The Geography of a “Legal” Boston Marathon

by Sean Hartnett

On Patriot’s Day of 2011, Geoffrey Mutai and Moses Mosop sped ahead of 35,000 competitors running the famous BAA Boston Marathon route that starts in Hopkinton and traverses 26 miles and 385 yards through a string of villages, college campuses, and local landmarks to the Boylston Street finish line.

Click on the graphic above for a detailed and zoomable graphic of what a record-legal Boston course might look like.

After 26 miles of fast paced running, Mutai sprinted away from his fellow Kenyan to break the tape, stopping the clock in 2:03:02, almost a minute faster than the World Record that Ethiopia’s Haile Gebrselassie set in the ’08 Berlin Marathon.

Yes indeed, Mutai’s performance was and is the fastest 26-miler ever run, but not quite a World Record, as his time goes into the record books as 2:03:02a with the “a” identifying a time run on an aided course not eligible for a WR ratification.

Why is the traditional Boston course considered ‘aided’ by IAAF rules and not a ‘legal’ course for record purposes? Looking forward, there are possibilities for revisiting the IAAF ‘aided’ rules, or for the BAA to modify its traditional course such that the elite fields would compete on a record-legal Boston route.

Certainly the Boston Marathon is certified as being the full marathon distance of 42,195 meters plus and additional 0.1% error compensation distance of 42.1meters. However, Boston’s longstanding traditional course fails both IAAF tests for an “aided” course in that its start and finish are 22.3 miles (35.91km) apart, far exceeding the allowable 13.1 miles (21.1km)—half the total race distance. Also, the elevation in Hopkinton is 459 feet (140m) higher than the finish, again far exceeding the allowable 1 meter of elevation drop per kilometer (138.4 feet).

Boston’s historic marathon predated any rules governing aided courses, and is one of the few such major courses as most of the big city marathons that emerged in the ’70s and ’80s were designed to adhere to the IAAF standards.

The root of road race rules are inherently the rules that govern track & field competitions, and marathon and other road race like track and some field competitions are considered either Loop or Point-To-Point events. Indeed these are somewhat arbitrary thresholds, half the race distance, and elevation drop of 1 meter per kilometer.

These standards are in place to guard against unfair advantages in terms of wind or gravity excessively aiding performance. Wind is the primary concern for aiding performance in sprint and horizontal jump events.

The effects of wind pushing athletes faster or farther is negated or evened off in events that involve a loop be it a full lap of the track or multiple laps as there may a headwind on the backstretch and a tailwind on the homestretch. Sprint events such as the 100 meters, 100 and 110 hurdles and the 200 meters and the horizontal jumps are considered point-to-point contested primarily in one direction, a direction that may be aided by a tailwind.

While point-to-point marathon courses are deemed aided whatever the wind, p-t-p track events are deemed to have a potential for an aiding wind and wind gauges are applied to these competitions measuring wind speed paralleling the direction athletes are running and jumping. It would be a bit of an overreach to require calm winds or a headwind, so it was deemed that a tailwind over 2mps (4.473mph) constituted an illegally-aided performance.

Elevation is less of a concern on the track as it is seen as a level playing field with minimal elevation change throughout the course. Rules do specify that for the horizontal jumps the takeoff board is the same elevation as the landing pit to guard against downhill jumps. Tracks are not perfectly flat in a lateral sense in that they have a prescribed pitch for drainage, but these are a usually a matter of inches and probably not enough to significantly aid performances.

Altitude has a greater gravitational impact on performance yet there is no IAAF standard pertaining to altitude (note: T&FN thinks there should be).

To better understand the application of the IAAF aided-course rules we can employ a Start Distance Index (SDI) that measures the distance separating the start and finish as a percentage of the total race distance. A legal SDI would be less than 0.50.

Downhill standards are quantified with an Elevation Drop Index (EDI) that divides a course’s net elevation change by maximum allowed change. A legal EDI would be greater than or equal to -1.000 – as in the allowable 1.0 meter of elevation drop per kilometer [continues below].

Click on the graphics below to view and read them at full size.


Calculating the SDI and EDI for the London Olympic Track and major marathons differentiates legal from aided events. The London 400 meter race and multiple lap races on the track have virtually the same start and finish line and SDI and EDI indexes of 0. Likewise the loop courses of the Berlin and Chicago Marathons have indices very close to 0, and very legal status.

Shifting to point-to-point events, the London Olympic 100m has and EDI of 0, but an SDI of 1.000, twice the allowable separation and necessitating the use of a windgauge. Both London and New York are legal courses with EDIs and SDIs within limits. New York’s SDI is 0.4773 just below 0.5, and London’s EDI is a legal -0.6794 and like Boston most of that drop is in the first four miles. Boston’s EDI of -3.3454 and SDI of 0.9155 are both far above the allowable and the other Marathon Majors.

Boston’s longstanding traditional course predates the rules that govern aided competitions on the track or the roads and the famed route like Mutai’s time seems doomed to an ‘aided’ label - despite being widely regarded as one of the most challenging if not difficult marathon routes. How is this the case, and is there any hope for a ‘legal’ Boston Marathon?

In a nutshell, Boston fate reflects that the IAAF ‘aided’ criteria considers a course based solely on the geography of the course start and finish positions, rather than a comprehensive appraisal of the geography of the whole 26.2 miles route. Course elevation criteria measures only the net elevation change between the start and finish without regard to characteristics of the course profile, total elevation climbs uphill, or even the physiological challenges of running on hilly terrain, and in particular the physiology of running a topographically complex marathon – which by most accounts Boston is.

With a drop of 459 feet Boston is surely a downhill marathon, and if its route followed a constant engineered grade from start to finish dropping roughly 10 feet each and every mile, it would rightly regarded as an aided route.

Yet Boston is not a grade that particularly aids runners, if anything its route seems engineered against them as most of the downhill segments feature short but steep elevation drops while uphill climbs are stretched out over the course. Almost all of Boston’s elevation drops are in the first four miles, the 15th mile, and the 23rd mile. However cruel Boston’s hills may be, the only way for it to become a legal course would be for the IAAF to change its rules to consider a more comprehensive evaluation of the topographic complexity of the course that taps modern GPS and DEM elevation data. It is debatable how this complexity may be defined, but chances are that any objective evaluation is likely to determine that Boston full route does not significantly ‘aid’ competitors.

Certainly, downhill running is an aid and I have to wonder how much faster you could run if we would spiral a track downhill at a 1 meter per kilometer pitch. How many of us even bother to read a time preceded by the words ‘downhill road mile.’ Boston certainly is a topographically challenging course, but I’m not sure that the legitimacy of marathon racing is well served by stretching an admittedly arbitrary, but already generous aid when compared to a flat track.

Likewise, assessments of the point-to-point characteristics of an aided marathon route can be modernized with multiple on-course wind measurements replicating the role of the wind gauge on the track.

As a start, digital rapid-fire weather stations can be positioned along the course affixed to the 5K chip stations and the start and finish. Timing guru Harold Mika has already started to work on integrating weather data into the split data recorded for every competitor, and wind recordings for the lead athletes could be averaged over the ten stations.

Track wind gauges could be employed along with the weather station anemometers to better define ‘aiding’ winds, or a crew could shuttle a track wind gauge along the course, setting up the gauge in advance of the lead pack at predetermined locations, record wind readings as the leaders run past, then move ahead and repeat the process such that 20-30 wind readings are recorded.

In either case, current technologies provide opportunities to replicate what is done on the track, such that point-to-points could be legal on calm days. Gauging the presence of aiding wind on the course would at least open the door to a more informed view of course-wide conditions and a determination if a marathon like an 100 meter race earns a wind-aided tag.

With a rigorous wind monitoring system a legal Boston might be possible on a calm day only if the IAAF would choose to revisit their start/finish separation and elevation drop rules. Short of that, the only recourse the BAA would have is to change their historic course. This is a bit of an obstacle in that if Boston stands for anything in the marathon world it is TRADITION, and changing the course messes with 117 years of marathon history.

Yet, a refined ‘elites’ course may be the best option, and the masses will continue the tradition, running the standard Hoptinkin to Boston route as the aided course tag most profoundly affects times clocked by elite athletes. London and New York City both already employ multiple starts which blend together the field after a few miles, and several viable starting options are possible with the Boston route that utilize the current route and meet the IAAF criteria of being within 13.1 miles of the finish and netting less than a 142 foot elevation drop.

Precise locations and elevations are critical to course certification so perhaps a little explanation of available resources is needed. Historically, two types of elevation data sets were used to determine, surveyed elevation points or ‘bench marks,’ and topographic maps. Standard USGS 1:24,000 scale topographic maps were the best common source for both types of information and were accurately used to interpolate ‘spot’ elevations, and construct topographic profiles of race routes at a scale of 2.64 inches per mile.

The surveyed bench marks and topo maps have replaced by high-grade GPS and Digital Elevation Models (DEMs). High grade Differential GPS calculate LAT-LONG-ELEVATION points with rather accurate elevations, and highly accurate relative elevations for points recorded within an hour or two.

The DEMs are a 3D rendering of terrain based on square grid cells or tiles each with a specified elevation. The Model part of the DEM represents that this topographic surface is interpolated from surveyed elevation points, from contour lines derived from surveyed elevation points. Newer generation DEMs have been generated by the Space Shuttle and the evolving Gold Standard are derived from very accurate LIDAR surveys.

While well have to wait for LIDAR data for the Boston course, I had conducted a dGPS survey of the course in 2006 utilizing a Trimble ProXR consisting of over 8,000 readings. Also utilized was the USGS DEM employed by Google Earth with grid cell elevations rounded to 1 foot. Tapping both the GPS and Google Earth data, Start and Finish positional and elevation data was calculated for several alternative route scenarios.

Starting the elites at the finish line and running the course backwards to a turn-around at the half-marathon would constitute an ideal loop course, that it until you wrap your mind and legs around the idea of running Heartbreak Hill coming and going. Legal but brutal might not be the best change.

My German colleague Helmut Winter (Professor of Marathon, de) likes the idea of starting at 25K and running backwards to a 12.5K before U-Turning to complete the course. With this route all of the 5K chip stations would remain intact, but the initial seven mile marks would be off.

You could also start at the halfway chip mat and run backwards to the ¼ marathon mark at 6.55 miles and U-Turn. The start would be located in Wellesley but would require locating the first six mile marks, and the 5K and 10K timing stations.

Pushing further up the course, starting at 12 miles may offer the best combination of maximizing the old course while maintaining legal status. The start position is just within the limit at 12.8 miles from the finish, and the start elevation is also just within the legal limit at 142 feet—within a half foot of the allowable drop.

Click here to see the course profile of Boston's time-honored traditional route versus the suggested 12-mile-mark start course that would be record legal. Once opened, you can zoom with your browser to see details.

While just inside the IAAF limits, the 12 mile elite start also has some great logistical advantages in that the starting point is within 200 meters of the Wellesey College track and athletic field – an ideal warm-up and staging area for the elite start. The run past Wellesey has always been a vibrant aspect of the Boston Marathon and adding the elite start would be a grand sendoff.

While changing the start makes Boston legal, it is still no cupcake. The start at 12 miles trends uphill, but the initial 15 miles would be classified as flat to rolling, with a steep drop in the 16th mile. The final ten miles from the Charles River bridge to the finish with the trudge up and down Heartbreak Hill still remains as one of the toughest stretches on any marathon course, and any non-aided time would be well earned.

The 12 Mile Elite Start course being just inside the legal limit, serves to underscore some of the problems with the IAAF criteria. Consider a race day with winds of 10-20 mph from the west. With the 12 Mile start the elites would run 6 miles into a headwind, then U-Turn with 29.2 miles of tailwind to the finish. Most likely the 6 miles of headwind would have added 30-60 seconds to Geoffrey Mutai’s time if he had run this route, but his ‘legal’ time would have been aided by 20 miles of tailwind. In many ways, this 'legal' Boston route may be the best reason to revisit the aided rules.

One thing is certain about Boston’s long-standing traditional course, it has generated a tradition of debate over just how much a tailwind and downhill topography aids performance. Whether or not the IAAF revisits the aided criteria for Start-Finish separation and elevation drop, or the BAA revisits their traditional course, it is possible that some day in the future we may see elite fields competing on a non-aided BAA Boston Marathon.


April 15, 2013

Copyright © 2013 Sean Hartnett/Track & Field News