Tuesday, June 28, 2016

Fun with Standard Deviation



The graph above shows how many standard deviations from the field average my event times were for each of my triathlons (z-score). Negative is slower and positive is faster than the average. As you can see, my first race in Chaska was a turd. Since then my sprints have been above average, my olympics below average (but respectable) and this graph makes my 70.3 look like a fail.

Look at the graph a bit closer and you'll see that my total time for my last two sprints we almost the same with respect to the field average.
Hopkins:     0.368 standard deviations = 64th percentile and
New Bri Tri: 0.330 standard deviations = 63rd percentile

Similarly, the rankings on my two Olympic length races are nearly the same.
Bemidji:        -0.341 = 37th percentile and
Lake Marion: -0.374 = 35th percentile

Each of those races had their own quirks (length, wind, humidity, hills, give-up factor (Marion), holding back factor (New Bri)), but the numbers are too close to resist.  So, I had an average drop off of 0.706 standard deviations from my sprint times to my my Olympic times.  If I had the same drop off going from the Olympic to the 70.3 length, that would have put me at -1.064 = 14th percentile. As it happens, my time came in at -1.309 = 19th percentile when compared to the average.  It's more so the back of the pack instead of the middle of the pack, but I'll take it.

Also, my Liberty long course bike didn't drop off at all when compared to my Olympic effort.
Bemidji (20 miles):  -0.562 = 29th percentile
Marion (26 miles):   -0.515 = 30th percentile
Liberty (55 miles):   -0.562 = 29th percentile

I kinda laid an egg on the Liberty swim, but we'll try to get'em next time. Anyhoo, even without the numbers I'm happy with my first 70.3 effort; the numbers just make it a little more fun.


No comments:

Post a Comment