Basic HTML Version
© 2014 Pacific Crest
341
YOUR
TURN!
Scenario:
The following data set includes the WHIP ratio and the number of wins for
American League teams during the 2013 MLB season. WHIP corresponds to
the number of walks plus hits per inning pitched. The higher a team’s WHIP,
the worse the performance of the pitchers. What is the likely WHIP ratio for a
team that had only 50 wins during the season?
Team
WHIP Wins
Baltimore
1.32 85
Boston
1.30 97
Chicago Sox
1.33 63
Cleveland
1.33 92
Detroit
1.25 93
Houston
1.49 51
Kansas City
1.27 86
LA Angels
1.38 78
Team
WHIP Wins
Minnesota
1.41 66
NY Yankees
1.31 85
Oakland
1.22 96
Seattle
1.33 71
Tampa Bay
1.23 92
Texas
1.28 91
Toronto
1.34 74
O
ops
! A
voiding
C
ommon
E
rrors
●
Assuming the relationship is linear when it is not
Example
: Ice Cream Sales based upon Temperature
$0
$100
$200
$300
$400
$500
$600
$700
10
15
20
25
30
35
40
45
Sales
Temperature °C
Why?
The data could be linear from 10 to 25 degrees or between 30 and 45 degrees but not
between 10 and 45 degrees. The correlation will be close to 0 and linear regression
line be flat around $300 to $400 which doesn’t provide any useful information or
predictive capability.
7.4 Simple Linear Regression