Cadel versus Schleck

output

After 167 of 167 riders:
  Andy Schleck's predicted time is 60:05  (based on 2010 TDF ride for yellow, mean - 0.5 std.dev)
  Cadel's predicted time is 58:22  (based on 2008 TDF ride for yellow, mean - 1.3 std.dev)
  Cadel would be faster by 1:43

tdf2011.py

#!/usr/bin/env python
# -*- coding: UTF-8 -*-

from math import sqrt
from urllib2 import urlopen

def meanstdv(x):
    """ Returns the mean and standard deviation of the given values.
        From <http://www.physics.rutgers.edu/~masud/computing/WPark_recipes_in_python.html>
    """
    n, mean, std = len(x), 0, 0
    for a in x:
      mean = mean + a
    mean = mean / float(n)
    for a in x:
      std = std + (a - mean)**2
    std = sqrt(std / float(n-1))
    return mean, std

def secs(x):
  if ":" in x:
    h,m,s = x.split(":")
    return int(h,10)*3600 + int(m,10)*60 + int(s,10)

def schleck(mean,std):
  return int(mean - (0.5 * std))

def cadel(mean,std):
  return int(mean - (1.3 * std))

def get_current_times():
  f = urlopen("http://www.letour.fr/2011/TDF/LIVE/us/2000/chrono/25.html")
  return [secs(x[28:-6]) for x in f.readlines() if 'col_5' in x][2:]

times = get_current_times()
mean, stddev = meanstdv(times)
st = schleck(mean, stddev)
cd = cadel(mean, stddev)

print "After %d of 167 riders:" % len(times)
print "  Andy Schleck's predicted time is %d:%02d  (based on 2010 TDF ride for yellow, mean - 0.5 std.dev)" % divmod(st, 60)
print "  Cadel's predicted time is %d:%02d  (based on 2008 TDF ride for yellow, mean - 1.3 std.dev)" % divmod(cd, 60)
print "  Cadel would be faster by %d:%02d" % divmod(st - cd, 60)