Road Race Stats - Marathons & Other Running Races

Saturday, June 04, 2005

Boulder 10K and High Altitude Effects

With an elevation of around 5300 feet, the Bolder Boulder 10K may provide some insights into how high altitude effects running. Before crunching the numbers, I first did some research.

I found this useful CoolRunning thread in which a runner asked about what difference he might experience when running the Johannesburg marathon which is at a 6000ft elevation. The replies gave some factors to consider such as the acclimatization time. One reply mentioned that the oxygen levels are about 93% of what they are at sealevel so that can be used as a rough estimate of slowdown. The replies also provided some useful resources such as this Runworks calculator and this Peak Performance article on high altitude training.

Due to the format of the Boulder's 10K results, I wasn't able to get total averages. So I decided to look at the times of individuals and see how much faster they had run in a previous lower-altitude 10K.

First, I looked at the elite runners. The TD Banknorth Beach to Beacon 10K held last August 1st near Portland Maine seemed to provide a good comparison to the Boulder 10K. The Beach to Beacon is near sealevel and the temperature that day was warmer than last week's Boulder 10K. So you would think that any slowdown at Boulder should be due to the altitude.

I found 3 elite men and 1 elite woman who ran both races. All four had faster times at Beach to Beacon 10K. Below is a comparison of the times:

Runner2005 Boulder 10K2004 Beach to Beacon 10KChange
Thomas Kiplitan29:3227:391:57
Gilbert Okari29:3927:352:04
Tekeste Kebede29:3629:110:25
Masako Chiba (F)33:5431:512:03
Average Change1:37

So for these 4 elite athletes, the average slowdown was 1:37. The Runworks Running Calculator predicted about a 46 second slowdown.

I then looked at average runners. I pulled down the Texas runners who had run the last 3 Bolder Boulder 10Ks from the Bolder Boulder Citizens Race Results page. I then used this data with the data from the last three years of the Austin Capitol 10K results to find those runners who had run both races in the same year. So if you ran the Austin Capitol 10K, how much slowdown should you expect when you run the Boulder 10K?

Austin is not at sealevel but it's close. Its elevation is around 500ft. The temperature differences between the races were minimal for 2005 and 2003. Like Boulder, Austin temps during the race were in the 50's. In 2004, Austin was warmer with a race time temp of 71°F.

I thought I would find more runners who had run both races. I found only 9 this year, 14 in 2004, and 14 in 2003. The sample is a little larger than the elites but still not as much as I had hoped. Another factor that may complicate comparisons is the Boulder finish times. There is no distinction between chip and gun finish times. Boulder uses a wave system to start, so I'm assuming the Boulder times are essentially chip times.

The tables below provide the detailed results. I tried to summarize these results in two ways. First I took the average of the runners whose times were within 10 minutes of their Cap10K and Boulder 10K times. I assumed other factors besides elevation may have affected those runners who had more than 10 minute differences. I also looked specifically at runners who ran both under 60 minutes (in addition to less than a 10-minute delta.) Below are the results.

Slowdown of Runners at Boulder 10K vs Austin's Cap10K
Excluding those with >10-min deltas:
2005 : 7 runners averaged: 4:23
2004 : 11 runners averaged: 0:52
2003 : 14 runners averaged: 1:56
Average: 2:24
Excluding those with >10-min deltas and times over 60 minutes:
2005 : 3 runners averaged: 1:05
2004 : 7 runners averaged: 0:57
2003 : 11 runners averaged: 1:33
Average: 1:12

Including over-60-minute runners, the average of 2:24 minute slowdown is close the the slowdown predicted by Runworks calculator of 2:16 for a 60-minute runner. The average 1:12 slowdown of those who ran under 60 minutes is quite a bit less. Perhaps if I had a larger sample, it would be closer. The detailed results are below.


Boulder vs Austin ChangeNumberPercent
over 10 min faster0
5 to 10 min faster0
2 to 5 min faster0
0 to 2 min faster0
0 to 2 min slower222.2
2 to 5 min slower222.2
5 to 10 min slower333.3
over 10 min slower222.2


Boulder vs Austin ChangeNumberPercent
over 10 min faster17.1
5 to 10 min faster0
2 to 5 min faster0
0 to 2 min faster535.7
0 to 2 min slower214.3
2 to 5 min slower428.6
5 to 10 min slower0
over 10 min slower214.3


Boulder vs Austin ChangeNumberPercent
over 10 min faster0
5 to 10 min faster0
2 to 5 min faster0
0 to 2 min faster321.4
0 to 2 min slower642.9
2 to 5 min slower214.3
5 to 10 min slower321.4
over 10 min slower0

Technorati Tags: , ,


Post a Comment

<< Home