For Whom the Dogs Bark

70F, 100%, N 1 mph.  Same as all week, will be similar tomorrow then start cooling off some more.  But 70F is not bad, I still sweat like a hog but it doesn't seem to have any other effect this late in the summer training cycle.  The plan today was to do 10 easy.  I ran the course that I had previously mapped out at 10.02 miles, but I took my new Garmin 210.  I liked several things about it.  It is the size of a regular watch, small and light, but there are some drawbacks.  The illumination has an automatic fade after about 10 seconds, so every time you want to see the watch you have to turn it back on.  I run 80% of my miles in the dark, so this won't work unless there is a setting that changes the fade function.  Also, the numbers on the display are small, hard to read for my bleary early-morning eyes. 

A much worse problem is accuracy.  It measured today's course at 9.7 miles.  My course has a total of 29 cul de sacs and 45 turns of 90 degrees or more.  So if a GPS device isn't taking measurements at short intervals it will measure this course significantly short.  I double-checked my previous computer measurement (10.02 miles) by putting the Google Earth setting at the smallest scale and measuring every step of the run. It took an hour, but there is a point to it.  It's not about distance -- there is no difference in training effect between running 9.7 miles or 10.0 miles.  The computer measurement, and I can think of no reason why it is not accurate at that setting, was 9.97 miles.  So the 210 measured short by 2.5%, which for this run was the difference between 8:43 per mile and 8:57 per mile, 14 seconds per mile, more than 6 minutes over the length of a marathon.  I can't train intelligently with that much discrepancy.  My old Garmin 305 measured the first 4 miles of this course exactly the same as Google Earth, so I will send it in for repair -- the 210 despite its attractive appearance is not going to cut it, although I will take it out again tomorrow on a straighter course to see what happens.

Today's exercise also supports my beef with course measurements by race directors.  If a GPS device is inaccurate, it is because it measures short, not long.  It measures in straight line segments; the more often it takes a reading the shorter the segment and the more accurate the measurement.  The +/- differential for satellite inaccuracy is a statistical error that doesn't favor short or long, and over a long course it will average out to almost zero.  So if you run the tangents and a course still measures long, it is long.

On the bright side, my average heart rate for the run was 142, max 151, which is really good news.  I have not gotten a heart rate reading in at least a month.  I had planned to do this as a recovery run at 9 flat because I have a hard run tomorrow, but I ran it at 8:43 with no significant heart rate drift, quite a bit better than earlier in the summer when I was recovering from my back injury.  Back then I was happy to get 10 at a 9:20 pace with 150 max.  So 8:43 today after running garages yesterday and 20 on Wednesday represents bona fide progress.