Average pace in WKO "wrong", and how accurate do we need to be?

[Deleted User]

Hi everyone.  Been thinking about my pace data from the Nov. OS running test and have found some strange things.

• First, I noticed that the average pace as given by WKO is not a simple distance divided by time calculation.  In my run test I did 3.11 miles in 24:57, and the data tool gives an average pace of 8:01/mile.  However, WKO give an average pace of 8:15 per hour!!  I could not figure out why this was, then found this on the training peaks forum: "I know that this seems like an easy problem for us to fix, but it is actually pretty complicated.  We do not calculate avg pace by just dividing the duration and distance.  Instead, we look at the raw data given to us by the device, alayze the time series points, take each speed point and weight it for the delta to the next speed point, then avg that data.  We have done that in this case, but with the data given us it does not match the result if you take just the distance and duration and divide.  The number we are providing is the most accurate we can provide given the data we get from the device without ignoring the time series and just basing it on the total distance and duration.  Sorry for any confusion this causes." Just thought I would point that out to anyone else puzzled by this.
• Second, I say I ran 3.11 miles (a 5K) because that is what I mapped in MapMyRun.  My Suunto footpod actually said I ran 3.06 miles.  That is only off by 1.6%, and I doubt any other GPS or Footpod devices can be much more accurate.  However, using 3.06 miles I get and average pace of 8:09 min/mile, instead of 8:01 min/mile.  A difference of 8 seconds!  So I was going slightly faster than my foot pod said, and I covered slightly more ground (and/or mapmyrun is off).  I'm concerned because the data tool help file says to "stay as close as possible to the prescibed speed".  Do I just shoot for the speeds as best I can as provided by the data tool, and don't worry about the error?  Or should I add 8 seconds to the prescribed paces to account for the calibration error, and stay as close as possible to those, at least for the Z4 work?
• Finally, how do you guys set up pace zones in WKO or Training Peaks.  They want a range of paces, yet the data tool gives exact numbers.  I think I saw on one of these posts to go + or - 5 seconds.

Thanks for the help everyone!

Jim

mgraffeo

Jim, the short answer is that it's near impossible to run to an exact pace in training to the second, so being accurate to the second doesn't matter so much. Take a time for your 5k, make it on a course that's as repeatable as possible, put it in the data tool, and go with what it tells you. You'll see if you are improving.

As for zones in WKO, ignore them. They're irrelevant. The only thing that matters is to put your FTP in on the bike, and your Z4 pace from the data tool in for the run.

[Deleted User]

Thanks Mike. Only thing that matters because that's what it uses to do the TSS charts? I figured that I would enter my time for the 5k for threshold pace. No good? Use the Z4 pace?

Nemo Brauch

FWIW- the issues you describe with using a GPS device is one of the reasons I've been thinking about heading back to the track for my OS workouts. As much as I like the flexibility of running intervals on the road wherever I am, there's nothing quite like the truth you get with a simple clock against the 400 meter track

mgraffeo

The IF calculation is based on the FTP you put in. The ranges are for information, but not used for any calcs.

Threshold should be your 1-hour pace. Much closer to Z4, whereas 5k pace is usually in the 20-30 minute range.

mgraffeo

Nemo, that's why I love my treadmill!

[Deleted User]

A track would probably be better - just tough to find one that I could reliably use all winter in upstate NY!

Thanks again Mike - that makes sense.