Sunday, July 27, 2014

Career Touchdown Comparisons

Career Touchdowns

Consider two datasets – career touchdowns for the top two running backs chosen in the NFL draft and career touchdowns for the top two quarterbacks chosen in the NFL draft. 

How do the shapes of the distributions for career TDs for top-two running backs and career TDs for top-two quarterbacks differ?  

Data:  The data for this post can be found at


The datasets cover the top two running backs picked and the top two quarterbacks picked for every draft from 1970 to 2002.

Analysis: 

Below I present estimates for the mean, median, standard deviation, skew, kurtosis, and percent of observations less than 10% of the average for our sample of running backs and quarterbacks.

Running
  Backs
Quarterbacks
Mean
41.00
103.53
Median
27.00
82.50
Standard Deviation
36.98
101.54
Skew
1.17
1.26
Kurtosis
1.07
2.15
Percent < 10% of average
7.81%
21.88%
Minimum
0
0
Maximum
164
491



Average TDs is 41 for our sample of running backs versus 103 for our sample of quarterbacks.

The median TDs for our samples are lower than the mean TDs.  Skew is higher for quarterbacks than running backs. 

The reason for the larger skew for quarterbacks is that a larger number of quarterbacks have relatively few TDS.   (Many top-pick quarterbacks don’t get any playing time.) 

I believe the higher kurtosis of the quarterback career TD total statistics largely stems from the fatter left tail of the quarterback career total statistics but perhaps there are other differences in the data sets driving the difference in kurtosis statistics.



What are the potential implications of this comparison towards the decision to use a top draft pick to obtain either a quarterback or a running back?

The risk of coming up empty is a lot higher when drafting quarterbacks than when drafting running backs.


#NFL DRAFT
#KURTOSIS
#SKEW
#SKEWNESS



No comments:

Post a Comment