Run Pacing – How to Predict a Pace for Your First Marathon.
In two of my recent articles I have discussed the importance of pacing. I started with a look at the importance of getting the pace right and subsequently went on to review some of the evidence from an elite runners perspective. I concluded that article by suggesting that if even or negative pacing was good enough for Mo Farah then it should be good enough for all of us. So the question then for the first time marathon runner is often how do I calculate my pace? The answer is often to use one of the freely available pace calculators so I wanted to look at some of those and see what differences there were in any of them. One of my runners recently completed a 10km race in 41:27 so I used her time to try and ascertain her predicted marathon time finish. With this known time we could therefore better gauge how to pace the marathon
You can find a number of online predictors and it is a good place to start with those. Just put marathon pace predictor in your search engine of choice and you will have a number that you can browse through. Using my runners 41:27 10k time you will find that most give you times the 3:12 to 3:17 mark. The final two metrics I looked at are two that I am very familiar with. The first was oft quoted by the late great Frank Horwill who stated that if you take you 10km time, multiply it by 5 and then subtract 10 minutes you will get your predicted marathon time. Looking at our time of 41:27 this would give a marathon prediction of 3:17.
The final metric is known as Daniels Running Formula as worked out by Professor Jack Daniels, Phd. Using his formula our 41:27 gives us a prediction of 3:10 albeit the exact read across is for a time of 41:21.
If you look at Elite runners that we in the UK will know well and start with Paula Radcliffe then you will see that Paula has a track PB of 30:01 set in August 2002 and a road PB of 30:21 set in Feb 2003; so one of these times is set on a track in the Summer and one on the road in the Winter. Were they set at altitude, was there a head wind, were they uphill/downhill we don’t know – I would hope the track was flat though! But you can see all of these factors will affect finish time and if we are then using that finish time to predict a performance over a marathon then we are introducing errors potentially. Using her 30.21 10km road time she went on to run the worlds best marathon time of 2:15:25 two months later in April 2003. Daniels Running Formula predicts Paula would have actually run a 2:21:26. Daniels does admit himself in his book that the times he predicts are ones that with adequate training an athlete “could” achieve so you would expect and actually you normally see non elite athletes not achieving their potential over the marathon distance. Most athletes I suspect are not adequately prepared to run to that potential. The marathon is a long way!
Professor Tim Noakes’ Central Governor Theory may give an answer to that but we’ll leave that for another post. My explanation for Paula’s result compared to her Daniels prediction is that like the under prepared athlete the predictors accuracy is less at the elite athletes level and probably an accuracy versus result graph would form an S shaped curve with good correlation for the majority of athletes in the flatter portion of the “S”. I checked this theory against Mo Farah’s 28:08 10km win in the Europeans, yes I appreciate that is a track race but I wanted a recent result; Daniels would predict would give a 2:10 marathon time and Mo ran a 2:08 in the London Marathon.
I am sure the algorithms behind these predictors were never designed to give pinpoint accuracy a bit like dating websites (This is not from experience!) but as you can see they do put you very much in the right area in terms of time. Our athlete with a 41:27 is never going to run a 2:30 marathon (unless she gets significantly faster) but equally she is never going to run a 4:30 ….or is she? If she gets her pacing wrong and goes out to quickly then I’d suggest anything is possible in a marathon!
So what was the result of this look at predictors? Well I think the science behind them, particularly Daniels, is pretty well founded and I have always considered Frank Horwill’s predictor to be a true accurate picture for the majority of runners. So for our 41:27 runner you could predict a finish time between say 3:12 to 3:20. You could then set the runner off at a pace to hit the bottom of the band so in this case 7:38 per mile to give a 3:20 finish. Remember as always even pacing means increased effort. If the runner has something left in the tank at 21 miles then she can pick up the pace but in her first race why not set a time then go out and knock some time off it in your second. Train Smart!