Re: RC: Re: RE: RE: Cutoff Times

[David LeBlanc <dleblanc@mindspring.com>]

I've written a script that has pulled down the results for 1058 AERC rides,
and then processed the HTML pages to pull out a lot of interesting
information.  You'll be seeing more posts about this once I get done
crunching more of the data.

I've been primarily looking at completion rates, and what factors affect
them, but ride speed is also interesting.  Unfortunately, not everyone
reports individual times, so I can't work with that, but everyone does
report the winning time.  Couple this with the distance, and you've got
speed for the winner.  If I'm not mistaken, this is _ride_ time, which
excludes vet check holds.

For both 25's and 50's, winning speeds ranged from 4MPH to about 15MPH.
100's 
ranged from 5MPH to almost 12MPH.

Average winning speed was 8.9 MPH for 25 mile races, 9.6 MPH for 50 mile
races, and 8.2 MPH for 100s.

I find this interesting, since it is the reverse of what I would have
expected to find between 25's and 50's.  One of the reasons I embarked on
this project (other than geekiness) was that someone wrote me and was
concerned that perhaps LD riders were overriding their horses, and that LD
(particularly fast rides) could pose a larger danger to horses than longer
distances where you presumably have to ride slower.  So far, the data isn't
supporting that conclusion, but more on that later.  Perhaps overriding
one's horse is more a function of people than ride distance, and the horse
is more likely to emerge undamaged from being overridden for 25 miles than
50, but I'm guessing.

I wouldn't have expected the _winning_ speeds for 25's to be lower than the
winning speeds for 50's, but that's what the data tells me.

Another interesting bit of information is that completion rate cannot be
predicted from the winning rider's speed - there's no valid statistical
relationship there. To be precise, I don't know if the F-statistic is
valid, but the R^2 is really low - in English, this means that there may be
a valid relationship, but that it doesn't predict very much of the
differences.  The interesting thing is that the lowest completion rate
within a given range of winning speeds does go up as ride speed increases -
for example, above a 12 MPH  winning speed, there were no races with less
than 2/3 of the riders completing.  For winning speeds of less than 8 MPH,
completion rates dropped as low as 25% (rides where no one completed
excluded).

At 09:10 AM 2/3/00 PST, Nancy Mitts wrote:
>Yes, Truman, that is exactly how I figure it, using the average amount of 
>hold times we run into. Add in the time it takes to stop for water, extra 
>sponging, etc. I consider 5 mph the MINIMUM average speed it takes to finish 
>in time.
>I believe we did a 100 one time that allowed 3 hours to cover the 1st 21 
>miles. 7 mph. Seems slow enough? 


David LeBlanc
dleblanc@mindspring.com