Birkie Stats

Pictured here is a generic graph that relates finish time to age and depicts the various performance elements that will be examined in upcoming figures. The points on the graph are the mean finish times at the mid-point age of each of the age groups (excluding age groups from 80 on ).

First, a broad-brush summary can be obtained by dividing the performance curve into a younger period (21-52) and an older period (57-77) and calculating the average change in finish time/year for each period. The dividing point between the periods is somewhat arbitrary, but it works for most of the figures that follow. As will be seen, there really is a relatively long period of sustained performance that is then followed by a more marked decline starting somewhere within the ages from 50 to 60 .

The broad-brush approach, however, does not capture all the information that is retrievable from these data. The performance trends are much better depicted by modeling the data with polynomial equations. The reason for this is that fitted polynomials highlight the critical age periods when skiers are slowing down or speeding up. For example, the third-degree polynomial model used in this schematic detected two changes in direction. There was a short initial period where skiers were slowing down with age (finish times were increasing) that ended at point A (~age 25). From point A to point B (~age 45) skiers were actually skiing faster as they aged. Then, from point B on, skiers slowed down forevermore with every new birthday. Points A and B are referred to as local maxima and local minima, and they can be determined using differential calculus (more).

(More on aging athletes)

Skiers can hold out for a long time. It appears that good performance can be maintained throughout much of middle age and even longer, at least for endurance events like the Birkie. 

For the freestyle event, men have remarkable staying power through their middle years. There is an inflection point at 25 years, indicating that thereafter speeds are slowing down, but men add on average only 38 sec/year to their finish times during their younger period. In their older period, each birthday adds about 2 1/2 min.

For the classic race there is a pretty significant slow down until age 40 and then times level out. The local minimum at 47.3 years is the point when speeds start slowing again. The initial slow down may be, in part, an artifact stemming from the fact that many of the younger classic skiers have only skied the Birkie since the separate classic trail was established in 2008 (more). This issue will come up again in later figures.

Ignoring the initial blip in finish times, during their younger period, classic skiers add an average 44 sec to their time each year. In their older period, the increase is around 2 3/4 min/year.

Note: Finish times are standardized to a 55k classical race or a 51k freestyle race.

Women show the same general pattern as men. For the freestyle event there is an initial slowing until the inflection point at age 33.6 after which times remain fairly steady until the late 40s. For the classic event, speeds slow down until the age of 36, then speeds actually increase until age 50.3.

Ignoring the initial blips, the increase in finish time during the younger period averages 29 sec/year for freestyle and 44 sec/year for classic. The slow down during women's older period is more pronounced than seen in men - a little over 6 min/year for freestyle and 7 min/year for classic.

Note: Finish times are standardized to a 55k classical race or a 51k freestyle race.

So far, all we’ve looked at is the performance of an average skier. What about the best skiers in each age group? Are they better at maintaining performance as they age than the average skier? 

The following graphs show the same data as in the previous two graphs but for the mean of the top three finishers in each age group (ignoring the top overall finishers who have been removed from this analysis). Looking at only the top three skiers in each age group also eliminates the artifact of shifting age demographics encountered in the previous graphs.

These "elite-skier" graphs show remarkable staying power for top cross country skiers, especially classic skiers. For males in their younger period, the increase in finish time for freestyle averages about 39 sec/year. The inflection point at 37.3 is the age at which finish times begin their inexorable increase.

Classic male skiers in their younger period actually speed up as they age by by an average of -25 sec/year! While there is a slight slowing until the local maximum at age 26.8, there is a long period of faster and faster finish times until age 48.2.

On average over their older period, males in the freestyle event add a little under 6 1/2 min/year to their finish times, while classic skiers add a little under 7 1/2 min/year.

Even though the possible effect of shifting age demographics is removed by considering only the top 3 finishers in each group, the curve for classic skiers remains distinct from that of freestyle skiers because of the initial slowdown to the mid-20s. As will be seen in the next figure this distinction will remain for elite females as well.

Note: Finish times are standardized to a 55k classical race or a 51k freestyle race.

For women in their younger period, the increase in finish time for the freestyle event averages 36 sec/year. The inflection point at age 34 represents the age at which finish times will increase forevermore.

As for men in the classic event, women in their younger period actually speed up as they age, and do so at nearly twice the average rate for men, -43 sec/year. There is an initial slow down until the age 25.4, but then speeds ramp up again until age 45.5.

For elite women, it seems that the performance decline that comes with advancing age starts at an earlier age than for men and proceeds more rapidly. Women in their older period add a little under 10 1/2 min to their finish times each year for both the freestyle and classic events.

Note: Finish times are standardized to a 55k classical race or a 51k freestyle race.

If you are a "main pack" Birkie skier and wonder what effect age is having on your results, you can use this table to estimate your Birkie finish time based on your previous year's time. For example, if you are male, were of age 32 in last year's Birkie, and you ski the freestyle event, you can expect the aging effect will add 28 seconds to your finish time from the previous year.

The numbers in this table are derived from the polynomial models used in the previous graphs.

If you are an age-group contender for the Birkie and wonder what effect age is having on your results, you can use this table to estimate your Birkie finish time based on your previous year's time. For example, if you are female, were of age 41 in last year's Birkie, and you ski the classic event, you can expect to ski the Birkie 45 seconds faster than the previous year.

The numbers in this table are derived from the polynomial models used in the previous graphs.

Another way to look a longevity is to examine how performance declines with age after the point of inflection or local minimum (depending on the shape of the fitted polynomial), which represents that age after which speeds invariably slow down at an ever increasing rate. From the previous graphs, these points for the classic event range from ages 45.5 to 50.3, and for the freestyle event, the range is 25.0 to 37.3. So it seems reasonable to examine the decline in performance relative to a base level of performance set by the mean finish times of the 45-49 age group for the classic event and a base level set by the 35-39 age group for the freestyle event. Each point on this figure is merely the mean base level finish time divided by the mean finish time for each age group (represented by its mid-point age). The points were modeled with a second-order polynomial (more). 

For the classic event, female performance falls to a little under 70% by the time skiers reach their late 70s, while that of males falls to around 83%. It is apparent in this figure that female performance degrades more rapidly than male performance for the classic event.  This is a common observation in studies of aging master athletes across a variety of athletic endeavors (see Baker et al., 2010).

The performance decline curves for the freestyle event are similar to those for the classic event in that, again, female performance declines faster than male performance, although this conclusion seems particularly reliant on the polynomial model. The last point on the graph for males at age 77 pulls the model away from the curve for females and may exert undue influence.

It is not possible to compare  rates of decline for classic versus freestyle events since the baseline levels are different.

Performance decline analysis can also be applied solely to the top 3 finishers in each age group (ignoring the top overall finishers). For the classic event, the first notable observation is that performance decline extends to nearly 50% for both males and females by age 77 (i.e., finish times are nearly double those of the base level at age 45-49). This is likely due to the fact that these elite skiers are near their optimum level of performance and close to their prime age at the base level of 45-49, while the average skier depicted in the two previous figures operates at somewhat less than maximum potential at this age. Therefore, average performance of all skiers, as a percentage of the baseline, declines at a slower rate than for elite skiers.

The performance of females seems to drop off more steeply than males, although both end up near the same level by age 77.

This figure is similar to the previous figure for elite classic skiers with the exception that performance decline dips even further, to below 40% for females and below 50% for males, by the late 70s. Female performance again seems to decline more rapidly than male performance with age.

Here is an example of how you might use this and the previous three figures. Say you are a 62 year-old male who has entered the freestyle event. You've been competitive in your age group in the past and occasionally make it to the top 3. What finish time should be your goal to show that you are doing better than your physical condition, skill level, and chronological age would predict? 

From this figure, you note that if you are at the the fractional level of 0.8 or higher, your are doing well for your age and ability. Since the base level finish time for elite freestyle male skiers is 2:19:06, you should be happy to finish at 2:54 or better.

Repeating the same exercise, but for a recreational (not top 3) male freestyle skier of age 62, would suggest that a finish time of 4:48 would be an appropriate age-adjusted goal. 

This table uses the polynomial models that were fitted to performance degradation rates to predict how old skiers are when they decline by 25%, 50%, 75%, and 100% from their peak performance years (as defined by the base levels used in the previous figures). Recall that a 50% reduction corresponds to a doubling of finish time relative to the base level. A 100% reduction means that a skier is incapable of finishing the Birkie anymore.

The predictions are probably within reason for the 25% and 50% levels of decline, and perhaps the 75% level for the top 3 skiers. But the 100% level is way outside the range of the data and is provided merely as a conversation piece. In fact, the 75% level doesn't make much sense for the non-elite skiers since the predicted finish times may be beyond the cut-off times for completing the Birkie.

Still, the models do show that cross country skiers have the  potential to ski Birkie length events-well into their 80s and beyond.

In general, classic race times may have decreased somewhat over the past 16 years, while the same cannot be said of the freestyle race. From 2008 on, the classic and freestyle events had separate start waves and a separate trails to OO. So while the distances for the two races were equal to each other up to 2008, in 2008 the classic trail was about 2 km longer, and from 2009 on, about 4 km longer than the freestyle trail.

Finish times, however, in this graph are standardized to a 55k classical race and a 51k freestyle race for all years, so the convergence of the finish times for the two events probably indicates that after 2008 classic skiers were able to increase their overall pace, especially during the first half of the Birkie, because of less interference with skate skiers. The downward slope of the trend line suggests that the classic finish times have decreased at a rate of 4 min 24 sec/year.

There were several notably slow races, particularly 2001 when 8 inches of new snow fell from midnight until midmorning on race day. In 2002, the classic race was plagued by warm conditions and a badly deteriorated track. In 2014, 18 inches of snow fell in the two days before the race, and the temperature was well below zero at the start. 

More Birkie history

Note: Finish times are standardized to a 55k classical race or a 51k freestyle race.

There is a steady increase in finish times from the elite wave to wave 7 for both events, with each wave finishing about 27 minutes slower than the one preceding it for classic racers and about 23 minutes slower for freestyle racers.

For waves higher than 7, many racers may be experienced skiers, but they may also be first-time Birkie skiers without finish times from qualifying events that would have placed them in one of the earlier waves.

Note: Finish times are standardized to a 55k classical race or a 51k freestyle race.

For this database, each skier's wave placement was determined by bib number although in a few instances additional steps were needed to separate elite racers from wave 1 racers. The Birkie includes several honorary waves for founders, those who have skied 35 Birkies, and those older than 70. These honorary waves were not included in this table.

1997 was the first year when the classic and freestyle racers were distinguished from one another in the results, even though they still had common start waves based on prior finish times regardless of technique. Therefore, there were relatively few classic skiers in the elite and lower number waves until 2008 when the two races had separate starting waves based on separate finish-time criteria, and the new classic trail was first used. 

The current (2017) policy for assignment to the elite wave is that the top 200 male and 60 female finishers from the previous year's Birkie freestyle race, and the top 100 male and 40 female finishers from the previous year's classic race are automatically assigned a starting position in the elite waves for the current year. This policy has changed over the years, and it also seems that there have been special situations when notable skiers are allowed into the elite waves without having skied the Birkie in prior years.

The size of the first five waves of the freestyle race have remained relatively constant over the years, although it hard to really determine this since the Kortelopet skiers are excluded from this summary.

It's readily apparent that the classic race has become more popular, especially since 2008. 

The proportion of women has increased slightly over the years. Women made up 19% and 22% of the total classic racers in 1999 and 2016 respectively. For the freestyle race, women were 15% and 19% of the total in the two years. The proportion of women totaled over both races has increased from 15% in 1999 to 20% in 2016.

The proportion of classic skiers has increased dramatically over the years, from 4% (35 females and 150 males) in 1999 to 34% (434 females and 1547 males) in 2016. The classic event became more attractive after the new classic trail was established in 2008.

The increasingly popular classic race has been responsible for the much of the significant growth in total number of Birkie skiers. 

(data table)

The age of individual skiers was not provided in the PDF results for many years, so for the summaries involving ages, a gross-scale mean age was calculated based on the age-group classification of each skier.

The mean age over both genders and both techniques has not changed much, from 41.3 in 1999 to 44 in 2016. The mean age of men has increased fairly steadily over the years while that of women has remained steady or slightly declined in recent years. 

On average, classic skiers are about 2.4 years older than freestyle skiers, and women are about 5.1 years younger than men.

Further age breakdown:

Classic - female = 41.6

Classic - male = 47.9

Freestyle - female = 38.7

Freestyle - male = 43.6

The largest age groups for men are 45-49 for freestyle and 50-54 for classic. The picture is very different and a little more complicated for women. The youngest age group, 18-24, is actually the largest age group for freestyle and the numbers in each successive age group decrease slightly until 35-39. A second peak occurs at 45-49. The overall trend for classic is similar but the largest age group is 50-54. 

Age-group distributions are strikingly different for men and women, and these differences have changed over time, as can be seen in the following figures.

In the first 8-year period from 1999 to 2008, the freestyle race was a great deal more popular than the classic race for both men and women. Few women entered the classic race.

In the most recent 8-year period from 2009, the classic race attracted more skiers than in the earlier period for both men and women. The was also a marked increase in younger women doing either event. The largest age group for women was the 18-24 group for both the classic and freestyle events.

All things considered, the separate classic trail in 2008 and renewed interest in classic skiing has been very good for the Birkie.

While the average age of Birkie skiers has not changed much over 16 years, the distribution among age groups has changed dramatically.

There are far more skiers in the upper and lower age groups in 2016 than there were in 1999. This may be a very positive demographic trend for the future – a steady stream of new, young skiers, and an evermore fit and dedicated group of older skiers.

It may also be that the same group of skiers in their early 40s who were so prominent in 1999 have continued doing the Birkie. Their prominence is now, 18 years later, reflected in the older age groups.

In this graph, the number of younger (<30 year old) and older (50+ year old) skiers are compared over years for the classic race. It is apparent that there has been a dramatic increase in older men, and this increase has been going on for the last 16 races.

There is also an increase in younger and older women as well as younger men, but not nearly to the degree seen for older men. 

As seen for the classic race, the number of older men in the freestyle race has increased over years, but at a pace not much higher than for younger and older women and younger men.

Combined with the previous graph, it appears that older men are becoming more and more prominent in the Birkie, especially the classic race.

Each point in this figure depicts the difference in mean finish times between women and men for each year from 1999 to 2016 for the classic and freestyle events. All skiers are included.

The trendlines are pitched slightly downward suggesting that women are speeding up and/or men are slowing down such that the difference between genders is decreasing at a rate of 37 sec/year for freestyle and 12 sec/year for classic. These rates, however, are small enough, especially for the classic event, that they are essentially meaningless statistically.

Note: Finish times are standardized to a 55k classical race or a 51k freestyle race.

This figure also represents the difference between women's and men's finish times, as in the previous figure, but the difference is presented only for the top 3 finishers in each age group. 

It is obvious that a there has been a substantial narrowing of finish times for the classic race - 3 min 33 sec/year. This may be the result of a growing number of well-trained younger women skiers entering the classic event particularly after 2008 when the new classic trail was established.

While the freestyle trend line also sloped downward, the rate was statistically insignificant.

Note: Finish times are standardized to a 55k classical race or a 51k freestyle race.

For the tables in the "More demographics" section, the age group category "NA" applies those skiers for which it was not possible to assign an age group due to missing or undecipherable records. 

Also note that the age group categories have changed a bit over the years. For example, the 16-18 and 19-24 age groups only existed in 1999 and 2001. After 2001, racers had to be at least 18 years old (although it seems that a few 17 year olds sneaked in for some years).

For the age group summaries provided in Birkie stats, the top overall finishers, indicated by "TOP 3" or "TOP 6" were removed prior to analysis.

For the tables in the "More demographics" section, the age group category "NA" applies those skiers for which it was not possible to assign an age group due to missing or undecipherable records. 

Also note that the age group categories have changed a bit over the years. For example, the 16-18 and 19-24 age groups only existed in 1999 and 2001. After 2001, racers had to be at least 18 years old (although it seems that a few 17 year olds sneaked in for some years).

For the age group summaries provided in Birkie stats, the top overall finishers, indicated by "TOP 3" or "TOP 6" were removed prior to analysis.

For the tables in the "More demographics" section, the age group category "NA" applies those skiers for which it was not possible to assign an age group due to missing or undecipherable records. 

Also note that the age group categories have changed a bit over the years. For example, the 16-18 and 19-24 age groups only existed in 1999 and 2001. After 2001, racers had to be at least 18 years old.

For the age group summaries provided in Birkie stats, the top overall finishers, indicated by "TOP 3" or "TOP 6" were removed prior to analysis.

For the tables in the "More demographics" section, the age group category "NA" applies those skiers for which it was not possible to assign an age group due to missing or undecipherable records. 

Also note that the age group categories have changed a bit over the years. For example, the 16-18 and 19-24 age groups only existed in 1999 and 2001. After 2001, racers had to be at least 18 years old.

For the age group summaries provided in Birkie stats, the top overall finishers, indicated by "TOP 3" or "TOP 6" were removed prior to analysis.

For the classical race, there has been only one woman in the 75-79 age group over the last 16 years - Barbara Klippel from Hayward, WI., skied the classic race in two years, 2008 and 2012.

Note: Finish times are standardized to a 55k classical race or a 51k freestyle race.