tech guy in midtown

I was curious to see how my simple strategy for the weighted fantasy football pickem would do in previous years.  I found an archive of NFL point spreads, over under lines, and game results on goldsheet.com.  I don’t know how accurate the data are, but I wrote a little Python script to parse it.  The formatting of the data is pretty rough, so I’ve put a CSV version up on Google Docs.

Here’s a tabular summary of the results.  The key column is Percent, which shows the percentage of available points (MaxPosib) that my strategy got in each year.

    Year      Min      Max      Avg    Total   MaxPosib    Percent
==================================================================
    1993       27      103    60.94     1097       1466      74.83
    1994       46       95    64.82     1102       1500      73.47
    1995       50      114    81.18     1380       1747      78.99
    1996       51      119    82.82     1408       1748      80.55
    1997       52      114    82.24     1398       1745      80.11
    1998       56      120    86.94     1478       1734      85.24
    1999       65      115    91.06     1548       1959      79.02
    2000       62      113    90.53     1539       1959      78.56
    2001       43      117    90.29     1535       1960      78.32
    2002       62      136    95.59     1625       2088      77.83
    2003       72      134    98.53     1675       2088      80.22
    2004       67      129    95.47     1623       2088      77.73
    2005       77      127   101.47     1725       2088      82.61
    2006       65      134    89.59     1523       2091      72.84
    2007       56      136   103.41     1758       2091      84.07
------------------------------------------------------------------
 average    56.73   120.40    87.66  1494.27    1890.13      78.96
std.dev.    12.20    11.84    11.67   189.42     211.33       3.39
     min    27.00    95.00    60.94  1097.00    1466.00      72.84
     max    77.00   136.00   103.41  1758.00    2091.00      85.24

So I expect to score 79% of the total points available this year if I use my strategy of assigning weights to the favored team, ordered by point spread.

The next step is obviously to try some other strategies and see what the results are.  First I want to see how poorly a completely random strategy does.