The influence of football head coaching change in the Football Championship Subdivision: an evaluation of the NCAA Academic Progress Rate.
Johnson, James E. ; Pierce, David A. ; Tracy, Daniel R. 等
College football has become one of the largest and most celebrated
forms of sport (Smith, 2011). Participation at the highest level of
collegiate football includes institutions that comprise the National
Collegiate Athletic Association (NCAA) Football Bowl Subdivision (FBS).
These institutions are given the most media attention as they compete
for the national championship through a variety of high-profile bowl
games that are accompanied by large financial payouts. For these
reasons, FBS football programs receive a tremendous amount of attention
both from a fan and research perspective (Rogers, 2008; Sander, 2011).
At the level below the FBS is the category of institutions
belonging to the Football Championship Subdivision (FCS). The FCS was
formerly known as Division I-AA, and classified as a lower tier because
institutions in this subdivision are less equipped to participate at the
FBS level. For example, in 2011 there were 22 FBS programs with an
average profit of $7.4 million. In contrast, only 18 total FCS athletic
programs have seen a financial surplus in the six years from 2005-2011.
This widening gap of revenue can largely be credited to revenues, media
rights, and conference distributions (NCAA, 2011).
A microcosm of the disparity between FCS and FBS are head coaching
salaries. It is widely understood that FCS head coaches receive fewer
benefits and less compensation than their FBS counterparts. Most recent
data reveals the average coach salary at the FBS level ($1.64 million)
is more than ten times that of FCS coaches ($95,000). Coaches in the
Southeastern Conference alone make an average of $2.54 million (USA
Today, 2013). Despite these differences, FCS football programs are
expected to compete against FBS programs on a regular basis. This can
place a tremendous amount of pressure on FCS head football coaches due
to the ever-present emphasis on winning (Sperber, 2001). Furthermore,
the lack of financial resources means that FCS institutions have less
personnel and infrastructure to aid in the athletic and academic
endeavors of its student-athletes. For example, it is not uncommon for
FBS institutions to have large academic support centers and teams of
advisors to aid student-athletes with course selection, study skills,
tutoring, and learning disabilities. If FCS coaches have a version of
these services, it is often on a much smaller scale (Judge, Petersen,
& Johnson, 2013).
Identifying the limitations faced by FCS coaches is important
because it implies that FCS coaches are more likely involved in the
day-to-day athletic and academic pursuits of their student-athletes.
This point is especially salient when examining the influence of a coach
through a theoretical lens. Coaches are a superlative example of what a
leader is, especially from an influential and behavioral perspective. In
these regards, virtually every organizational theory emphasizes the
importance of a leader and their instrumental position to influence
change (Chelladuri, 2009; Gilmore, 2003; Herold, Fedor, Caldwell, &
Liu, 2008). With this premise in mind, there are two specific theories
which are particularly relevant to explain why a leadership (i.e.,
coaching) change occurs. First is common sense theory (Grusky, 1963),
which posits that team performance is largely influenced by the actions
of the coach. This theory implies that when a team is performing poorly,
a change in coach will likely improve performance. Therefore, the
relationship between coaching change and team performance has a negative
relationship (i.e., as team performance goes down, the likelihood of a
coaching change goes up). The second theory, known as ritual
scapegoating (Gamson & Scotch, 1964), suggests that coaching changes
are necessary after poor performance to alleviate anxiety of various
stakeholders. Unlike common sense theory, however, ritual scapegoating
implies that "the actual impact of the manager on team performance
is minimal and therefore has little consequence" (Fabianic, 1994,
p. 136). Thus, coaches are used as scapegoats for poor performance.
Although common sense and ritual scapegoating theory help to
explain why a coaching change might occur, few theories can explain how
such a change might impact student-athletes. One specific theory,
Complex Adaptive System theory (CAS; Eidelson, 1997), indicates that the
leader is one vital piece of an intricate system where a disruption of
one part of the system would cause a ripple-effect through the whole
system. CAS theory is particularly relevant in this regard because it
notes the strong relationships manifested between leaders and followers.
Based on this theory, the more involved a coach is with their
student-athletes, the more influence that coach has to impact the
overall system. Furthermore, CAS theory supports that the impact of a
change would be influenced by the magnitude of the change (e.g., a
relatively seamless positive coaching change based on coaching success
vs. a negative change based on years of underperformance), as well as
the ability of the new leader to minimize the disruption in the system
(e.g., hiring a coach from the previous staff who continues the same
procedures, or replacing the entire staff and implementing new
procedures). Given this line of thinking, it is reasonable to conclude
that if a coaching change does occur, it is likely to impact a variety
of areas within the system, including both athletic and academic
pursuits. For FCS football student-athletes, such an academic impact
would likely be observed in the Academic Progress Rate (APR).
The APR is the most contemporary academic metric used by the NCAA
and was implemented in 2004 as a way to gauge the semester-by-semester
academic performance of individual teams. The factors used to compute
APR scores are eligibility and retention. All student-athletes who
receive athletics-based financial aid are evaluated at the beginning of
each semester to determine if they are eligible, and if they returned to
the institution to pursue their education. In each semester,
student-athletes earn one point for eligibility and one for being
retained. Therefore, in an academic year students can earn a total of 4
points. To determine a team APR score, points earned are divided by
points possible and multiplied
by 1000. This calculation will result in a number between 0 and
1000, which is the APR score (Brown, 2005). Because APR scores are team
scores and calculated using dichotomous metrics for individuals (i.e.,
retained or not and eligible or not), it is a less sensitive measure
than individual GPA. For example an athlete may be eligible and retained
in a given semester, but their GPA might have dropped significantly. In
this example the team APR score would not suffer, but the individual GPA
could.
The penalty structure for low APR scores has the potential to be
severe. The current cutoff score for penalties is 925, but will be
changed to 930 in 2015-16 (Hosick, 2011). If a team falls below 925 in a
single year, and has one or more students ineligible and not retained
(i.e., 0 for 2), an immediate penalty is triggered which can result in
up to a 10 percent loss of scholarships. Harsher penalties occur if
teams do not meet a score of 900 (NCAA, 2010).
Analysis of Division I football APR scores is not flattering.
Similar to the men's sports of basketball and baseball, Division I
football has the lowest APR scores of any NCAA sport at 948. When
Division I is divided into FBS and FCS, the scores are 952 and 944
respectively (NCAA, 2012a). Thus, FCS football has the lowest APR scores
of any NCAA sport. To date, there is little empirical evidence to
explain what influences APR scores. Most of the previous research
investigating academic performance has focused on GPA or graduation
rates. Those studies, which often focused on the largest and most
successful football and men's basketball programs, reveal the most
athletically successful teams produce some of the lowest GPAs and
graduation rates (Amato, Gandar, Tucker, & Zuber, 1996;
Christianson, 2004; Hosick, 2009; Institute for Diversity and Ethics in
Sport, 2012a, 2012b; Shapiro, 1984). Similar research on the
relationship between academic achievement and winning has not been
replicated using APR in place of GPA and graduation rates.
Within the small body of APR literature, one study indicated that
coaching change was a relatively powerful predictor of semester APR
scores (Johnson, Wessel, & Pierce, 2012). However, the Johnson et
al. study examined 652 Division I student-athletes using a variety of
variables without consideration given to a specific sport (e.g.,
football). As a follow-up to the Johnson et al. (2012) study, Johnson
(2012) investigated the general relationship among a head football
coaching change and APR scores at a single FBS institution between 2005
and 2010. The findings indicated coaching change was the only
significant predictor of APR among other factors that included distance
from home, major, race, and playing time. Because these results were
limited to one institution, further investigation of coaching changes at
multiple institutions and NCAA divisions, while simultaneously examining
the characteristics of such a change, was encouraged (Johnson, 2012).
Subsequently, the most recent study by Johnson, Blom, Judge, Lee,
Pierce, and Ridley (2013) expanded on the Johnson (2012) study by
investigating if all FBS head football coaching changes (N = 160)
between 2003-04 and 2010-11 had an impact on APR scores. As recommended
in the Johnson (2012) study, Johnson et al., (2013) also examined
characteristics of the change that may have been significantly impactful
to team APR scores. Specifically, Johnson et al. (2013) reviewed the
nature of the change (positive vs. negative), type of hire (internal vs.
external), and APR's relationship to a team's winning
percentage. The significant results from the Johnson et al. (2013) study
supported that APR scores are significantly lower in the year of a head
coaching change, internally hired coaches produce higher APR scores than
externally hired coaches, and teams with the highest winning percentages
produced the greatest APR scores. These results supported the importance
of a head coach at the FBS level, but were limited only to the highest
level of college football (i.e., FBS). Johnson et al. (2013) proposed
that high levels of resources and the commercialized nature of FBS
football may render these results unique. Thus, it was recommended the
research be replicated at varying levels of collegiate football.
Therefore, the present study will attempt to determine if the impact of
a coaching change, as well as the variables surrounding such a change,
apply to FCS football head coaching changes as they did to FBS coaching
changes. Given the aforementioned differences between FBS and FCS
football, the fact that FCS football comprises the lowest APR scores of
any NCAA sport, and the limited research conducted on FCS football
programs in general, this research is prudent.
Research Question/Hypotheses
The current study attempted to systematically replicate (Sidman,
1960) the Johnson et al. (2013) FBS study by investigating all FCS head
coaching changes between the 2003-04 and 2010-11 academic years. The
following research questions and three corresponding hypotheses were
formulated as the basis for this study and adapted from the Johnson et
al. (2013) study:
RQ 1: Does a FCS football head coaching change impact APR scores?
H1: The APR score in the year of a head coaching change will be
significantly lower than the average APR score.
H2: Teams with a positive coaching change will demonstrate
significantly higher APR scores than a negative coaching change.
H3: A coaching change that results in an internal hire will have
APR scores significantly higher than a coaching change that results in
an external hire.
RQ 2: Does a FCS football team's athletic success relate to
APR scores?
H4: Teams with the highest winning percentages (year of coaching
change and overall) will produce significantly lower APR scores than
teams with the lowest winning percentages.
RQ 3: What factors predict APR scores for a FCS football team in
the year of a coaching change?
H5: All variables (i.e., APR average, type of hire, nature of
change, year of change, winning % in year of change, and winning %
average) under investigation will be significant predictors of APR
scores in the year of a coaching change.
Method
The current study sought to investigate all FCS football head
coaching changes (N = 120) and the corresponding APR scores during the
academic years of 2003-04 to 2010-11 (8 years) amidst a variety of
potential intervening variables (e.g., type change, type of hire,
winning percentage) to determine if a football head coaching change is
significantly related team APR scores. The academic year of 2003-04 was
selected as the starting point of this study because APR data was first
collected during this academic year.
Operational Definitions
The following are operational definitions used in the current study
adapted from Johnson et al. (2013).
1) APR (year of coaching change) = the single-year APR score earned
during the academic year in which a head coaching change occurred.
2) APR (average) = the average APR score for the eight academic
years under investigation (2003-04-2010-11) minus the APR score for the
year in which a head coaching change occurred.
3) Internal/External = identifies from where a new coach was hired
after the head coaching change. Internal hires were coaches already on
the coaching staff that were promoted to head coach (not interim
status). External hires were from outside of the exiting coach's
staff.
4) Positive/Negative = identifies the circumstances by which the
coaching change occurred. Positive coaching changes occurred as a result
of successful coaching tenures (e.g., leaving for a more prominent
coaching position after success, retired voluntarily with a history of
success, or was promoted to athletic director as a result of past
accomplishments). Negative coaching changes occurred as a result of
unsuccessful coaching tenures (e.g., being fired, resigning after a lack
of success, death, scandal, or other negative circumstances where
resignation or termination occurred).
5) Year of Change = the academic year in which the head coaching
change occurred (July 1-June 30).
6) Month of Change = the month in which the head coaching change
occurred.
7) Win % (year of coaching change) = the total number of wins
divided by the total games played during the academic year in which the
head coaching change occurred.
8) Win % (average) = the total number of wins divided by the total
games played during the academic years under investigation
(2003-04-2010-11).
Procedures
All data utilized in the current study was archival. The NCAA
Division I Head Coach APR Portfolio (NCAA, 2013) was used to log all FCS
head football coaching changes (N = 120) between the 2003-04 and
2010-2011 academic years. This publicly accessible online portfolio
contains APR scores and employment ranges of all head coaches. Each
coaching change, even if multiple coaching changes occurred at the same
institution, was considered a unique occurrence. Information regarding
internal/external hires and timing of change was gathered from
university websites. Information determining positive or negative
coaching changes, as well as winning percentage, was mined from
institutional and media sites documenting the coaching change. This
information was treated as manifest content due to the public nature of
coaching changes.
For data analysis, descriptive statistics were analyzed using
frequency totals and measures of central tendency. Pearson correlations
were conducted to determine relationships between continuous variables
(i.e., APR scores, year of change, and winning percentage), while
point-biserial correlations were conducted for relationships between
categorical (i.e., nature of change and type of hire) and continuous
variables. After descriptive and correlational analysis, hypothesis one
was tested using a paired samples t-test. Hypotheses two and three were
tested using independent t-tests. Hypothesis four was evaluated using
two one-way analyses of variances (ANOVAs) to compare the top, middle,
and bottom groups based on single year and average winning percentage.
The final hypothesis was tested by ordinary least squares multiple
regression analysis. Alpha levels were set at .05.
Results
Before addressing the hypotheses, descriptive statistics were
calculated to establish context and patterns. Table 1 demonstrates
descriptive information for APR scores (year of change and average
score), type of hire (internal vs. external), nature of change (positive
vs. negative), and winning percentage (year of change and average). Of
note, there is a lower mean APR score in the year of a coaching change
(M= 922.87) than mean APR scores in general (M= 929.69), an overwhelming
majority of new coaches came from outside of the previous coaching staff
(80.83%), a large number of coaching changes were classified as negative
(84.17%), and overall average winning percentage at the FCS level was
quite low (.36). Table 2 adds an additional layer of descriptive data by
displaying cross tabulations for nature of change (i.e., positive vs.
negative) and type of hire (i.e., internal vs. external). This
additional layer of analysis indicated the majority of coaching changes
(68.3%) were negative changes that lead to an external coaching hire.
Table 3 displays the timing (year and month) of FCS head coach changes
cross tabulated with type of hire (internal vs. external) and nature of
change (positive vs. negative). Although the number of coaching changes
varied from year to year, the mean number of coaching changes per year
was 15. For the month of change, 54.2% of the coaching changes happened
in November (n = 24) and December (n = 41). Table 4 extends the
descriptive analysis by demonstrating correlation coefficients for FCS
head coaching change variables. The strongest correlations for APR in
the year of a coaching change occurred with APR average (r = .65).
Hypothesis one predicted APR scores for the year of the coaching
change would be significantly lower than the average APR scores. Results
from the paired t-test revealed the mean APR scores during the year of a
coaching change (M= 922.87, SD = 40.78) was significantly lower than the
average APR score (M = 929.69, SD = 33.33), t(119) = -2.35, p = .02,
thus supporting hypothesis one. For hypothesis two, an independent
samples t-test was used to determine if a positive coaching change would
demonstrate higher team APR scores than a negative coaching change. The
results supported the hypothesis indicating significantly higher APR
scores when teams had a positive coaching change (M= 941.53, SD = 38.68)
rather than a negative coaching change (M= 919.36, SD = 40.39), f(l 18)
= 2.209, p = .03. The 95% confidence interval for the difference in
means ranged from 2.3 to 42.04. Hypothesis three predicted a coaching
change that results in an internal hire would have APR scores
significantly higher than a coaching change that results in an external
hire. The independent samples t-test revealed no significant differences
between APR scores when the coach was hired internally (M= 917.43, SD =
38.92) versus externally (M= 924.15, SD = 41.3), t(118) = -.71, p = .48.
Hypothesis four predicted that teams with the highest winning
percentage (year of coaching change and overall winning percentage)
would produce lower APR scores (year of coaching change and average APR
scores) than teams with lower winning percentages. The independent
variable of winning percentage was divided into three evenly distributed
groups known as the bottom third, middle third, and upper third. Table 5
demonstrates APR scores (average and year of coaching change) based on
winning percentage for the FCS programs in this study compared to FBS
programs found in the study by Johnson et al. (2013). For each winning
percentage group, APR scores for the FCS are dramatically lower than for
the FBS. This is most evident for APR scores in the year of a coaching
change where FCS programs produced a score of 915.55 and FBS programs
produced a score of 949.19 (Johnson et al., 2013). For FCS programs in
the current study, the first of two one-way ANOVAs (APR in the year of
coaching change) yielded no significant differences based on winning
percentage, F(2, 96) = .25,p = .78. The second one-way ANOVA (APR
average) was significant, F(2, 96) = 4.81, p = .01. Post hoc Tukey HSD
analyses revealed that average APR scores for the upper third winning
percentages (M= 938.22, SD = 25.13) were significantly higher than
average APR scores for teams in the middle third (M = 929.7, SD = 24.9)
and bottom third (M= 914.24, SD = 41.84). There were no significant
differences between the middle and bottom third. Given the only
significant difference for winning percentage occurred between the top
third and bottom two thirds of FCS programs (APR average), and no
winning percentage differences occurred for APR in the year of the
coaching change, hypothesis four was mostly rejected.
For the fifth hypothesis, it was forecasted that APR average, type
of hire (internal vs. external), nature of change (positive or
negative), year of coaching change, and winning percentage (year of
change and overall) would be significant predictors of APR scores in the
year of a coaching change. Results from the least squares multiple
regression analysis can be seen in Table 6. This analysis supported the
linear combination of these predictor variables as significant, F(18,
91) = 6.18, p < .01. The sample multiple correlation coefficient was
.74, indicating approximately 51.5% of the variance in APR scores can be
accounted for by a linear combination of predictors. Average APR score,
however, was the only variable significant enough to aid in the
prediction equation, accounting for 45.2% of the variance alone. These
results partially support hypothesis five.
Discussion
The current study sought to systematically replicate the Johnson et
al. (2013) study on FBS football coaching changes to determine if FCS
head football coaching change and winning percentage impacted APR
scores.
Coaching Change Context
The descriptive statistics were first calculated to determine
patterns and provide context of FCS head coaching changes and winning
percentages. The findings supported several patterns that have largely
been anecdotally understood in high-level collegiate coaching. First,
most new coaches (80.83%) are hired external to the previous coaching
staff. This finding signifies that programs gain a fresh start with a
new coaching staff that brings their own philosophy. This is not
surprising considering that football staffs are routinely hired and
fired as units, and that their employment is often dependent on their
team's athletic accomplishments (Fee, Hadlock, & Pierce, 2006).
This finding is nearly identical to FBS programs (Johnson et al., 2013)
where 80.62% of coaches were hired externally.
The second noteworthy descriptive finding indicates the majority
(84.17%) of coaching changes are negative. This finding was also
expected considering football coaching changes are routinely public
events that coincide with winning percentage. In fact, results from this
study indicated a significant correlation (r = -.49, p < .01) between
winning percentage in the year of a coaching change, and the nature of
the change. Thus, as winning percentage decreases, the likelihood of a
negative coaching change increases. This point is especially relevant
for FCS football programs because during the time frame of this study
FCS programs demonstrated only a .46 winning percentage overall, and
only a .36 winning percentage in the year of a coaching change. These
low winning percentage numbers are likely due in part to games played
against larger FBS programs. The importance, however, lies in supporting
the link between winning and coaching changes in FCS football because it
provides empirical support for what is commonly understood among the
casual fan, which is that football coaches are often fired for losing
games. This point is further solidified when one considers that the
majority of new coaching hires (68.3%) are made externally after a
negative coaching change. Furthermore, this finding also supports common
sense and ritual scapegoating theories suggesting that a coaching change
is necessary because a new coach is likely to improve the situation
(Grusky, 1963), or because stakeholders need to reduce anxiety by
replacing the person of blame (Gamson & Scotch, 1964).
Descriptive statistics for the timing of coaching changes also
warrant discussion. Given that college football starts in August and
ends in late fall (November or December), it is not surprising that most
coaching changes occur in months coinciding with the end of the season
or closely after. In fact, 77% of FCS football head coaching changes
happened in the months of November, December, and January. This finding
was similar to the 82.5% of changes that occurred during the same three
months for FBS programs (Johnson et al., 2013). This outcome supports
that wins and loses often dictate whether a coaching change occurs, and
further support the correlation between winning percentage and the
nature of the change (i.e., positive or negative). In other words, the
results of the football season appear to dictate when a coaching change
occurs, thus reinforcing the emphasis on winning for FCS football
coaches. Furthermore, making a coaching change immediately after the
football season allows athletic administrators time to make an informed
hiring decision in time to prepare for the next season.
Overall Impact of Coaching Change
Beyond the descriptive information provided by this study, three
research questions and five corresponding hypotheses were addressed. The
first three hypotheses were created to determine if FCS football head
coaching change impacted APR scores. Hypothesis one was supported,
indicating that APR scores are significantly lower in the year of a
coaching change than during years when a coaching change did not occur.
This result mirrors findings from the Johnson et al. (2013) study
indicating both FBS and FCS football programs see lower APR scores in
years where a coaching change occurs. This finding also supports the
importance of a coach within the framework of transformational theory
(Herold et al., 2008) and CAS theory (Eidelson, 1997), demonstrating
that when a change in leadership occurs, it has an impact on the other
components in the system. The support for CAS theory is further
evidenced by the nonlinear nature of relationships in the theory, thus
indicating that when a change in leadership occurs it does not have a
direct linear impact on the subordinates, but rather the ripple effect
that influences key parts of their lives (e.g., athletic or academic
policies). This is best exhibited by correlations that are significant,
but not particularly strong, as seen in this study.
In a practical sense, results from hypothesis one can further be
explained by examining the importance of a head coach relative to their
student-athletes. The influence of a head coach, beginning in the
recruiting phase and progressing to the daily athletic and academic
policies, can be intense (Brubaker, 2007; Giacobbi, Roper, Whitney,
& Butryn, 2002; Leslie-Toogood & Gill, 2008). For example, 41%
of Division I college football players indicated their coach was the
primary reason for selecting their institution (NCAA, 2011). This
powerful statistic implies that many student-athletes have a strong
affinity for their coach before arriving on their respective campuses.
Once on campus, coaches fill a variety of leadership roles. Coaches have
been ascribed characteristics associated with teachers, guardians,
business mentors, counselors, disciplinarians, injury evaluators, and
emotional caretakers (Amorose, 2003; Bradley, 2005; Brubaker, 2007;
Gagne, Ryan, & Bargmann, 2003; Lattman, 2008; Lewis, 2004;
Schilling, 2007). These powerful leadership characteristics demonstrate
why the Division I APR portfolio was created, and permanently links a
coach with their APR history. This sentiment is echoed in the stance
held by the chair for the NCAA Committee on Academic Performance, Walter
Harrison, who concluded that coaches "not only recruit
student-athletes to their institutions but also have the closest
relationship with individual student-athletes of any other adult at a
college or university" (Hosick, 2010, [paragraph]8).
The result from hypothesis one is also important to understand in
the context of APR scores. The difference between mean APR scores in the
year of a coaching change (922.87) and the mean APR scores in years
without a coaching change (929.69) is notable. Although a difference of
6.82 APR points may not appear large on a scale of 1000, it is important
to remember the average APR scores for the lowest men's sport in
the NCAA (944-FCS football) is only 39 points below the highest
men's sport of gymnastics at 983, and only 46 points below the
highest women's sport of field hockey at 990 (NCAA, 2012a). Thus,
6.82 APR points accounts for 17.4% and 14.8% of the spread between all
men's teams and women's teams, respectively. In practical
terms, this means that 6.82 points can make a dramatic difference for
teams. This is especially true for FCS football because they have the
lowest average team APR scores, which positions them closest to the
penalty cutoff of 925. Moreover, the new penalty structure implemented
in 2015-16 will increase the penalty cutoff to 930 (Hosick, 2011), which
is .31 of a point above the average APR score in years where there is no
coaching change, and 7.13 points above the average APR in years where
there is a coaching change. Therefore, if the status quo continues in
FCS football, the average FCS team would be under the 930 threshold.
Teams that experience a coaching change would be well under the 930
mark, and thus would be at an elevated level of risk for accruing
penalties. Pragmatically, athletic administrators can use the results
from hypothesis one, as well as FBS results from Johnson et al., (2013)
to provide programming efforts in times of coaching transition at both
FCS and FBS institutions. Decisions about academic support, interim
leadership, and personnel changes can be made with these results in
mind.
Impact of Positive and Negative Coaching Change
Hypothesis two predicted that a team with a positive coaching
change would have significantly higher APR scores than a team who
experienced a negative coaching change. At the FBS level, Johnson et al.
(2013) found that no APR differences existed based on a positive or
negative change. At the FCS level, however, there were significant
differences in APR scores. Programs experiencing a negative coaching
change produced significantly lower scores than programs with a positive
change. Comparing the results between FBS and FCS programs allows one to
draws some potential conclusions.
First, CAS theory implies that within a complex system, roles and
interactions are nonlinear and unique to each system or context. A
leadership role, while often powerful and likely to influence a great
deal of the other components, can vary in the potential impact it has
for any particular system. Given the identified differences in
resources, roles, and challenges faced by FCS coaches relative to their
FBS counterparts, it is logical to conclude FCS coaches may play a more
salient role in the system, and subsequently impact student-athletes to
a greater degree. In other words, because FCS coaches do not have the
amount of resources, media engagement, and external pressures of FBS
coaches, they are forced to be more engaged and prominent in the
day-to-day endeavors of student-athletes. Hence, the amount of
interaction between FCS coaches and their student-athletes may be
greater than that of FBS coaches, causing a more sensitive reaction to a
coaching change. With the amount of academic support resources available
at the FBS level (Judge et al., 2013), as well as the other resources in
place that might serve to neutralize the nature of the change, this is a
plausible explanation.
Second, there are inherent differences in the student-athlete
experience between the FBS and FCS levels. Beyond the facility and
personnel resources, the number and nature of scholarships are
different. At the FBS level, there are 85 full scholarships available.
At the FCS level, however, there are only 63 total scholarships, and
those may be distributed in whole or part to no more than 85 players
(NCAA, 2012b). Therefore, in FCS football fewer students receive full
scholarships, and some receive only partial scholarships. Additionally,
FBS football tends to have more prestige associated with being a member,
and in general has a higher caliber of athlete. Thus FBS
student-athletes may produce a higher level of athletic identity,
especially in the highly commercialized sport of football (Steinfeldt
& Steinfeldt, 2012). Consequently, when a negative coaching change
happens, it is reasonable to conclude that FCS student-athletes may be
able to part ways from their program somewhat easier than FBS
student-athletes who may feel pressure to remain on their team for
scholarship money or prestige. If FCS football student-athletes can
disengage from their respective programs easier than FBS
student-athletes, the explanation for differencing APR scores may be
attributed to differences in retention points.
Impact of Internal and External Hires
The third hypothesis predicted a coaching change that results in an
internal hire will have APR scores significantly higher than a coaching
change that results in an external hire. Unlike FBS programs where an
internal hire was significant (Johnson et al., 2013), hiring internally
or externally at the FCS level did not produce different APR scores. For
FCS programs, this finding appears to contradict Roach and Dixon's
(2006) contention that trust and stability are established when making
an internal hire. This finding also appears to be counterintuitive to
CAS theory which would predict that the lower the amount of disruption
in a system, the more likely the status quo would be maintained.
Therefore, if a version of the status quo is maintained (i.e., internal
coach is hired), a dip in the APR scores as a result of an external head
coaching change could be avoided according to CAS theory.
It is difficult to determine why an internal head coaching change
would be academically beneficial for FBS programs and not for FCS
programs. At least part of the reason may lie in the relatively high
amount of negative coaching change that occurs at the FCS level (84.17%)
relative to the FBS level (68.8%; Johnson et al., 2013). Evidence from
this study demonstrated a significant correlation between winning
percentage (both year of a coaching change and average) and nature of
change, suggesting that coaches are often fired as a result of their
team's poor athletic performance. This finding again supports both
common sense and ritual scapegoating theories, as well as evidence by
Fee et al., (2006) who suggested that entire coaching staffs are fired
after poor athletic performance. Therefore, if 84% of the time the
previous coach was released due to poor performance, the team morale was
likely in an already negative state. This logic is somewhat supported by
the very low APR scores associated with internal hires (917.43) vs.
external hires (924.15). So, if FCS programs already have a negative
team atmosphere due to the lack of athletic success, it appears that APR
scores could be low no matter who is hired. This explanation was not
supported by this study but warrants future consideration. The practical
importance of testing hypothesis three, however, lies in the fact that
internally and externally hired coaches at the FCS level are likely
stepping into a scenario where winning is generally difficult to
accomplish and APR scores are dangerously low.
Relationship between Winning and APR Scores
The second research question attempted to determine how winning was
related to APR scores. Given previous research that found the most
successful athletic teams produce some of the lowest GPAs and graduation
rates (Amato et al., 1996; Christianson, 2004; Hosick, 2009; Institute
for Diversity and Ethics in Sport, 2012a, 2012b; Shapiro, 1984) it was
predicted that the winningest FCS football teams would produce the
lowest APR scores. However, analogous to FBS programs (Johnson et al.,
2013), the winningest FCS programs demonstrated the highest average APR
scores. Unlike FBS programs, however, there was no significant
difference in APR scores during the years of a coaching change based on
winning percentage (See Table 5). These findings for FCS programs
suggest that on average more successful teams generate higher APR
scores, except in the years of a coaching change. From a practical
perspective, these findings are particularly dire for programs with the
bottom third winning percentages where the average APR scores were
914.24, and a dreadful 907.26 in the year of a coaching change. New
coaches, particularly those hired externally into a program with a
history of losing, are in an especially difficult situation due to the
extremely low APR scores. Additionally, it is important to remember that
APR scores are calculated using eligibility and retention, which is an
overall less sensitive metric than GPA. Therefore, it may be that APR
were not impacted by winning percentage because students were eligible
and retained, but their overall academic performance might have suffered
(i.e., GPA might have decreased).
One explanation for the contrast between significantly higher APR
scores on average compared to no significant APR differences in the year
of a coaching change is that coaching change neutralizes the impact of
winning. If a coaching change produces significantly lower APR scores,
which is the case based on results from hypothesis one, then it is
possible that the most successful programs are impacted more by the
coaching change due to leader characteristics that positively influence
both athletic and academic achievements simultaneously. In other words,
if a coach is athletically successful (i.e., high winning percentage),
then they are likely influential in other areas of the program due to
the same personality characteristics that make them a successful
football coach (i.e., influence on social policy or academic
achievement). Thus, when a coaching change occurs after winning seasons,
the impact may be felt more strongly because an influential leader has
departed. This logic is supported by common sense theory, as well as the
large difference between the average APR scores (938.22) and APR scores
in the year of a coaching change (914.55) for the winningest programs
(see Table 5). Although the current study cannot determine if coaching
personality is the reason there are no significant differences in APR
scores during the year of a coaching change, it is clear that in years
when a coaching change does exist, winning does not appear to impact APR
scores. Thus, coaching change appears to neutralize the impact of
winning.
Another potential explanation for the findings from hypothesis
four, which was also noted for FBS programs (Johnson et al., 2013), was
the sample included only programs that have experienced a coaching
change during the eight years under investigation. A total of 36 FCS
teams did not have a coaching change, and thus were not incorporated in
this study. These teams included traditionally successful programs with
an entrenched coach and an established culture. Therefore, the sample in
the current study differs from samples in much of the previous
literature relating athletic success and academic achievement because it
did not include many of the institutions that were traditionally
successful in FCS football. This point is reinforced when one considers
that 84.17% of the coaching changes were negative. Thus, the winningest
teams in this study made coaching changes (likely negative changes)
during the eight-year time frame of this study. If the study had
included all FCS institutions without taking coaching changes into
account, results might have resembled prior studies to a greater degree.
Predicting APR Scores
Research question three, and the corresponding fifth hypothesis,
was analyzed using least squares linier regression to determine which
variables best predicted APR scores in the year of a coaching change.
Despite significant correlations with four of the six variables in this
study (i.e., APR average, nature of change, winning percentage in the
year of a coaching change, and average winning percentage), only average
APR was influential enough to aid in predicting APR in the year of a
coaching change. This finding is important because it suggests that
although four of the six variables under investigation were
significantly related to APR scores in the year of a coaching change,
three of those variables were not powerful enough to help make a
prediction. When one examines the correlation coefficients, it becomes
clear the reason for this finding is due to the relatively weak r values
for nature of change, average winning percentage, and winning percentage
in the year of a coaching change, thus suggesting weak to moderate
relationships. Moreover, the r value of .65 for average APR was strong,
thus explaining how this variable eclipsed the remaining three
correlated variables, and was the only variable that was a significant
predictor of APR in the year of a coaching change.
The importance of average APR on the influence of APR in the year
of a coaching change can also be quantified using the B values from the
regression equation. A B value of .51 for average APR reveals that APR
in the year of a coaching change increases .5 for each point increase in
the team's average APR score, assuming all other variables were
held constant. Perhaps more importantly, one could expect APR scores in
the year of a coaching change to be roughly 50% of what the APR scores
would be on average. Although this number appears to predict a very low
APR score in the year of a coaching change, it is important to remember
these numbers assume all other variables would be held constant. In the
real world, there are a multitude of variables that would likely
increase APR scores (e.g., student motivation, academic support, social
influences). This finding supports CAS theory by suggesting there are a
variety of influences in the complex system of FCS college football that
have direct or indirect influences on APR scores. Because many of these
extraneous variables exist no matter if a coaching change happens or
not, it provides contextual support for the reason why average APR is
significantly higher than APR scores in the year of a coaching change,
and why the only variable to predict APR in the year of a coaching
change is average APR scores. In other words, a coaching change does
produce significantly lower APR scores, but those scores will be similar
enough to the average scores to reasonably predict them if a coaching
change occurs.
These findings for the final research question resembles results
for FBS football programs (Johnson et al., 2013) in that average APR
scores was a significant predictor of APR scores in the year of a
coaching change. However, year of change was also a significant
predictor for FBS programs suggesting the farther from the initial
implementation of APR scores (2003-2004), the higher APR scores would
be. Johnson et al. explained this trend by noting that FBS programs have
likely adapted to the APR by committing more resources to
student-athlete support services and advising. This is certainly
probable considering athletic programs at FBS institutions have been
found to have three times the full-time academic support staff, three
times the tutors, eight times the tutor budgets, and nearly four times
the overall dedicated academic space as non-FBS institutions (Judge et
al., 2013). Additionally, when one considers the low APR scores found in
the current study, as well as the low APR scores for FCS football in
general (NCAA, 2011, 2012a), it is easy to discern there is a large
discrepancy between FBS and FCS programs. Perhaps more pragmatically,
the fact that FBS programs have seemed to find ways to improve their APR
scores while FCS programs have not suggests a continually growing divide
between the haves and the have-nots of Division I football.
Limitations/Suggestions
The current study investigated only FCS football programs during an
eight-year time frame. Although the current study can draw many
parallels with the results from FBS football programs examined by
Johnson et al. (2013), FCS football programs that have experienced a
coaching change is a specific population. Division I football, even at
the FCS level, is considered a high level of athletic competition, and
thus offers a unique set of characteristics. Generalizing these findings
to other sports, NCAA divisions, or coaching patterns is recommended.
With this limitation in mind, this line of research could be expanded to
investigate other sport contexts to determine how they might differ from
FBS and FCS football. The sport of basketball, for example, has more
coaches per player than football which could have an impact on how
leadership is conceptualized.
An additional limitation of the current study is the small number
of variables used to investigate the circumstances of a coaching change.
Capturing all factors that influence a coaching change, or APR scores,
for any given member of a football team is not possible. Although this
study does identify general patterns for FCS football programs using
specific coaching change variables, there are other factors that likely
contribute to the findings. For example, Johnson et al. (2013) noted
that subjective views of athletic success, specific academic resources,
or identifying GPA as a dependent variable could yield additional
information. This type of future research could add to a more
comprehensive understanding of coaching changes at the college level,
and could serve to ignite leadership change research in other levels of
sport.
Conclusion
The current study systematically replicated work by Johnson et al.
(2013) investigating the impact of FBS football head coaching changes on
APR scores. The findings from the current study suggest FCS programs are
somewhat similar to FBS programs, but differ in a few key areas.
Similarities include lower APR scores in a year of a coaching change,
higher APR scores for the winningest teams, and average APR scores being
the most powerful predictor of APR scores in the year of the coaching
change. Differences were found regarding nature of change (positive vs.
negative) and type of hire (internal vs. external). Specifically, for
FBS programs the nature of change was not significant, but an internal
hire produced higher APR scores (Johnson et al., 2013). For FCS
programs, the type of hire was not significant, but positive coaching
changes produced higher APR scores. A final difference was that year of
coaching change was a significant predictor of APR scores for FBS
programs, but was not a significant predictor for FCS programs, thus
suggesting FBS programs have better acclimated to the APR metric over
time.
Practically, these results provide important information for
stakeholders of Division I college football. By knowing there is a
significantly lower APR score in years when a coaching change exists,
administrators and academic support personnel working at FCS
institutions can be deliberate in their efforts to provide additional
support during times of coaching transition. The extent of the support
can be measured against the nature and type of coaching changes. For FBS
student-athletes, increased assistance would be warranted when a new
external staff will be hired. This could include directing an interim
coach to be vigilant with their oversight of academic pursuits during
times without a head coach (Johnson et al., 2013). For FCS programs,
more support during a negative coaching change would be warranted. Given
that negative coaching changes are usually the result of a poor winning
percentage, and that teams with the lowest winning percentage tend to
have the lowest APR scores, it would be advantageous to start additional
efforts for support during an unsuccessful season and throughout the
process of hiring a new coach.
References
Amato, L., Gandar, J. M., Tucker, I. B., & Zuber, R. A. (1996).
Bowls versus playoffs: The impact on football player graduation rates in
the National Collegiate Athletic Association. Economics Education
Review, 15, 187-195.
Amorose, A. (2003). Reflected appraisals and perceived importance
of significant others' appraisals as predictors of college
athletes' self-perceptions of competence [Electronic version].
Research Quarterly for Exercise and Sport, 74, 60-71.
Bradley, B. (2005). The coach as guidance counselor. Coach and
Athletic Director, 75(3), 45-46.
Brown, G. T. (2005). Implementation of penalty structure triggers
new terminology, consequences, questions. NCAA News. Retrieved from
http: www.ncaa.org/.../Division-)-1/
APR+101+-+2-14-05+NCAA+News?pageDesign=Printer+Friendly+NCAA+News+And
+Updates
Brubaker, K. (2007). Coaching and teaching our athletes. Coach and
Athletic Director, 77(3), 30-33.
Chelladurai, P. (2009). Managing organizations for sport and
physical activity: A systems perspective. Scottsdale, AZ: Holcomb
Hathaway
Christianson E. (2004, April 29). NCAA board of directors adopts
landmark academic reform package. Retrieved from
http://www.ncaa.org/wps/ncaa7Content
Eidelson, R. (1997). Complex adaptive systems in the behavioral and
social sciences. Review of General Psychology, 1, 42-71.
Fabianic, D. (1994). Managerial change and organizational
effectiveness in major league baseball: Finding for the eighties.
Journal of Sport Behavior, 17, 135-147.
Fee, C. E., Hadlock, C. J., & Pierce J. R. (2006). Promotions
in the internal and external labor market: Evidence from professional
football coaching careers. Journal of Business, 79(2), 821-850.
Gagne, M., Ryan, R. M., & Bargmann, K. (2003). Autonomy support
and need satisfaction in the motivation and well-being of gymnasts.
Journal of Applied Sport Psychology, 15, 372-390.
Gamson, W. A., & Scotch, N. A. (1964). Scapegoating in
baseball. American Journal of Sociology, 70, 69-72.
Giacobbi, P. R., Roper, E., Whitney, J., & Butryn, T. (2002).
College coaches' views about the development of successful
athletes: A descriptive exploratory investigation. Journal of Sport
Behavior, 25, 164-180.
Gilmore, T. N. (2003). Making a leadership change: How
organizations and leaders can handle leadership transitions
successfully. New York, NY: Authors Choice Press.
Grusky, O. (1963). Managerial succession and organization
effectiveness. American Journal of Sociology', 69, 21-31.
Herold, D. M., Fedor, D. B., Caldwell, S. & Liu, Y. (2008). The
effects of transformational and change leadership on employees'
commitment to a change: A multilevel study. Journal of Applied
Psychology, 95(2), 346-357. doi: 10.1037/0021-9010.93.2.346
Hosick, M. B. (2009, May 6). Latest APR jumps three points.
Retrieved from http://www. ncaa.org /wps/ncaa?ContentID=49718
Hosick, M. B. (2010). Head coach APR database released. NCAA News.
Retrieved from http://www.ncaa.org/wps/wcm/connect/public/ncaa/resources/latest+news/2010+news+stories/august+latest+news/head+coach+a
pr+database+released
Hosick, M. B. (2011). DI board adopts improvements in academic
standards and student-athlete support. NCAA News. Retrieved from
http://www.ncaa.org/wps/wcm/connect/ public/ NC
AA/Resources/Latest+News/2011
/October/DI+Board+of+Directors+adopt+changes
+to+academic+and+student-athlete+welfare
Institute for Diversity and Ethics in Sport. (2012a). Keeping score
when it counts: Academic progress/graduate success rate study of 2012
NCAA division I women s and men's basketball tournament teams.
Retrieved from http://www.tidesport.org/Grad% 20Rates/
2012%20Women%27s%20Basketball%20Toumament%20Teams%20Study.pdf
Institute for Diversity and Ethics in Sport. (2012b). Keeping score
when it counts: Assessing the 2011-12 bowl-bound college football teams:
Graduation rates improve; Racial gaps persist. Retrieved from
http://www.tidesport.org/RGRC/2011/2011%20FBS%20Bowl% 20S
tudy%28FinalFinal%29.pdf
Johnson, J. E. (2012). An examination of factors impacting the
Academic Progress Rate of first-year football student-athletes at a
Football Bowl Subdivision institution. Indiana AHPERD Journal, 41(2),
11-16.
Johnson, J. E., Blom, L. C., Judge, L. W., Lee, D., Pierce, D. A.,
& Ridley, M. J. (2013). The impact of Football Bowl Subdivision head
coaching changes on NCAA Academic Progress Rate. Journal of Issues in
Intercollegiate Athletics, 6, 131-154.
Johnson, J. E., Wessel, R. D. & Pierce, D. (2012). The
influence of selected variables on NCAA Academic Progress Rate. Journal
of Issues in Intercollegiate Athletics, 5, 149171.
Judge, L. W., Petersen, J., & Johnson, J. E. (2013, April). An
examination of Division I Athletic-Academic Support Services,
Facilities, and Staffing. Paper presented at the College Sport Research
Institute Conference on College Sport, Chapel Hill, NC.
Lattman, M. (2008). The role of junior college coaches: More than
the sport. Coach and Athletic Director, 77(1), 110-112.
Lewis, D. K. (2004). Personal and situational bases for
coaches' causal attributions for the recovery outcome of injured
athletes. Unpublished doctoral dissertation, Michigan State University,
East Lansing.
NCAA. (2010). Academic Progress Rate penalties list. Retrieved from
http://ncaa.org/wps/
wcm/connect/public/NCAA/Academics/Division+I/APR+penalties+list
NCAA. (January, 2011). Examining the student-athlete experience
through the NCAA GOALS and SCORE studies. Presentation made at the NCAA
National Convention, San Antonio, TX. Retrieved from
http://www.ncaa.org/wps/wcm/connect/public/ncaa/ pdfs/2011/
examining+the+student+athlete+experience+through+the+ncaa+goals+and+score+ studies
NCAA.(2012a). National and sport group APR averages, trends and
penalties. Retrieved from
http://www.ncaa.org/wps/wcm/connect/public/ncaa/pdfs/2012/apr+2012+trends
NCAA. (2012b). NCAA Division I manual. Overland Park, KS: NCAA.
NCAA. (2013). NCAA Division I Head Coach APR Portfolio. Retrieved
from http://fs.ncaa. org/Docs/newmedia/public/rates/index2.html
Roach, K., & Dixon, M. A. (2006). Hiring internal employees: A
view from the field. Journal of Sport Management, 20, 137-158.
Rogers P. C. (2008). The quest for number one in college football;
The revised bowl championship series, antitrust, and the winner take all
syndrome. Marquette Sports Law Review, 18(2), pp. 286-307.
Sander, L. (2011, June 15). 22 elite college sports programs turned
a profit in 2010, but gaps remain, NCAA report says. The Chronicle of
Higher Education. Retrieved from http://
chronicle.com/article/22-Elite-College-Sports/127921/
Schilling, E. (2007). A game plan for mom and dad. Coach and
Athletic Director, 77(5), 4852.
Shapiro, B. J. (1984). Intercollegiate athletic participation and
academic achievement: A case study of Michigan State University
student-athletes, 1950-1980. Sociology of Sport Journal, 1, 46-51.
Sidman, M. (1960). Tactics of scientific research. Oxford England:
Basic Books.
Smith, R. A. (2011). Pay for play: A history of big-time college
athletic reform. Champaign, IL: University of Illinois Press.
Sperber, M. (2001). Beer and circus: How big-time college sport is
crippling undergraduate education. New York: Henry Holt.
Steinfeldt, M., & Steinfeldt, J. A. (2012). Athletic identity
and conformity to masculine norms among college football players.
Journal of Applied Sport Psychology, 24(2), 115129.
USA Today. (2013). USA TODAY sports college football coaches
'salaries database. Retrieved from
http://www.usatoday.eom/story/sports/ncaaf/2012/l
1/19/ncaa-college-football-head-coaches-salary-database/1715543/
James E. Johnson
Ball State University
David A. Pierce
Indiana University Purdue University
Daniel R. Tracy
Megan J. Ridley
Ball State University
Address correspondence to: James Johnson, Ball State University,
School of Physical Education, Sport, and Exercise Science, 2000 W.
University Ave., HP 223, Muncie, IN, 47306. Email: jejohnsonl@bsu.edu
Table 1
Descriptive Information for Variables Associated with Head
Coaching Change in FCS programs
Variable n % M SD
APR (year of coaching change) 120 100 922.87 40.78
APR (average) 120 100 929.69 33.33
Internal Hire 23 19.17
External Hire 97 80.83
Positive Coach Change 19 15.83
Negative Coach Change 101 84.17
Win % (year of coaching change) .36 .25
Win % (average) .46 .21
Table 2
Cross Tabulations for Nature of Change and Type of Hire in FCS
programs
Positive Change Negative Change
Internal Coach Hired 4 (3.3%) 19 (15.8%)
External Coach Hired 15 (12.5%) 82 (68.3%)
Note: Total coaching changes = 120
Table 3
Timing of FCS Head Coach Changes Cross Tabulated with Type of
Hire and Nature of Change
n % Positive Negative Internal External
Year
2003-04 9 7.5 0 9 2 7
2004-05 10 8.3 1 9 1 9
2005-06 22 18.33 2 20 1 21
2006-07 18 15 1 17 4 14
2007-08 16 13.33 8 8 4 12
2008-09 12 10 2 10 3 9
2009-10 19 15.8 4 15 5 14
2010-2011 14 11.66 1 13 3 11
Month
January 12 1 5 7 3 12
February 10 8.3 3 7 3 10
March 3 2.5 1 2 0 3
April 2 1 0 2 2 2
May 1 .08 0 1 0 1
June 5 4.2 0 5 1 5
July 15 12.5 0 15 2 15
August 3 2.5 0 3 2 3
September 3 2.5 0 3 3 3
October 1 .08 0 1 0 1
November 24 20 2 22 0 24
December 41 34.2 8 33 7 41
Table 4
Correlations for FCS Head Coaching Change Variables
APR Nature of Type
(year of Change of Hire
coaching APR (positive/ (internal/
change) (average) negative) external)
APR (year of coaching --
change)
APR (average) .65 ** --
Nature of Change -.2 * -.2 * --
(Positive/Negative)
Type of Hire (internal .07 .07 .02 --
/external)
Year of Change .02 -.14 -.15 -.14
Win % (year of 27 ** .26 ** _ 49 ** -.13
coaching change)
Win % (average .28 ** .24 ** -.37 ** -.1
2003-2011)
Win %
(year of Win %
Year of coaching (average
Change change) 2003-2011)
APR (year of coaching
change)
APR (average)
Nature of Change
(Positive/Negative)
Type of Hire (internal
/external)
Year of Change --
Win % (year of .01 --
coaching change)
Win % (average .12 .69 ** --
2003-2011)
* p < .05. ** p < .01
Table 5
FCS and FBS APR Scores Based on Winning Percentage
FCS Programs FBS Programs
(current study) (Johnson et al., 2013)
Winning % APR (Year of APR APR (Year of APR
Coach Change) (Average) Coach Change) (Average)
Top 1/3 914.55 938.22 * 949.19 * 949.78 *
Middle 1/3 920.97 929.7 932.07 940.65
Bottom 1/3 907.26 914.24 934.11 940.98
* = significantly higher APR scores than bottom 2/3 of teams in same
category
Table 6
Summary of Least Squares Regression for Variables Predicting APR
in the Year of a Coaching Change in FCS programs
Variable B Std. Error Beta t sig
APR (average) .51 .14 .43 3.56 <.01 **
Type of Hire (internal 5.17 8.19 .05 .63 .53
vs. external)
Nature of Change -3.26 9.87 -.03 -.33 .74
(positive vs. negative)
Year of Change 1.70 1.437 .09 1.182 .24
Win % (year of change) 15.9 16.39 .1 .97 .34
Win % (average) 17.17 16.5 .09 1.04 .3
* p < .05. ** p < .01
