Marathon Conversion Tables…….or To Hell With Algorithms

By David Chalfen

It’s that time of year, everyone has their marathon goals coming into view. The vast majority, at a wide range of performance levels, will have a half marathon in the build up and, amongst other data will be using this to indicate their likely marathon time. I have seen so many tables and articles about this subject and whilst some are very sensible, none seem to pick up all the relevant threads. So, in no particular order and with absolutely no claim to be authoritative or exhaustive, here are some thoughts. As context, in recent years I have coached people to close to 100 marathons per year, about 70% men and 30% women, and about 95% of these runners are finishing between 2.20 and 3.50. I’m fully aware that the median finish time in the London Marathon is now around 4.45 and about 1% of my coachees are at this level. So here goes:

Many conversion tables seem to expect fantastic marathons, and are predicated on a 3.45 or 4.10 marathoner having the same relationship between their VO2 max and their running economy (both are empirically measurable) as the elites. I can’t imagine why their calculations are done on this basis given what we know about training frequency; volume; and longevity and how this interplays between VO2 max, anaerobic threshold and running economy. That is, physiologically, probably the biggest and by far most frequent factor in why most charts just don’t work for so many below the ‘good club’ level. I always wish I could chat to one of the coaches whose tables show that with an 86 minute half marathon you should nip sub 3 or with a 1.54 half you can get sub 4, and ask them to show me that this is the norm amongst their coachees. In those very rare cases where these wonderfully tight conversions apply then either the runner is on a notable improvement curve whereby their half PB is antiquated on marathon day, or the course/weather on their best half marathon was not as ideal as the target marathon. It’s not a conversation I am likely to have but I suspect that Paula Radcliffe was in that category and I have chatted at length to an Olympian and 2.12 guy who was such a runner.
“Women are better than men at converting to marathon.” At elite level this isn’t the case – look at the world’s best women in the sub 2.23 category, look at their 10k and Half PBs and compare then with fine level men with the same marathon PBs and the conversion is very similar. The reason? I suspect that at this level most of the men are very well trained and experienced – they train and race along the lines of the very best but don’t have their innate ability, and so pan out at the level of the worlds best women. At around 3.00 to 3.15 level we see a difference – I suspect , and I generalise here, that the typical 3.00 women is closer to her limit than a 3.00 male; her marathon is ‘worth’ around 2.43 for a guy. So she’s likely to have relatively better running economy (through training volume and longevity) and less likely to do silly things on pace judgement. In the UK we have now around 140 women sub 3 per year (annually increasing) and some 2500 men – so its logical that a woman at approaching national level is likely to be ticking more boxes than a ‘mere’ solid club level senior.
“Older runners are better than younger runners”. I think that the stats which take in the full range of marathon finishers are correct in their conclusion, that is, that generally the older runners are more clued up, probably with more volume in the marathon build up and historically and correspondingly less likely to do rash things early doors. But I wouldn’t expect, say, a 48 year old in 2.55 shape to have any intrinsic advantages over a 23 year old in 2.55 shape simply because of being older, if physiologically both seem to be in the same state of fitness for 26.2 miles.
Tables that have a significant number of people beating their likely target time. Dooh. Surely if you do a marathon you have an uppermost realistic target where if everything goes right on the day, that’s what you get. If you are minutes ahead of this then whoever has been advising you (in most case, ‘self’) doesn’t really know their onions. The whole goal setting thing is not for this article but if you are anywhere between a 3.00 to 3.59 type and you are thinking in 10-15 minute time zones then you are missing a trick. 15 minutes in a marathon is around 35 sec per mile. If you do a 10k race would you be looking at that sort of woolly time frame for pacing?
None of the tables quantify or, typically, acknowledge the mental factor. It’s not yet at a stage where it is measurable but it seems likely both that the fastest runners are mentally better able to handle the rigours of the marathon than those many miles behind them; and, that some runners of equal physiological ability may have mental strengths that their apparent equals don’t have. Not yet, at least. These are trainable of course but the benefits of such training currently aren’t quantifiable.
Similarly, whilst we all know that some people are genetically better suited to distance running by virtue of innately high VO2 max and/or response rate to training at very high percentages of VO2 max, maybe when it comes to that quirk of marathon racing– one’s efficiency at utilising fatty acids as fuel – some people respond slightly better than others to training this trait. It would be illogical if the same training stimulus produced exactly the same level of response and adaptation in everyone. That said, I don’t think the margins of variation are huge so if you are an 81 minute half marathoner who keeps not dipping under 3 hours then I don’t think you can just sigh at genetic bad luck!

So where do I get off on this? I actually find amongst the range of levels I coach at marathon that an old formula of 5 x 10k time minus 10 minutes is pretty helpful – at the sharper end you may well ‘beat’ this by 2 to 3 minutes, but not more if your 10k time is current, all things being equal.

If you want some fairly conventional numbers:

70 - 2.27-28
80 - 2.50
84 - 2.59
90 - 3.15
96-97 - 3.30
1.50/51 – 3.59