A proposed method of automatically calculating X-country course
difficulty from team race data.
    Background:
  The only reference I could find of anybody trying to rate XC courses
was Marc Bloom's Harrier publication (Vol 18 No 4 Nov. 12 1993)
       The Harrier, PO Box 41, Marlboro, NJ 07746
BTW this is an excellent publication and appears to be the definitive source
for High School XC info.( You also get great information from
Walt Murphy's X-country X-press.PO box 750994, Forest Hills, NY 11375) 

      He proposed the following for standard 5k courses:
	     Grade  Boys (best times) Girls (best times)
		A       15:00           18:00
		B       15:15           18:15
		C       15:30           18:30
		D       15:45           18:45
		E       16:00           19:00
		F       16:15           19:15
		G       16:30           19:30         
      These are average times produced by the best runners in a state.
 
	If we assumed an easy "A" course was equivalent to a an outdoor 5k
on a track, then a 15:00 would have an equivalent Purdy point value of 832.
This value would correspond to a 4:18 mile time, 9:20 two mile or
8:38 3k.  These values are not to far from Downin's and Mortimer's best times
(3k) : 8:32.04, 8:33.80. (Downin & Mortimer were 1 & 2 in the Ft Locker
HS championship at Balboa Park, San Diago, CA in times 14:58, 15:03)
An 8:32 3k has a purdy pt. value of 852, assuming Downin's performance 
at the footlocker final (14:58 836 Purdy) was at least as good.
Then running a 5k at Balboa Park course is 1.019 (856/832) as difficult as a 3k
track performance. Typically XC courses are more difficult than a track
but some courses, like on flat asphalt roads can be easier than a track.

   With this in mind I propose the following difficulty factors for
average courses.

       The proposed system uses a numerical difficulty rating based on 
scaling the purdy points for a course.
			time   Purdy  difficulty factor
		A       15:00  832    1.020  (849/832)
		B       15:15  806    1.053  (849/806)       
		C       15:30  781    1.087  (849/781)      
		D       15:45  757    1.122  (849/757)      
		E       16:00  733    1.158  (849/733)      
		F       16:15  711    1.194  (849/711)      
		G       16:30  688    1.234  (849/688)       

   Hence to calculate a X-country runner's true performance in Purdy points,
   one calculates the purdy points for time and distance, and then
   multiplies that value times the course difficulty factor.
  
  If we assume that Balboa Park is an "A" course with a difficulty factor
 of 1.020 then Matt Downin's time of 14:58 had a purdy point value of
      14:58 (836 X 1.020) = 853 

  This method of using purdy points to establish course difficulty factors
can be used to caculate "relative course" difficulty values if a number of
runners (like a X-country team) run on two courses within a short
period of time (say 14 days). If you make some assumptions like the average
performance should be about the same. The more runners you have the
more accurate the course difficulty becomes.

    I have written a program to do this and hope to establish a course
database on the internet.  Since there will always be runners that
have "good" races and "bad" races, my program throws out the hi and lo
performance changes when comparing the runners two race performances.

 Using the 16 boys and girls from each regional Footlocker championship
 the "relative" course values compared to Balboa Park are as follows:
				Relative Diff.     True dif. (Balboa 1.020)
 Kenosha Wi.                         0.991          1.011  (.991X1.020)
 Van Cortland Park, NY               1.037          1.058
 MaAlpine Greenway, Charlotte, NC    0.940           .959
 Woodward Park, Fresno, Cal.         0.982          1.002

   This says that the course in Charlotte is the "easiest". Now it could be     
 that all the runners from the south just had "bad" days when they went  
 to California. 
   Assuming the courses stay the same using more runners to "average"
   in the performances will produce more accurate results.

   Stay tuned for more information about the X-country Course database.
   Please e-mail me if you would like more info or have any comments
   about this proposed system or you are interested in these programs.
		  Patrick Hoffman
		  phoffman@cs.uml.edu

To see the output of the program calculations for the 4 regional HS footlocker and the same 64 runners at Balboa click here detail on Balboa XC-course ratings for each runner you will see the time, pace and purdy point for each race. The differences of the purdy points for the comparitive races will also be shown.