAL preparation is so sound why are your race performance so erratic?

    Personally I think a big part of the problem is your pacing strategy. You mask this by putting enormous amounts of mental commitment into races, but when this falters so does your performance.

    More sensible pacing strategies are much easier to follow through on the day and achieve the best that you can achieve on the day. Good pacing enables you to be strong through the whole race, if you get it right it becomes pretty easy to run strongly right through to the end of the race.

    How do I know.... well in all of my races in the last 9 months- four races, three ultras and one marathon I nailed them all, in the only two race that I had done before I did big PB's. I didn't need massive mental commitment on the day, the great races just happened naturally. A big part of this success was pacing using my HR monitor.

    Oh.. for my marathon PB I did a playful 3 minute negative split :-)

  8. Paul's comments are on the nail. At least the database should be limited to people who have done (say) a faster marathon than either of their previous two.

    I think there is only one way to resolve this dispute though, through a carefully controlled twin study.

    First, recruit 10 pairs of identical twins (of various ages) from the running community who (a) have been running for about the same number of years, (b) have similar current 5k times, (c) live in the same town, and (d) are about the same weight. Ask them to do an identical marathon training programme (that's why they have to live in the same town) and then run the same marathon on the same day in the same footwear. Randomly assign one of each pair of twins to an even-split pacing strategy and the other to a positive-split strategy devised by Stu after taking the best scientific advice he can find on the subject.

    To control for placebo effect, at the start of the process show each twin some literature supporting the positive-split view and some supporting the even-split view, and ask them which strategy they think is more likely to work for them. Ask them again a week before the marathon in case their views have changed. Then when it comes to assigning pacing strategies ensure that about half of the runners are doing the strategy they believe will work for them and half are not.

    My predictions:

    - Overall there will be no significant correlation between beating your twin and using either one pacing strategy or the other.

    - There will be a significant correlation between beating your twin and using the positive-split strategy for younger runners.

    - There will be a significant correlation between beating your twin and using the even-split strategy for older runners

    - There will be a significant correlation between between beating your twin and using the positive-split strategy for older runners.

    - There will be a significant correlation between beating your twin and using the pacing strategy that told the investigator you believed would work best for you (placebo effect).

    I am sure some funding body would be interested ..

  9. <a href=" Cabs to London

  10. Very helpful if you run in miles :) Any chance of having a conversion to km? These numbers mean absolutely nothing to me :(