Uphill
Bike Races: Whiteface Mountain vs. Mount Washington
by Peter O'Reilly
Have you raced your bike up Whiteface Mountain and are wondering how you would fare doing the same for Mount Washington? Or maybe you have climbed Mount Washington with your bicycle and would like to know what finishing time to expect for Whiteface? If so, then you have come to the right place. Please read on.
This web page serves a simple purpose of predicting your expected finish time racing up one of these mountains using your past experience for the other. All that is needed is your finish time for the climb that you had completed.
At this moment you may be wondering how this is so. If you are hard-core cyclist, you may suggest one calculate a very accurate time using such information as: their average power threshold, average speed, distance traveled, coefficient of drag and the combined weight of the rider and bicycle.
While the calculated result will be quite accurate, this is only true if your inputs used in the calculation are themselves accurate. At a minimum, you will need your bike equipped with a power meter. Additionally, you must be mathematically inclined to perform such calculation. For many, this is not an easy task.
An alternative approach is used here. It uses the law of large numbers. No. Not large like this, 86482736. Instead, statistical analysis of past Whiteface Mountain and Mount Washington uphill bicycle race results have been compiled. The data collected is from the respective race inception, 1997 for Washington and 2001 for Whiteface, through 2005. This makes for a complete population of race results data for these two events.
A statistical linear regression model was derived from a sample consisting of cyclists who did both uphill races respectively within the same year. Looking at the graph below, this particular model fits the sample data quite well.
For those interested in some technical detail, the sample size is 160, which is a sufficiently large sample. This is so when considering this sample is taken from an already select sample of cyclists who are willing to ride their bicycles 8 miles up a mountain.
The correlation coefficient is 0.84443. One would expect these two races to be more positively correlated, e.g. a value closer to 1. For those who raced up Whiteface or Washington or both, you know that these climbs can be merciless for the ill-equipped or ill-prepared.
For instance, Whiteface Mountain has an average grade of 8% over 7.9 miles. Mount Washington's average grade is 12% over 7.6 miles. Mount Washington has the additional challenge of a gravel road section at the end of the race, which incidentally is also the steepest portion of the racecourse. It is also known for potentially harsh weather such as high winds. (It holds the world record for the highest recorded surface wind speed of 231 miles per hour). Incidentally, the weather for the Whiteface Mountain bike race has been less than favorable the past two years, 2004 and 2005, with high winds and excessive heat, respectively.
All it takes is one element like insufficient bicycle gearing, a big gust of wind to blow you off your bike or muscle cramps and your race time will suffer just as much as you did cycling up the mountain. For such folks, they learn quickly after just one race. The collected race data reflects this fact. Disparate race times are reflected in the regression data and explain the softening effect on strongly correlated race results.
Anyway, without further ado, below you will find the "Expected race results calculator." The only information needed is your known Whiteface Mountain or Mount Washington finishing time. This value must be expressed in minutes. Since this is an estimate calculation, entering a whole number is sufficient, but not absolutely necessary. For example, enter: 63.5 if your Whiteface time is 1 hour, 3 minutes and 30 seconds.
So what do these numbers mean? In technical terms the low/mid/high values represent the expected finish time within a 95% confidence interval (t-Statistic). In other words, you can expect your time, on average, with a good deal of certainty, to be within the high and low range of numbers provided here. The word “expected” is used here in a statistical context.
As noted earlier, this model is based on the experience of cyclists who’ve done both races, Whiteface and Washington within the same year. (Performing other types of quantitative comparative analysis, like race results across different years, has not been explored).
Of course, it goes without saying, if you lose say 25 pounds of body weight subsequent to racing up Whiteface or decide to race up Washington with a 39x23 low gear and with the rear brakes rubbing, well then, it is assured, with 100% confidence, that this model will provide you with an inaccurate result.
For questions or comments, please contact Peter O'Reilly.
Last updated 17 March 2006