Mount Washington Lottery and Miscellany

Fist things first, the Mount Washington Road Race lottery opens today and will go until March 15th. If you have any interest in running one of the World's truly classic mountain races, make sure that you enter before the 15th of March. It is rumored that about twice as many runners enter the lottery as can be allowed (for logistical reasons) to enter the race. This means that if you enter the lottery, you have a 50-50 chance of getting in. There is a lotter bypass system as well.

Second things second. I was recently interviewed for the inov8 team blog and one of the questions that I was asked was

"Living (at low) altitude(,) do you feel altitude is overrated and/or that focus on speed can make up for the lack of altitude acclimation? It seems as though it did not affect Blake once again at Mt. Washington last year."

My answer was as follows.

My feeling is that training at high altitude allows an athlete to better develop their cardiovascular system and training at low altitude allows an athlete to better develop their leg strength and speed. Certainly, someone training for a race that requires speed (shorter than, say 3km) would be best served by training at lower altitude while someone who is training for a marathon or ultra would be best served by training at high altitude. In training for the mountains, I honestly do not know which is better. Personally, I love the feeling of overall fitness that I get from training at high altitude (I have spent periods of training at high altitude in the past); I also love the feeling of slamming uphills at low altitude.

As for the Mount Washington race specifically, I do not think that athletes coming from low or high altitude training situations necessarily have the advantage. An analysis of past result suggests this as well. Of the 96 performances of 65 minute or faster on the hill, 29 came from athletes training at high altitude (4999 feet above msl) and 67 came from athletes training at low altitude (42’ to 1400’ above msl). Taking into consideration that the race is in a region where most of the population lives at low altitude (thereby adding a very difficult to quantify independent variable), this seems like a pretty even split. What is really odd is that no one coming from between 1400’ and 4999’ has ever run faster than 65 minutes. Weird.

It should go without saying that athletes that are trained at high altitude have a notable advantage in races that are held at high altitude. Anyone who says otherwise is stupid or lying.

I had always had a gut feeling that neither"athletes coming from low or high altitude training situations (do not) necessarily have (an) advantage" but I thought that I would look into it because, well, I like to be able to make informed statements rather than just going with my initial "gut" reaction. So, to make an informed decision, I make a bivariate plot of most of the times run under 65 minutes vs. the training altitude of the runner responsible for the time. The "training altitude" serves essentially to more Derrek Froude from his low altitude home of NZ to the high altitude of Boulder, CO and moves Daniel Kihara from his high altitude home of Kenya to the low altitude of Philli Metro.

A few other notes on the graphs: i. For the sake of simplicity, the altitude for each athlete is based on the altitude of their hometown as listed in the race results thus, anyone from Boulder, CO would be listed at 5202' even though, obviously, not everyone in Boulder lives at exactly 5202'. ii. Matt Cull, Mark Berman, David Ezersky, Peter Pfizinger are not plotted because their geographic affiliations were not listed in the race results; all are low altitude at Cull (I believe) was living in VT at the time and the others represented either BAA or GBTC leading me to believe that they were coming to the race from low altitude training centers as well, iii. I chose 65 minutes as a cut off because I did not want to spend too much time on this and because I thought that it would be a good round number that would separate the national-level performances from the regional-level ones- since the race is, after all, held in New England, including too many runners in the analysis would actually make the results less meaningful (Eric Morses 65:00 from 2001 is included in the plot), and iv. the black plot represents the performance of Eddy Hellebuyck who tested positive for EPO in 2004. The point representing the 64:49 that he ran in 1995 is black because it would be unfair of my not to point out the the performance of a known drug cheat.

Anyway, one of the things that becomes immediately striking about the plot is the dearth of sub 65 performance from runners training between 1400’ and 4999’.


I placed two lines on the second graph- one a hypsographic curve showing the cumulative percentage for the world's landmass and a curve representing global hypsographic demography (the elevation at which humans inhabit the Earth). I did not have specific curves for the USA but I would suspect that they would look very similar to the world curves. The curves, of course demonstrate that there, of course, more land surface near sea-level but also that there is more land (and people) in the 1400’ and 4999’ range than there is in the 5000'+ range. So where are all of the runners?


One more figure shows the locations of sub-65 performances geocoded onto a map of the continental US. This shows what is essentially a bimodal distribution with the primary cluster in New England and a notable secondary cluster along the Front Range cities of the Rocky Mountains of CO and NM.