Multiple Coefficient of Determination

Least Squares Method

 FINAL EXAM 40 Marks

Unit 1: Single Sample Z test for mean

Unit 2: Single Sample Z test for proportion
Unit 3: t test
Unit 4: F test
Chapter 14, 15 Unit 5: Regression Analysis (linear, multiple regr)
Chapter 13 Unit 6: ANOVA
Unit 7: Chi Square test for Independence
Unit 8: Chi Square test for Goodness of Fit
X1 X2 Y a. Develop an estimated regression equation relating y to x1. Predict y if
30 12 94 b. Develop an estimated regression equation relating y to x2. Predict y if
47 10 108 c. Develop an estimated regression equation relating y to x1 and x2. Pre
25 17 112 and x2 =_x0002_ 15.
51 16 178
40 5 94 SUMMARY OUTPUT
51 19 175
74 7 170 Regression Statistics
36 12 117 Multiple R 0.47066597982237
59 13 142 R Square 0.22152646456215
76 16 211 Adjusted R Squar 0.12421727263242
Standard Error 38.4374261644808
Observations 10
Y =fun(x1)+c estimated reg equ
y=fun(x2) ANOVA
y = fun(x1,x2) df SS MS
Regression 1 3363.414 3363.414
equ R2 adj p value Residual 8 11819.49 1477.436
Total 9 15182.9

Coefficients Standard Error t Stat

Intercept 85.2171016102166 38.35196 2.221975
X2 4.32148806218767 2.864161 1.508815

SUMMARY OUTPUT

Regression Statistics
Multiple R 0.81242546193166
R Square 0.66003513119487
Adjusted R Squar 0.61753952259422
Standard Error 25.4009168303564
Observations 10

ANOVA
df SS MS
Regression 1 10021.25 10021.25
Residual 8 5161.653 645.2066
Total 9 15182.9

Coefficients Standard Error t Stat

Intercept 45.0593689924234 25.41815 1.772724
X1 1.94357118624901 0.493161 3.941046
SUMMARY OUTPUT

Regression Statistics
Multiple R 0.9620421485435
R Square 0.92552509557419
Adjusted R Squar 0.90424655145253
Standard Error 12.7096421575264
Observations 10

ANOVA
df SS MS
Regression 2 14052.15 7026.077
Residual 7 1130.745 161.535
Total 9 15182.9

Coefficients Standard Error t Stat

Intercept -18.3682675757143 17.9715 -1.022077
X1 2.0101852240573 0.247119 8.134483
X2 4.73781182041831 0.94844 4.995375
 ating y to x1. Predict y if x1= _x0002_ 45.
lating y to x2. Predict y if x2= _x0002_ 15.
ating y to x1 and x2. Predict y if x1= _x0002_ 45

F Significance F
2.276521 0.169781

P-value Lower 95%Upper 95%Lower 95.0%

Upper 95.0%
0.057006 -3.222682 173.6569 -3.222682 173.6569
0.169781 -2.283279 10.92626 -2.283279 10.92626

F Significance F
15.53184 0.00429

P-value Lower 95%Upper 95%Lower 95.0%

Upper 95.0%
0.114211 -13.55498 103.6737 -13.55498 103.6737
0.00429 0.806339 3.080803 0.806339 3.080803
 6.980515

F Significance F
43.4957 0.000113

P-value Lower 95%Upper 95%Lower 95.0%

Upper 95.0%
0.340762 -60.86412 24.12758 -60.86412 24.12758
8.193E-05 1.425842 2.594529 1.425842 2.594529
0.001573 2.495108 6.980515 2.495108 6.980515
a. The estimated regression equation is

ŷ = 45.06 + 1.94x1

An estimate of y when x1 = 45 is

ŷ = 45.06 + 1.94(45) = 132.36

b. The estimated regression equation is

ŷ = 85.22 + 4.32x2

An estimate of y when x2 = 15 is

ŷ = 85.22 + 4.32(15) = 150.02

c. The estimated regression equation is

ŷ = -18.37 + 2.01x1 + 4.74x2

An estimate of y when x1 = 45 and x2 = 15 is

ŷ = -18.37 + 2.01(45) + 4.74(15) = 143.18

Weekly Gross Televison Newspaper
Revenue Advertising Advertising
($1000s) ($1000s) ($1000s)

96 5 1.5
90 2 2
95 4 1.5 The owner of Showtime Movie Theater
92 2.5 2.5 as a function of advertising expenditure
95 3 3.3
94 3.5 2.3 a. Develop an estimated regression equ
94 2.5 4.2 the independent variable.
94 3 2.5 b. Develop an estimated regression equ
advertising as the independent variable
c. Is the estimated regression equation
the same in part (a) and in part (b)? Inte
d. Predict weekly gross revenue for a w
and $1800 is spent on newspaper adve

SUMMARY OUTPUT

Regression Statistics
Multiple R
R Square
Adjusted R Square
Standard Error
Observations

ANOVA

Regression
Residual
Total

Coefficients
Intercept
Televison
Advertising
($1000s)
Newspaper
Advertising
($1000s)

The regression equation is

Revenue = 83.2 + 2.29 TVAdv + 1.30 Ne
wner of Showtime Movie Theaters, Inc., would like to predict weekly gross revenue
unction of advertising expenditures. Historical data for a sample of eight weeks follow.

elop an estimated regression equation with the amount of television advertising as

dependent variable.
velop an estimated regression equation with both television advertising and newspaper
tising as the independent variables.
he estimated regression equation coefficient for television advertising expenditures
me in part (a) and in part (b)? Interpret the coefficient in each case.
dict weekly gross revenue for a week when $3500 is spent on television advertising
1800 is spent on newspaper advertising?

MARY OUTPUT

egression Statistics
0.958663
0.919036
0.88665
0.642587
8

df SS MS F Significance F
2 23.43541 11.7177 28.37777 0.001865
5 2.064592 0.412918
7 25.5

Coefficients
Standard Error t Stat P-value Lower 95%Upper 95%Lower 95.0%
Upper 95.0%
83.23009 1.573869 52.88248 4.572E-08 79.18433 87.27585 79.18433 87.27585

2.290184 0.304065 7.531899 0.000653 1.508561 3.071806 1.508561 3.071806

1.300989 0.320702 4.056697 0.009761 0.476599 2.125379 0.476599 2.125379

egression equation is
ue = 83.2 + 2.29 TVAdv + 1.30 NewsAdv
The regression equation is
Revenue = 88.6 + 1.60 TVAdv

Predictor Coef SE Coef T P

Constant 88.638 1.582 56.02 0.000
TVAdv 1.6039 0.4778 3.36 0.015

S = 1.215 R-Sq = 65.3% R-Sq(adj) = 59.5%

Analysis of Variance

Source DF SS MS F P
Regression 1 16.640 16.640 11.27 0.015
Residual Error 6 8.860 1.477
Total 7 25.500

b.

The regression equation is

Revenue = 83.2 + 2.29 TVAdv + 1.30 NewsAdv

Predictor Coef SE Coef T P

Constant 83.230 1.574 52.88 0.000
TVAdv 2.2902 0.3041 7.53 0.001
NewsAdv 1.3010 0.3207 4.06 0.010

S = 0.6426 R-Sq = 91.9% R-Sq(adj) = 88.7%

Analysis of Variance

Source DF SS MS F P
Regression 2 23.435 11.718 28.38 0.002
Residual Error 5 2.065 0.413
Total 7 25.500

c. No, it is 1.60 in part (a) and 2.29 above. In part (b) it

d. Revenue = 83.2 + 2.29(3.5) + 1.30(1.8) = $93.56 or $93,560

 F P
11.27 0.015

F P
28.38 0.002

(a) and 2.29 above. In part (b) it represents the marginal change in revenue due to an increase in television advertising with newspaper advertising held co

8) = $93.56 or $93,560
with newspaper advertising held constant.
Team Conf Yds/Att Int/Att Win% The National Football League (NFL) records a varie
Arizona CarNFC 6.5 0.042 50 and teams. To investigate the importance of passin
Atlanta FalNFC 7.1 0.022 62.5 team, the following data show the conference (Co
Carolina P NFC 7.4 0.033 37.5 attempt (Yds/Att), the number of interceptions thr
Cincinnati AFC 6.2 0.026 56.3 of games won (Win%) for a random sample of 16 N
Detroit Lio NFC 7.2 0.024 62.5 (NFL website, February 12, 2012).
Green Bay NFC 8.9 0.014 93.8
Houstan TeAFC 7.5 0.019 62.5 a. Develop the estimated regression equation that
IndianapoliAFC 5.6 0.026 12.5 of games won given the average number of passin
Jacksonvill AFC 4.6 0.032 31.3 b. Develop the estimated regression equation that
MinnesotaNFC 5.8 0.033 18.8 of games won given the number of interceptions t
New EnglanAFC 8.3 0.02 81.3 c. Develop the estimated regression equation that
New OrleanNFC 8.1 0.021 81.3 of games won given the average number of passin
Oakland RaAFC 7.6 0.044 50 of interceptions thrown per attempt.
San Franci NFC 6.5 0.011 81.3 d. The average number of passing yards per attem
Tennessee AFC 6.7 0.024 56.3 and the number of interceptions thrown per attem
Washington NFC 6.4 0.041 31.3 regression equation developed in part (c) to predic
the Kansas City Chiefs. (Note: For the 2011 season
7 wins and 9 losses.) Compare your prediction to t
by the Kansas City Chiefs.
 ll League (NFL) records a variety of performance data for individuals
tigate the importance of passing on the percentage of games won by a
data show the conference (Conf), average number of passing yards per
e number of interceptions thrown per attempt (Int/Att), and the percentage
%) for a random sample of 16 NFL teams for the 2011 season
ary 12, 2012).

ated regression equation that could be used to predict the percentage

the average number of passing yards per attempt.
ated regression equation that could be used to predict the percentage
the number of interceptions thrown per attempt.
ated regression equation that could be used to predict the percentage
the average number of passing yards per attempt and the number
wn per attempt.
ber of passing yards per attempt for the Kansas City Chiefs was 6.2
nterceptions thrown per attempt was .036. Use the estimated
developed in part (c) to predict the percentage of games won by
s. (Note: For the 2011 season the Kansas City Chiefs’ record was
Compare your prediction to the actual percentage of games won
 a. The Minitab output is shown below:

The regression equation is

Win% = - 58.8 + 16.4 Yds/Att

Predictor Coef SE Coef T P

Constant -58.77 26.18 -2.25 0.041
Yds/Att 16.391 3.750 4.37 0.001

S = 15.8732 R-Sq = 57.7% R-Sq(adj) = 54.7%

Analysis of Variance

Source DF SS MS F P
Regression 1 4814.3 4814.3 19.11 0.001
Residual Error 14 3527.4 252.0
Total 15 8341.7

Unusual Observations

Obs Yds/Att Win% Fit SE Fit Residual St Resid

14 6.50 81.30 47.77 4.24 33.53 2.19R

R denotes an observation with a large standardized residual.

b. The Minitab output is shown below:

The regression equation is

Win% = 97.5 - 1600 Int/Att

Predictor Coef SE Coef T P

Constant 97.54 13.86 7.04 0.000
Int/Att -1600.5 484.6 -3.30 0.005

S = 18.3008 R-Sq = 43.8% R-Sq(adj) = 39.8%

Analysis of Variance

Source DF SS MS F P
Regression 1 3652.8 3652.8 10.91 0.005
Residual Error 14 4688.9 334.9
Total 15 8341.7

Unusual Observations

Obs Int/Att Win% Fit SE Fit Residual St Resid

 8 0.0260 12.50 55.93 4.60 -43.43 -2.45R

R denotes an observation with a large standardized residual.

c. The Minitab output is shown below:

The regression equation is

Win% = - 5.8 + 12.9 Yds/Att - 1084 Int/Att

Predictor Coef SE Coef T P

Constant -5.76 27.15 -0.21 0.835
Yds/Att 12.949 3.186 4.06 0.001
Int/Att -1083.8 357.1 -3.03 0.010

S = 12.6024 R-Sq = 75.2% R-Sq(adj) = 71.4%

Analysis of Variance

Source DF SS MS F P
Regression 2 6277.0 3138.5 19.76 0.000
Residual Error 13 2064.7 158.8
Total 15 8341.7

Source DF Seq SS
Yds/Att 1 4814.3
Int/Att 1 1462.8

Unusual Observations

Obs Yds/Att Win% Fit SE Fit Residual St Resid

8 5.60 12.50 38.57 5.32 -26.07 -2.28R

R denotes an observation with a large standardized residual.

d. The predicted value of Win% for the Kansas City Chiefs is

Win% = - 5.8 + 12.9(6.2) – 1084(.036) = 35%

With 7 wins and 9 loses, the Kansas City Chiefs won 44% of the g
al St Resid
53 2.19R

dardized residual.
sas City Chiefs won 44% of the games they played. The predicted value is somewhat lower than the actual value.
Ship Overall Shore ExcuFood/Dining The Condé Nast Traveler Gold List for 2
Seabourn Odyssey 94.4 90.9 97.8 ships (Condé Nast Traveler website, M
Seabourn Pride 93 84.2 96.7 each ship received based upon the res
National Geographic E 92.9 100 88.5 Choice Survey. Each score represents t
Seabourn Sojourn 91.3 94.8 97.1 excellent or very good on several criter
Paul Gauguin 90.5 87.9 91.2 An overall score was also reported and
Seabourn Legend 90.3 82.1 98.8 Seabourn Odyssey, has an overall score
Seabourne Spirit 90.2 86.3 92 for Food/Dining.
Silver Explorer 89.9 92.6 88.9
Silver Spirit 89.4 85.9 90.8 a. Determine an estimated regression e
Seven Seas Navigator 89.2 83.3 90.5 score given the score for Shore Excursi
Silver Whisperer 89.2 82 88.6 b. Consider the addition of the indepen
Natinal Geographic Exp 89.1 93.1 89.7 regression equation that can be used t
Silver Cloud 88.7 78.3 91.3 Shore Excursions and Food/Dining.
Celebrity Xpedition 87.2 91.7 73.6 c. Predict the overall score for a cruise
Silver Shadow 87.2 75 89.7 Food/Dining Score of 90.
Silver Wind 86.6 78.1 91.6
SeaDream II 86.2 77.4 90.9
Wind Star 86.1 76.5 91.5
Wind Surf 86.1 72.3 89.3
Wind Spirit 85.2 77.4 91.9
é Nast Traveler Gold List for 2012 provided ratings for the top 20 small cruise
ndé Nast Traveler website, March 1, 2012). The data shown below are the scores
received based upon the results from Condé Nast Traveler’s annual Readers’
rvey. Each score represents the percentage of respondents who rated a ship as
or very good on several criteria, including Shore Excursions and Food/Dining.
l score was also reported and used to rank the ships. The highest ranked ship, the
Odyssey, has an overall score of 94.4, the highest component of which is 97.8

ine an estimated regression equation that can be used to predict the overall
n the score for Shore Excursions.
er the addition of the independent variable Food/Dining. Develop the estimated
n equation that can be used to predict the overall score given the scores for
ursions and Food/Dining.
the overall score for a cruise ship with a Shore Excursions score of 80 and a
ng Score of 90.
 a. The Minitab output follows.

The regression equation is

Overall = 69.3 + 0.235 Shore Excursions

Predictor Coef SE Coef T P

Constant 69.300 4.799 14.44 0.000
Shore Excursions 0.23476 0.05659 4.15 0.001

S = 1.87028 R-Sq = 48.9% R-Sq(adj) = 46.0%

Analysis of Variance

Source DF SS MS F P
Regression 1 60.202 60.202 17.21 0.001
Residual Error 18 62.963 3.498
Total 19 123.166

b. The Minitab output follows.

The regression equation is

Overall = 45.2 + 0.253 Shore Excursions + 0.248 Food/Dining

Predictor Coef SE Coef T P

Constant 45.178 6.952 6.50 0.000
Shore Excursions 0.25289 0.04189 6.04 0.000
Food/Dining 0.24819 0.06161 4.03 0.001

S = 1.37650 R-Sq = 73.8% R-Sq(adj) = 70.8%

Analysis of Variance

Source DF SS MS F P
Regression 2 90.955 45.477 24.00 0.000
Residual Error 17 32.211 1.895
Total 19 123.166

c. yˆ  45.2  .253(Shore Excursions)  .248(Food/Dining)  45.2  .253(80)  .248(90) = 87.7

Thus, an estimate of the overall score is approximately 88.

45.2  .253(80)  .248(90) = 87.76
In exercise PR2, the owner of Showtime Movie Theaters, Inc., used multiple regression analysis
to predict gross revenue ( y) as a function of television advertising (x1) and newspaper
advertising (x2). The estimated regression equation was
e regression analysis
newspaper
 SSR 23.435
a. R2    .919
SST 25.5

n 1 8 1
Ra2  1  (1  R 2 )  1  (1  .919)  .887
n  p 1 8  2 1

b. Multiple regression analysis is preferred since both R2 and Ra2 show a

the variability of y explained when both independent variables are u
1
 .887
2 1

ce both R2 and Ra2 show an increased percentage of

ependent variables are used.
In exercise PR3, data were given on the average number of passing yards per attempt (Yds/Att),
the number of interceptions thrown per attempt (Int/Att), and the percentage of games won
(Win%) for a random sample of 16 National Football League (NFL) teams for the 2011 season
(NFL website, February 12, 2012).
a. Did the estimated regression equation that uses only the average number of passing
yards per attempt as the independent variable to predict the percentage of games won
provide a good fit?
b. Discuss the benefit of using both the average number of passing yards per attempt
and the number of interceptions thrown per attempt to predict the percentage of
games won.
 er attempt (Yds/Att),
age of games won
or the 2011 season

er of passing
f games won
a. r 2 = .577. Thus, the averages number of passing yards per attempt
variability in the percentage of games won. Considering the na
factors that might be related to the number of games won, this

b. The value of the coefficient of determination increased to R2 =

of determination is Ra2 = .714. Thus, using both independent v
fit.
 sing yards per attempt is able to explain 57.7% of the
on. Considering the nature of the data and all the other
ber of games won, this is not too bad a fit.

ation increased to R2 = .752, and the adjusted coefficient

ing both independent variables provides a much better
Player Team W L ERA SO/IP HR/IP R/IP
Verlander, DET 24 5 2.4 1 0.1 0.29
Beckett, J BOS 13 7 2.89 0.91 0.11 0.34
Wilson, C TEX 16 7 2.94 0.92 0.07 0.4
Sabathia, CNYY 19 8 3 0.97 0.07 0.37
Haren, D LAA 16 10 3.17 0.81 0.08 0.38
McCarthy, OAK 9 9 3.32 0.72 0.06 0.43
Santana, E LAA 11 12 3.38 0.78 0.11 0.42
Lester, J BOS 15 9 3.47 0.95 0.1 0.4
Hernandez,SEA 14 14 3.47 0.95 0.08 0.42
Buehrle, MCWS 13 9 3.59 0.53 0.1 0.45
Pineda, M SEA 9 10 3.74 1.01 0.11 0.44
Colon, B NYY 8 10 4 0.82 0.13 0.52
Tomlin, J CLE 12 7 4.25 0.54 0.15 0.48
Pavano, C MIN 9 13 4.3 0.46 0.1 0.55
Danks, J CWS 8 12 4.33 0.79 0.11 0.52
Guthrie, J BAL 9 17 4.33 0.63 0.13 0.54
Lewis, C TEX 14 10 4.4 0.84 0.17 0.51
Scherzer, DET 15 9 4.43 0.89 0.15 0.52
Davis, W TB 11 10 4.45 0.57 0.13 0.52
Porcello, RDET 14 9 4.75 0.57 0.1 0.57
Major League Baseball (MLB) consists of teams that play in the American League and the
National League. MLB collects a wide variety of team and player statistics. Some of the
statistics often used to evaluate pitching performance are as follows:
ERA: The average number of earned runs given up by the pitcher per nine innings.
An earned run is any run that the opponent scores off a particular pitcher except for
runs scored as a result of errors.
SO/IP: The average number of strikeouts per inning pitched.
HR/IP: The average number of home runs per inning pitched.
R/IP: The number of runs given up per inning pitched.
The following data show values for these statistics for a random sample of 20 pitchers from
the American League for the 2011 season (MLB website, March 1, 2012).

a. Develop an estimated regression equation that can be used to predict the average
number of runs given up per inning given the average number of strikeouts per inning
pitched.
b. Develop an estimated regression equation that can be used to predict the average number
of runs given up per inning given the average number of home runs per inning
pitched.
c. Develop an estimated regression equation that can be used to predict the average number
of runs given up per inning given the average number of strikeouts per inning
pitched and the average number of home runs per inning pitched.
d. A. J. Burnett, a pitcher for the New York Yankees, had an average number of strikeouts
per inning pitched of .91 and an average number of home runs per inning of .16.
Use the estimated regression equation developed in part (c) to predict the average
number of runs given up per inning for A. J. Burnett. (Note: The actual value for
R/IP was .6.)
e. Suppose a suggestion was made to also use the earned run average as another independent
variable in part (c). What do you think of this suggestion?
a. The Minitab output follows.

The regression equation is

R/IP = 0.676 - 0.284 SO/IP

Predictor Coef SE Coef T P

Constant 0.67575 0.06307 10.71 0.000
SO/IP -0.28385 0.07869 -3.61 0.002

S = 0.0602733 R-Sq = 42.0% R-Sq(adj) = 38.7%

Analysis of Variance

Source DF SS MS F P
Regression 1 0.047263 0.047263 13.01 0.002
Residual Error 18 0.065392 0.003633
Total 19 0.112655

b. The Minitab output follows.

The regression equation is

R/IP = 0.308 + 1.35 HR/IP

Predictor Coef SE Coef T P

Constant 0.30805 0.06036 5.10 0.000
HR/IP 1.3467 0.5407 2.49 0.023

S = 0.0682239 R-Sq = 25.6% R-Sq(adj) = 21.5%

Analysis of Variance

Source DF SS MS F P
Regression 1 0.028874 0.028874 6.20 0.023
Residual Error 18 0.083781 0.004655
Total 19 0.112655

Unusual Observations

Obs HR/IP R/IP Fit SE Fit Residual St Resid

1 0.100 0.2900 0.4427 0.0159 -0.1527 -2.30R

R denotes an observation with a large standardized residual.

c. The Minitab output follows.

The regression equation is

R/IP = 0.537 - 0.248 SO/IP + 1.03 HR/IP

Predictor Coef SE Coef T P

Constant 0.53651 0.08141 6.59 0.000
SO/IP -0.24835 0.07181 -3.46 0.003
HR/IP 1.0319 0.4359 2.37 0.030

S = 0.0537850 R-Sq = 56.3% R-Sq(adj) = 51.2%

Analysis of Variance

Source DF SS MS F P
Regression 2 0.063477 0.031738 10.97 0.001
Residual Error 17 0.049178 0.002893
Total 19 0.112655

d. Using the estimated regression equation in part (c) we obtain

 Analysis of Variance

Source DF SS MS F P
Regression 2 0.063477 0.031738 10.97 0.001
Residual Error 17 0.049178 0.002893
Total 19 0.112655

d. Using the estimated regression equation in part (c) we obtain

 P
02

P
23

esid
2.30R

residual.

P
001
 P
001
Refer to exercise PR7, where Major League Baseball (MLB) pitching statistics were
reported for a random sample of 20 pitchers from the American League for the 2011
season (MLB website, March 1, 2012).
a. In part (c) of exercise 10, an estimated regression equation was developed relating the
average number of runs given up per inning pitched given the average number of
strikeouts per inning pitched and the average number of home runs per inning pitched.
What are the values of
b. Does the estimated regression equation provide a good fit to the data? Explain.
c. Suppose the earned run average (ERA) is used as the dependent variable in part (c)
instead of the average number of runs given up per inning pitched. Does the
estimated regression equation that uses the ERA provide a good fit to the data?
Explain.
ed relating the

ning pitched.
a. The Minitab output in exercise 10 shows that R 2 = .563 and Ra2 = .512.

b. The fit is not great, but considering the nature of the data being able to explain slight
than 50% of the variability in the number of runs given up per inning pitched using j
independent variables is not too bad.

c. The Minitab output using ERA as the dependent variable follows.

The regression equation is

ERA = 3.88 + 12.0 HR/IP - 1.84 SO/IP

Predictor Coef SE Coef T P

Constant 3.8781 0.6466 6.00 0.000
HR/IP 11.993 3.462 3.46 0.003
SO/IP -1.8428 0.5703 -3.23 0.005

S = 0.427204 R-Sq = 62.5% R-Sq(adj) = 58.1%

Analysis of Variance

Source DF SS MS F P
Regression 2 5.1739 2.5870 14.17 0.000
Residual Error 17 3.1025 0.1825
Total 19 8.2765

The Minitab output shows that R 2 = .625 and Ra2 = .581

Approximately 60% of the variability in the ERA can be explained by the linear effe
and SO/IP. This is not too bad considering the complexity of predicting pithing perf
Ra2 = .512.

a being able to explain slightly more

up per inning pitched using just two

e follows.

P
0
3
5

58.1%

P
0.000

e explained by the linear effect of HR/IP

ty of predicting pithing performance.