Thursday 17 April 2014

Toughest Race in Scotland?

Mike Raffan's Facebook post was a little challenge not to be ducked.

"I was out for a long run today and was thinking to myself my legs are still tired after Tuesday's race. It was only 3,5 miles but hilly thus tough. That got me thinking: what is the toughest race in Scotland? I know what my answer is but I'll wait to see what some others think first, please justify your answer, don't just say "Tough Mudder 'cos it says so in the name and I found it hard". I don't mean the toughest one to win as that is any race with a strong field. What is the toughest race to finish? Hopefully after a few days and everyone has had time to answer, Andy Cole will write a blog post giving his analysis. Excluding the Spine as it is only in Scotland for a few miles."

I don't know Mike well but we've chatted a few times (the last time when he was acting as a kit-check marshal in last summer's Lakeland 100).  He's definitely no slouch as a runner, having a 18hr 18min West Highland Way under his belt among other good performances. Well sorry Mike, it took a bit longer than a few days, other things got in the way, but here goes.

Mike was referring to a post I did a while back called "The Toughest Race in the World?", in which I tried to compare races on the basis of how hard they are to finish. It still gets a regular stream of hits so it must be good for a bit of pub talk if nothing else. My contention was that if you know you can finish an event then how tough you choose to make it by going faster or slower is just a subjective game you play, but if there are some events that you can finish and some that you can't, then there must be a way of ranking just how hard a task finishing is for you. I thought of revisiting that with particular reference to the Scottish races, but time passes, the idle thoughts go on, and  I ought to be able to do better than that by now.

So I've tried to build a little model which could be applied to any race. Now I'm not claiming any stunning scientific underpinning here, just a bit of knowledge that I've gained from doing quite a wide variety of races and which I've tried to apply in a logical way. For sure it could be fine tuned, and there may be better places to start even, it's just a first hat in the ring if you like.

In the other post I went into some detail on what factors can make a race hard, but if you get bogged down in every possibility then you never really get started so I've just considered the most significant. A bit like, say, a rock climbing guide  -  grades are for good conditions and if you choose to do your climb in the rain then it's going to feel a bit harder.

The Model

It seems to me that there are three main physical quantities that set the overall work to be done when completing an event, and these are

1. The length  -  goes without saying, 100 miles is tougher than 50, all other things being equal.
2. The ascent -  or rather the ascent per mile, because for example  4000 ft of up in a 20 mile race is going to have rather more impact than 5000ft over 100 miles. I have called this the Climb Factor.
3. The ground underfoot  -  well compacted trail or springy grass is much more helpful to progress than trackless rocky hillside or knee-deep bog. I have called this the Ground Factor

So, if you take the Length of the course, and multiply it by the Climb Factor and the Ground Factor, you have a measure of the overall undertaking that the event presents. I've called this the Overall Effort (for me this is just a number, no units, it's just for comparison with other events).  Now the Overall Effort alone does not necessarily make a race tough. The UTMB course has a a pretty high Overall Effort rating by my scoring, but it's completed by lots of people every summer, many of whom have done little or no training for the trip - they just take 10 or 12 days to complete it. What makes the UTMB race tough is that instead of 10 days, you have only 46 hours to get round. The time factor is the key  - and this was the main conclusion from my earlier post.

So although the Overall Effort is an interesting enough value in itself, to rate how hard an event is to finish, you have to bring in the time pressure. I have done this by using two further factors, which are

4. The cutoff time  - again goes without saying. Now you could argue there is a difference between races which just have one final "must complete by" time and those which have intermediate cutoffs which normally force competitors to go faster earlier in the race. The theory behind the latter is that if you don't make the early cutoffs you have statistically no chance of overall completion. I actually don't agree with this at all, it just reflects what most runners do, rather than what they should do, rather like the positive/negative splits argument for marathons. We could debate this for a week, but to no point as it's too complicated for my model anyway. I'm only considering the final cutoff time.
5. Fatigue  -  it doesn't matter what the climb, ground, and so on are like, covering it when you've had no sleep for the past 30 or 40 hours is not the same as when you just started out. The longer a race goes on, the more impact fatigue has on your performance. I've called this the Fatigue Factor.

So if I take my Overall Effort, divide it by the the cutoff time (if you're finding a race near your limit, you'll be quite close to this time when you finish), then multiply by the Fatigue Factor, you (finally!) have a measure of how "Tough" the race is. Again this is just a number so one race can be compared with another.

Now, I could go on here about how I've devised each of these factors, but if your Myers-Briggs personality indicator is anything like mine you'll have had enough of this chat by now, and want to get on to the results! So I will explain the basis, but I'll do it at the end of this post for anyone that wants to carry on that far. So here we go.....

Results for some Scottish (and non-Scottish) races































Now another health warning here is that it's quite difficult to get accurate and consistent basic input for some events. If you look at the UTMB "qualifying events" list you see for example that the height gain given for the West Highland Way is 4500m, whereas the the sum of height gains for the Highland Fling and the Devil o'the Highlands is 5500m! I've taken info from original websites as the best available, and used a bit of judgement where it didn't seem consistent.

So, looking at the Scottish races, which was Mike's original question:

The Fling and the Cateran come out similar, with the Devil slightly tougher because of its higher climb factor and tighter cutoff relative to its length, but it's marginal and you could argue that these three are very similar. Finish one and you should be able to finish the others. The Double Cateran is significantly tougher than the WHW and this makes sense. They have similar climbing, but imagine you are just about to complete the WHW in 35 hours; to complete the double Cateran you would still have 15 miles to go, and you should have finished seven hours ago! This probably makes it Scotland's toughest ultra race  (not enough details to judge the new Cape Wrath event, but multi stage events are anyway generally easier to finish because of the recovery time allowed), but the biggest recognised "outing" in Scotland is the Ramsay Round, and I think this makes sense.

Comparing the Scottish races to those south of the border, the Hardmoors 55 and Lakeland 50 are not far off the Fling and Cateran as undertakings. The Lakeland 100 is up there with the Double Cateran, in spite of the 40 hour cutoff, because of its tough ground and fierce climbs. The Hardmoors 110 appears to be in the same league but I'm a bit sceptical that the total ascent figure is right on this one, nearly 20,000 ft seems a lot, but I'm sure Jon Steel could confirm whether this is actually true or not. Again, the mountain rounds are the big challenges. Another confirmation that this is about right is the completion rate.  Even the ultras generally recognised as gnarly normally manage about a 50% completion rate, whereas I think something like 75% of Bob Graham attempts end in failure, and maybe even more for the harder Rounds.

Abroad, it may be surprising that the Tor des Geants comes out easier than any of the Chamonix races but I think this is fair. It is a massive Overall Effort, but 150 hours is long enough for anyone that wants to complete it to take their time and do just that. It doesn't have the "must keep moving" urgency of the Mont Blanc events. 

I just put the Barkley Marathons in to confirm that however you try to rate it, it will always come out a class above anything else in the world. The only completers of this event so far have also had top ten finishes in UTMB, Hardrock, or have a similar pedigree.

A race that doesn't fit the model is the Spine, where conditions are everything. Completing the Pennine way in 7 days in summer would present the average mid-pack ultra runner with no real problems, a bit like the Tor des Geants. So how hard it is to complete with ice, snow, sub zero temperatures and constant wind I just don't know  -  maybe I'll have to try it one year(!).

I deliberately didn't read through all the replies to Mike's original post. I'll do that later, to see if anyone agrees with me!

So that's it. Just a bit of fun really. I'm there to be shot at. If you want any other races put into the same model, let me know.


Constructing the model (for those that are interested)

There is always a danger that when you apply some science to a very imprecise subject, you end up kidding yourself that you have a much better handle on what is happening than you actually have. So I have tried not to be too clever but just to apply some basic "finger in the air" rules on how to deal with the various inputs. It works as follows:

Climb Factor
This is 1 + (total climb in feet)/(length of the race in miles)/ 500.
Now (total climb in feet)/(length of race in miles) gives the climbing per mile. In fact it is only half the climbing per mile, because in most races the ascent is balanced by an equivalent descent so you only climb over half the race distance. This is where my figure of 500 comes from. I started from the assumption that if a runner can cover a mile on the flat in X minutes, he is likely to take 2X minutes if that mile includes 500ft of ascent and 500ft of descent. Play around with some figures and you'll see what happens. As a further refinement, I multiplied the resulting factor by 1.1 (ie adding 10%) if the nature of the race means that it has very steep and/or very long "in your face" climbs  -  1000ft gained on a well graded and easy to follow path (eg Devil's Staircase) is not the same proposition as 2 or 3000ft in one go over boulders, tree roots, scree, etc (Bovine Alp, Black Sail Pass, etc)

Ground
I gave each race a rating for difficulty of ground, on a 0-10 scale, where 0 would be a smooth (but maybe not metalled) road or path, and 10 would be Barkley type terrain (no path on 40degree slopes with knee-deep scrub, briars and fallen trees).  So forest tracks and occasional ploughed fields (Round Rotherham) would be 1, good trails (West Highland Way) 2, trails but with boulders and bog (Lakeland 100) 3,  grassy/stoney fells (BGR) 4,  and so on. My top rating for the UK was 6 for the difficult rocky and heathery ground encountered on the Ramsay. I then converted this into the Ground Factor by using 
Ground Factor = 1+ (ground rating)/20.

Fatigue Factor
It's difficult to put this into a linear equation, because fatigue hits you at particular points. You have a day out; you have a long day out; you miss a night's sleep; you miss two night's sleep; and so on. I think each of these events causes a step change in your ability to deal with what's going on around you. But then eventually, on the very long races, you get back into balance where you're just about getting enough rest to go on (within reason) indefinitely. So I brought the fatigue factor in via bands  - up to 17 hours Fatigue Factor = 1;     18-23 hours =1.25;    24-30 hours =1.5;   31-40 hours = 1,75;    41-60 hours = 2;   above 60 hours 2.25.

Time I got back to doing a bit of running again.